Tracking Hyperscale AI Data Center Growth with Satellite Imagery
The advancement of AI is inextricably linked to its physical foundations. Hyperscale data centers power the world’s most advanced AI models, and the demand for these facilities continues to accelerate. Governments and frontier AI developers alike are investing tens to hundreds of billions of dollars into hyperscale data center buildouts, sparking debates on the economic, environmental, and geostrategic impacts of such efforts. On the international stage, AI is increasingly considered a source of economic growth and an asset to military capabilities, resulting in global competition, particularly between the U.S. and China. In addition, the successful operation of data center facilities requires high quantities of energy, affecting local communities and the environment. Therefore, from grassroots community impacts to global geopolitical dynamics, understanding developing data center capacities is emerging as a critical analytical challenge.
The speed at which AI models are advancing necessitates the rapid growth of its supporting data center infrastructure. In tracking data center construction and operation, what can remote sensing—particularly, the use of electro-optical (EO) satellite imagery—meaningfully illustrate about data center development, and what can policymakers learn from this data moving forward? While multiple research organizations have utilized satellite imagery to track construction and ongoing activity at hyperscale data center facilities, no baseline assessment of its benefits and limitations is available in the public sphere. This research addresses that gap with two case studies: Khazna’s Ajman data center facility in the United Arab Emirates (UAE) and xAI’s Colossus data center site in Memphis, Tennessee. These examples reveal inconsistencies between commercial announcements and on-the-ground activity visible in satellite imagery, showing that official construction timelines, infrastructure permits, and company publications cannot be relied upon exclusively for insights into data center development. Satellite imagery provides a layer of independent verification that can be integrated to overall assessments of data centers, and this study illuminates where this imagery adds insight to inform future AI policy efforts. In particular, as discussions of international AI governance evolve, satellite imagery can be integrated as one verification method in a wider toolkit supporting the monitoring and implementation of future AI-focused agreements.
Acknowledgements
This work draws on Epoch AI’s Frontier Data Centers Hub project—an effort to which the author contributed as a Supervised Program for Alignment Research (SPAR) Fellow—and builds on information originally collected by other members of the project that was further expanded upon for the research. This report was written for and funded by the 2025 GovAI Fall Fellowship.
This report was edited for publication by the Federation of American Scientists, and the author would like to thank Matt Korda and Dr. Oliver Stephenson for their contributions and support in bringing this publication to completion.
Introduction
Data centers are an integral part of the AI ecosystem and impact finance, the environment, and international strategy. The total expected financial investment into data centers spans billions of dollars—and is expected to hit the trillions. The scale of these facilities is also massive: in July 2025, Meta CEO Mark Zuckerberg announced plans for the construction of the Hyperion data center, expected to have a one-gigawatt capacity and cover a land space almost the size of Manhattan. Thus, data centers have significant economic and industrial ramifications as they continue to be developed. The sheer size and scale of these facilities requires high amounts of energy, with these needs driving environmental concerns regarding the overall impact of AI. The Electric Power Research Institute in particular forecasts that data centers will use 9-17% of total U.S. electricity by 2030 (an increase from 4-5% today). In addition, on the international stage, high levels of AI technology have increasingly been seen as a marker for technological supremacy, with countries prioritizing not only the development of AI models themselves but their integration into military systems as national priorities. All of these elements have caused a degree of public backlash regarding data center construction across the globe, reflecting local concerns about energy needs and fears surrounding the unregulated advancement of AI having significant geopolitical destabilizing effects.
Despite this extensive media, political, and analytical attention, public information on data center facilities is often incomplete or opaque, generating uncertainties about their pace of development. In this context, satellite imagery is one tool for monitoring the development and operation of these sites as a form of independent verification. This report thus examines the role satellite imagery can play in tracking data center infrastructure development and operation using electro-optical (EO) imagery and two case studies: Khazna’s Ajman data center in the UAE and xAI’s Colossus data center in Memphis, Tennessee.
The following sections outline an analytical framework, assess observable indicators in satellite imagery, and apply them to these two case studies. The work compares visible features on imagery to commercially announced milestones, informing future forecasting efforts on data center buildout timelines and identifying areas where satellite-derived insights are limited or indeterminate. Recommendations for next steps are included in the conclusion of the report, and outline the role satellite imagery of data centers can play in commercial, government, and local communities. At the international level, discussions of potential multilateral AI governance frameworks are ongoing; satellite imagery is particularly suited to serve as one complementary layer to independently verify commitments under future AI-related agreements.
Comments on Existing Literature
Data center locations are not available in one comprehensive, publicly-available database. Instead, corporations themselves may choose to disclose the locations of their data centers through announcements or publications (i.e. Google’s Data Center directory). Other aggregators such as Data Center Map, Baxtel, and Pacific Northwest National Laboratory’s Open Source Data Center Atlas independently use permits, news articles, and other open source information to collect information and display data center locations. However, the completeness and accuracy of these listings vary. Thus, satellite imagery is one additional layer of independent information that can provide credibility to existing geographical information on data center facilities.
Several organizations have expressed interest in utilizing satellite imagery to track data center construction and operation. Notably, Epoch AI’s Frontier Data Centers Hub uses satellite imagery, permit data, and other open sources to track hyperscale data center buildout within the U.S.. In addition, Semianalysis has a data center analytical model that also relies on satellite imagery to provide commercial and industrial insights on data center activity for private companies. However, the Semianalysis model is not publicly available and is instead sold to companies for forecasting, business intelligence, energy usage, construction tracking, and environmental assessment purposes. Epoch AI’s methodology is publicly available, but does not detail how specifically to analyze the imagery itself, focusing on broader applications for energy usage calculations.
Therefore, while there is a clear interest in this type of analysis, there is still ample space to expand upon the impacts satellite imagery analysis can have in the AI field. This research provides a broad overview of satellite imagery’s capabilities for hyperscale data center tracking, providing an assessment of its benefits and limitations.
Satellite imagery has also long been integrated into media reports as a flashy and attractive visual for news articles, and data center construction is no exception. In October 2025, Business Insider published a high-level article illustrating the growth of U.S.-based data centers on satellite imagery. This displays an appetite in the public sphere for visual representations of data center expansion. Articles such as these lack the detail of traditional imagery analysis, and could benefit from additional context that formal analyses can provide. More generally, satellite imagery data is useful for a variety of stakeholders: governments can use satellite imagery of data centers to inform policy decisions regarding AI-related infrastructure expansion in the face of perceived international competition and strain on electricity grids. Frontier AI labs can integrate imagery assessments to stay informed of the progress of their (and their competitors’) data center development. In addition, the public can harness satellite imagery to increase transparency of local development projects and ensure that companies are adhering to their permit terms and legal obligations. This research in particular addresses the latter, displaying how satellite imagery adds transparency for verification of corporate data center operational activities.
Among researchers, some skepticism remains about the use of satellite imagery to track data center construction. In late 2024, Lennart Heim and Konstantin Pilz argued that satellite imagery of data centers can be limited in its usefulness to track the proliferation of computing power. Their assessment took a conservative view of satellite imagery’s utility, stating that elements such as chip quantity, chip type, and power consumption levels are difficult to extract. While it is true that EO imagery cannot directly observe GPUs or other indoor hardware, this research argues that satellite imagery is valuable for independent verification of visible infrastructure and construction progress, and for making broad data center capacity estimates.
Methodology
For this research, the author first assesses the principles underpinning the analysis of electro-optical (EO) imagery, outlining how it could reveal physical infrastructural patterns at data center facilities. The author then applies these concepts to two case studies: the Ajman Data Center in the UAE (to track buildout and construction) and xAI’s Colossus data center facility in Memphis, Tennessee (to illustrate feature identification). Note that this research focuses primarily on hyperscale data centers (i.e. those with at least 50-100 MW capacity). These types of facilities have a large geographic footprint—often spanning tens to hundreds of football fields—and are easier to visualize in publicly-available EO imagery than medium- to small-scale data centers. These might be housed in existing buildings or urban centers and therefore might not have the clearly-identifiable visual signatures expected at hyperscale data center campuses.
Opportunities and Challenges of Utilizing Satellite Imagery for Data Center Monitoring
Electro-Optical Imagery Basics & Uses
Overview of Electro-Optical (EO) Imagery
Electro-Optical (EO) imagery is satellite imagery that is collected via sensors that detect sunlight signatures in the visible and near-infrared portions of the electromagnetic spectrum. Therefore, depending on the collection parameters and post-processing, finished and published EO imagery products will generally look like a “photograph” from above to the naked eye. EO imagery is widely available commercially, both free of charge and for purchase. Commercial providers such as Airbus, Planet Labs, and Vantor are commonly-used sources of paid satellite imagery, with prices ranging from roughly $15 to several hundreds of dollars per square kilometer, depending on the area of interest, resolution, and provider. Free sources of satellite imagery can be acquired via platforms such as Google Earth Pro, European Space Agency’s Copernicus Sentinel program, and NASA Earthdata/Landsat. This is generally archival imagery or lower-resolution than resources available through paid providers. For more information on satellite imagery analysis basics, see FAS’s prior publication on the topic, A Guide to Satellite Imagery in the Nuclear Age: Assessing China’s CFR-600 Reactor Facility.
The spatial resolution of an image determines its level of granularity and detail. A high-resolution image is considered anything better than one-meter resolution: in other words, the length of each pixel represents one meter.At ten-centimeter resolution, therefore, one pixel represents ten centimeters on a side, corresponding to a 10-centimeter by 10-centimeter ground footprint.
Fig. 1. Comparison of satellite image resolutions at different ground sampling distances. Source: Wikimedia Commons, CC0 Public Domain.
At a typical industrial facility—for example, a nuclear reactor, water treatment plant, or coal processing facility—the following features can be identified in high-resolution satellite imagery:
- The overall footprint of the facility;
- Buildings;
- Security perimeters and entrances;
- Roads;
- Vehicles;
- Terrain changes and construction; and
- Other large-scale infrastructural features.
Fig. 2. Example of satellite imagery used to identify features at China’s CFR-600 nuclear reactor facility. Source: A Guide to Satellite Imagery in the Nuclear Age: Assessing China’’s CFR-600 Reactor Facility, Federation of American Scientists, December 2025
Figure 2 shows an example of EO satellite imagery being used to identify infrastructure at a nuclear reactor. Some of these principles can also be applied to monitoring data center development and construction, as discussed in the following sections.
Data Center Applications
Feature Identification
EO imagery can be used to identify specific infrastructural features of interest at a data center facility. The following generalities have been drawn from U.S.-based hyperscale data centers operated by frontier AI developers, and these features can be identified on imagery of at least 50-centimeter resolution. Some features from international facilities have also been included.
While the features outlined below display several visual patterns to make identification on satellite imagery feasible, it is vital to note that these features and their arrangements are highly variable, and the patterns outlined in this section are guidelines rather than strict rules. This is discussed further in the Limitations section below.
Appendix A contains a summary of the key features of a data center, the satellite resolution necessary to resolve them, their distinguishing elements, and potential obstacles.
Main Data Halls
The main building of a hyperscale data center facility houses the equipment used for AI applications, including racks, servers, and chips. These are generally sizable buildings that have a large geographic footprint. On satellite imagery, they appear as the “main” and largest building in an area. The main buildings may have white or pale-colored rooftops for reflectivity and temperature control as well. They also commonly have flat or low-sloped rooftops for ease of installation of HVAC equipment on top.
Fig. 3. Main data hall, du Data Center Silicon Oasis, Dubai, UAE, July 22, 2025. Coordinates: 25°7’58.32″N, 55°23’30.16″E. Image: Google © 2025 Airbus
Fig. 4. Meta Prometheus, New Albany, Ohio, U.S.A, July 2, 2025. Coordinates: 40°4’6.24″N, 82°45’1.34″W. Image: Google © 2025 Airbus
Fig. 5. Alibaba Zhangbei Data Center, Zhangjiakou, China, November 7, 2024. Coordinates: 41°11’34.73″N, 114°41’53.37″E. Image: Google © 2025 Airbus
Data center halls are identifiable both by their size and their surrounding infrastructure (discussed in the following sections). Note that, for a data center site, if a building takes up a similar large footprint but only trucks or loading areas are visible around its edges, it is likely a warehouse instead and not an IT-specific building.
Support/Administrative Buildings
Administration buildings will generally be built closer to an entrance or parking area at a data center site. They will have a much smaller footprint than the main data hall but will be located in its general vicinity.
On-site Grid Power
Substations and switchyards manage grid power for distribution to the rest of the data center facility. Substations—used for voltage transformation and power distribution—are generally distinguishable by the presence of power transformers (see Figure 6). Switchyards, on the other hand, reroute power without changing the voltage, and will generally not have transformers visible. In the U.S., they are generally identifiable by their appearance as open, flat, gray concrete lots with rod- or pipe-like features, thin metal infrastructure, and adjacent blocks. Expected features include steel gantries (frames providing stable support for other equipment), transformers (boxes used for electrical transfer), and busbars (metallic strips used for conduction and energy transfer). Thick cables or rod-like features might be visible connecting a substation to the main data hall, but this is not always necessarily the case, as both overground and underground feed systems could be in use. In addition, in higher-resolution imagery, power lines and their associated towers connecting to the substations could be visible.
Fig. 6. Granite Point Data Center Campus Substation, Bluffdale, Utah, U.S.A, June 1, 2025. Coordinates: 40°28’45.18″N, 111°54’26.34″W. Image: Google © 2025 Airbus
In some locations, the on-site grid power will be housed within a dedicated building. Indoor substations could indicate that they are gas-insulated switchgear (GIS) substations, medium-voltage (MV) distribution substations with air-cooling systems, indoor transformer halls (housed inside for noise reduction), or a combination of these.This is particularly true for hot and dry countries such as the UAE; substations in these environments could be located indoors for temperature control purposes. Therefore, substations and switchyards will not necessarily look the same at data centers around the world.
Main and Backup Power Generation
Data centers require significant amounts of energy to operate, typically ranging from tens to hundreds of megawatts (MW), with gigawatt (GW)-scale facilities now emerging. The high environmental cost of data center operation has placed them in the spotlight: from grassroots organizations to the United Nations, global communities have raised myriad concerns regarding the strain data center facilities place on electricity systems and water resources.
For hyperscale data center facilities, power generation infrastructure ensures that the main buildings have consistent access to adequate power. Main power generators come in various designs, but are often natural gas turbines (see Figure 7). They will appear larger than backup generators, and will have cylindrical exhaust stacks which appear circular on imagery. However, not all data center facilities will show a main power generation source. In some cases, the power plants supplying energy are offsite and not necessarily visible at the facility itself.
Fig. 7. Screen capture of OpenAI Stargate Abilene facility main power infrastructure outlined in pink within the Epoch AI Satellite Explorer, image dated September 26, 2025. Coordinates: 32°30’16.17″N, 99°46’31.20″W. Image: Epoch AI Satellite Explorer © 2025 Airbus via Apollo Mapping
Backup generators—often diesel—provide emergency power to the main hall if the main power fails. These generators will usually be numerous and in close proximity to the main data hall. They will appear as smaller, thinner rectangles arranged in rows. They will not have fan features (and therefore can be distinguished from cooling infrastructure). Their regular spacing and alignment along the main building perimeters are distinctive. See Figure 8 for an example.
Fig. 8. Screen capture of Amazon Madison Mega Site facility backup power outlined in purple within the Epoch AI Satellite Explorer, image dated October 4, 2025. Coordinates: 32°35’46.34″N, 90°5’40.50″W. Image: Epoch AI Satellite Explorer © 2025 Airbus via Apollo Mapping
Cooling Infrastructure: Chillers and Cooling Towers
Cooling infrastructure is vital for removing heat from data halls. Chillersactively remove heat from the data center cooling loop. Cooling towers are heat-rejection systems used to release heat to the atmosphere. Chillers and coolers will generally be colocated at a data center site; however, not all data centers require both. For example, water-cooled systems typically require cooling towers to release heat, while air-cooled systems do not, instead blowing air at a refrigerant-filled coil for heat dissipation. Cooling towers are distinguishable by their arrangement in rectangular or square clusters with large circular fan openings (see Figure 9). These fans are distinct on high-resolution imagery, spanning several feet in diameter.
Fig. 9. Examples of cooling towers. Left: Red Oak Google Data Center, Red Oak, Texas, U.S.A, August 24, 2025. Coordinates: 32°32’22.49″N, 96°47’31.47″W. Image: Google © 2025 Airbus
Middle: AirTrunk Iskandar Puteri, Iskandar Puteri, Malaysia, April 29, 2025. Coordinates: 1°26’6.45″N, 103°36’3.43″E. Image: Google © 2025 Airbus
Right: Switch Citadel Campus, Sparks, Nevada, U.S.A, June 19, 2025. Coordinates: 39°32’19.60″N, 119°26’12.77″W. Images © Airbus via Google Earth Pro. Image: Google © 2025 Airbus
Chillers, on the other hand, will usually be boxy or container-like in shape (See Fig. 10). Air-cooled chillers could show smaller circular fans arranged in a grid, but these are smaller in size and generally only distinct in ultra high-resolution imagery (higher than 30-centimeter resolution; see Table A for further details). Water-cooled chillers are generally housed indoors, and therefore would not be visible in EO imagery, but this is not universally the case at every facility.
Fig. 10: Examples of chillers. Left: Digital Realty ORD10 / Lakeside Technology Center, Chicago, Illinois, U.S.A, August 7, 2025. Coordinates: 41°51’14.35″N, 87°37’5.70″W. Image: Google © 2025 Airbus
Middle: Equinix LD5, Slough, United Kingdom, August 16, 2025. Coordinates: 51°31’2.18″N, 0°37’0.74″W. Image: Google © 2025 Airbus
Right: QTS Microsoft Fayetteville Data Center, Fayetteville, Georgia, U.S.A, October 31, 2025. Coordinates: 33°27’7.89″N, 84°31’18.58″W. Image: Google © 2025 Airbus
Construction and Change Over Time
Crucially, EO imagery can be used to track the construction of a data center over time. Using several satellite images of a data center collected over the course of months or years, the footprint of land use can be monitored chronologically. EO imagery is well-suited for tracking change over time, as progress is often significant enough to be seen on medium- to high-resolution imagery available publicly. Construction milestones are particularly useful for assessing future buildout: understanding the time required to build a data center can assist in planning for energy needs and estimating energy capacities. Case Study 1, the Ajman Data Center in the United Arab Emirates, is an example of satellite imagery used to track chronological data center expansion.
Case Study 1. Tracking Construction at the Ajman Data Center, UAE
Background & Context
The Ajman Data Center in the United Arab Emirates is one example of ongoing construction of a hyperscale data center facility. In October 2024, Khazna Data Centers—the largest data center construction firm in the Middle East and North Africa—announced plans to build the region’s largest data center facility in Ajman. Official announcements noted that the Ajman Data Center would be the first AI-optimized hyperscale data center facility in the region, with a footprint of 100,000 m2 and 20 data halls, each with 5 MW capacity (bringing the total capacity to 100 MW)., This is a significant achievement for the UAE, illustrating its capacity to manage large-scale AI compute needs, and is of particular interest given the United States Commerce Department’s November 2025 decision to allow the sale of advanced semiconductor chips to the UAE. The development of AI-focused data centers in the region, thus, highlights the potential integration and use of this exported hardware abroad.
Ajman is a well-documented data center facility, and there are several EO satellite images available in open sources capturing its construction over time. Using imagery from Google Earth Engine and Copernicus, this case study tracks the buildout of the Ajman Data Center over the course of 2024-2025, providing a comparison of commercial construction milestone announcements to real-world, on-the-ground data.
Chronological Assessment
Publicly-available imagery illustrates the development of the Ajman Data Center from May 2024 to August 2025. Figure 11, below, shows the milestones visible in satellite imagery in comparison to official commercial announcements by Khazna.
Fig. 11. Timeline of the Ajman Data Center construction compared to Khazna commercial announcements. Images: European Space Agency Copernicus Sentinel Data © 2024 and 2025 / Google © 2024 and 2025 Airbus
Fig. 12. Progression of construction at the Ajman Data Center. Images B and H © European Space Agency Copernicus Sentinel Data 2024 and 2025. Images A, C, D, E, F, G: Google © 2024 and 2025 Airbus
The above images (Figure 12) illustrate the chronological development at the Khazna Data Center with significant milestones visible. Below is a detailed evaluation of this change over time, image by image, highlighting the significant discrepancies between Khazna’s official announcements and the imagery.
Fig. 13. Ajman Data Center site, May 3, 2024. Coordinates: 25°23’22.57″N, 55°34’44.82″E. Image: Google © 2025 Airbus
Figure 13 shows the location of the Ajman Data Center facility. The infrastructure features to the far east of the site are an ongoing housing development. In this image, two construction clearings with material, vehicles, and equipment are visible in the central and southeastern portions of the image. The Etihad Water and Electricity Company (EtihadWE) Power Station is housed in the white-roofed rectangular facility to the northwest. At this point in time, the only signs of development are the temporary white storage units and the two fenced construction yards. Note that no official announcements had yet been made confirming this was the location of an upcoming data center.
Fig. 14. Ajman Data Center site, October 14, 2024. Coordinates: 25°23’22.57″N, 55°34’44.82″E. Image © European Space Agency Copernicus Sentinel Data
Fig. 15. Ajman Data Center site, November 16, 2024. Coordinates: 25°23’22.57″N, 55°34’44.82″E. Image: Google © 2025 Airbus
On October 15, 2024, Khazna formally announced the start of the Ajman Data Center project. According to the company, by October 15, the project had “broken ground” and was in the process of construction. “Breaking ground” is a subjective term, but typically refers to the start of official construction at a site, marked by the initial soil excavation and preparation. This commercial announcement is accurate: In Figure 14, October 14, 2024 Copernicus Sentinel-2 imagery (provided free of cost at lower ten-meter resolution via the Sentinel Hub) shows that the perimeter outline of the facility has been developed, and the dark brown linear feature within this perimeter is indicative of ground disturbances and corresponds accurately with the official announcement. Google Earth Pro shows higher-resolutionimagery for free, and the imagery available a month later (Figure 15) illustrates further detail in the development of the site, revealing that the dark linear feature from October 14, 2024 shows initial stages of a paved road. (Google Earth Engine does not display exact image resolution.)
Fig. 16. Ajman Data Center site, April 3, 2025. Coordinates: 25°23’22.57″N, 55°34’44.82″E. Image: Google © 2025 Airbus
On April 21, 2025, a Khazna press release reported that the “steel structures” of the data center were complete. However, this does not appear to be accurate. As of April 3, 2025, it would not be possible for the steel structures to be finished. Figure 16 shows that the foundations of the building have not yet been fully laid, with large stretches of beige terrain visible to the east. Therefore, the steel structures were not finished at this point in time. The eastern portion of the building hall is clearly still in development. This illustrates that the press release was optimistic or strategically timed to signal that the construction was ahead of true progress.
Fig. 17. Ajman Data Center site, May 28, 2025. Coordinates: 25°23’22.57″N, 55°34’44.82″E. Image: Google © 2025 Airbus
Fig. 18. Ajman Data Center site, June 10, 2025. Coordinates: 25°23’22.57″N, 55°34’44.82″E. Image: Google © 2025 Airbus
Fig. 19. Ajman Data Center site, August 22, 2025. Coordinates: 25°23’22.57″N, 55°34’44.82″E. Image: Google © 2025 Airbus
Significant construction progress is visible from May to August of 2025 (Figures 17, 18, and 19). Figures 17 and 18 show that the dark black foundational layer—likely asphalt or a waterproofing membrane—was overlaid with paler material, which are likely slabs of concrete. It is possible that Khazna defines “steel structures” as something different, and that the terminology may not refer to the thin metal support features that are visible on top of the foundational layer. However, if these metallic features are indeed “steel structures,” then they were only nearing completion around August 2025—far later than the April 2025 announcement. The steel structures may not even be entirely complete by the August 2025 image either, considering the easternmost portion of the main building still has pale foundational material entirely exposed.
Fig. 20. Ajman Data Center site, November 13, 2025. Coordinates: 25°23’22.57″N, 55°34’44.82″E. Image © European Space Agency Copernicus Sentinel Data.
As of November 13, 2025, the rooftop shows significant progress.
The announced startup timelines for the site also changed over the course of construction. In the October 2024 announcement, the company stated that the facility would be “operational by Q3 2025” and “completed within 15 months,” placing the startup date sometime between July and September 2025. With most data center facilities taking between 18-30 months to complete (from “concept to commissioning”), this is an aggressive timeline. The satellite imagery also suggests that this compressed timeline is not feasible, since the construction was not on track with a Q3 2025 start date.
By the April 2025 update, Khazna stated that “the initial project phase is due for completion in December 2026.” While the meaning of “initial project phase” is subjective, for general completion of the facility, this is a much more realistic timeline and is accurately represented by changes in the imagery. This suggests that companies may be more marketing-focused at the beginning of construction, changing their estimations as projects progress.
Case Study 2. Feature Identification at xAI’s Colossus Data Center Facility
Background & Context
Epoch AI’s Frontier Data Centers Hub, released November 2025, analyzes over a dozen hyperscale data center facilities to track energy usage, construction timelines, and compute. One feature, the Satellite Explorer, provides high-resolution EO imagery annotations of critical infrastructure and tracks change over time. The author of this report contributed to the satellite imagery analysis and annotations for the Satellite Explorer, and this study selects one of the Hub’s data center facilities—xAI’s Colossus installation in Memphis, Tennessee—and discusses methods for feature identification, outlining how this methodology can be applied to other large-scale data center sites as well.
xAI’s Colossus supercomputer is run by a hyperscale data center facility in Memphis, Tennessee. The facility boasts a capacity of 200,000 GPUs and is one of the largest operational AI training clusters worldwide. At this size and scale, this installation is ideal for identification of expected features at completed hyperscale data centers. While this example on Epoch AI’s hub can also be used for tracking construction and buildout (similarly to Ajman), because the facility is operational, it is used in this report specifically for feature identification purposes.
Feature Identification
Fig. 21. Screen capture of xAI Colossus Data Center facility annotations within the Epoch AI Satellite Explorer, image dated October 19, 2025. Coordinates: 35°3’35.09″N, 90°9’22.33″W. Image © Airbus via Apollo Mapping.
Main Building
The main building at this facility is determined by its central location in the image, flat white reflective rooftop, and large footprint in comparison to the surrounding infrastructure. While this type of main data hall might not be immediately identifiable on its own, its proximity to the other infrastructure and its size indicates its status as the main building.
On-site Grid Power
Three substations are visible at this facility, outlined in Figure 21 in orange. Each of these three shows the expected visual signatures: they are rectangular, gray concrete lots with thin steel structures (that may throw stark, visible shadows across the lot if the sun is not directly overhead). These are the gantries, transformer blocks, and busbars. This identification is further validated by open sources (Figure 22): the Memphis Light Gas and Water (MLGW) publicly reported the location of two of the substations by using Google Earth satellite imagery and making its own annotations, outlining the substation locations.
Fig. 22. Memphis Light Gas and Water (MLGW) imagery annotation and confirmation of xAI Colossus substation locations, July 9, 2024. Image date: April 17, 2023. Coordinates: 35°3’35.09″N, 90°9’22.33″W. Image: Google © 2025 Airbus
While it has its limitations, satellite imagery can be used for estimating power generation and capacity in some instances: for example, in July 2024, MLGW annotated Google Earth satellite imagery (Figure 22), revealing that the substation to the northeast of the facility had an 8 MW capacity and that the “future” substation located to the northwest would have 150 MW capacity. In Figure 21, a third substation located to the southeast is roughly the same size as the northwest substation. Therefore, it can be inferred that the third substation can also provide approximately 150 MW.
In May 2025, MLGW further announced that “xAI is receiving 150 MW of power at the Paul Lowry site. xAI requested an additional 150 MW at this location, for a total of 300 MW. A substation is being built to accommodate this request … MLGW anticipates that xAI will receive an additional 150 MW in Fall of 2025.” The announcement notes that 8 MW are being drawn from an “adjacent, pre-existing MLGW substation” (to the northeast) and 142 MW are being generated from the “newly constructed” substation (to the northwest). Therefore, the initial conclusion that the third southeastern substation has a capacity of approximately 150 MW was broadly accurate.
In this situation, EO imagery allows for the identification of the third substation and assists in energy estimations. However, it is only marginally accurate, since the northwestern substation is providing 142 MW and the southeastern substation will provide 150 MW. Therefore, exact estimations may not be feasible using satellite imagery.
On-site Main and Backup Power
Fig. 23. Close-up of main and backup power infrastructure annotations within the Epoch AI Satellite Explorer, image dated October 19, 2025. Coordinates: 35°3’26.33″N, 90°9’22.48″W. Image © Airbus via Apollo Mapping.
Fig. 24. Close-up of main and backup power infrastructure annotations within the Epoch AI Satellite Explorer, image dated October 19, 2025. Coordinates: 35°3’43.66″N, 90°9’21.15″W. Image © Airbus via Apollo Mapping.
Main (or on-site) power generators and backup generators will have differing visual indicators, as seen on Figures 23 and 24. For the Colossus facility, the main power generators are natural gas turbines. These types of turbines are distinguishable by the appearance of cylindrical exhaust stacks. In addition, just north of the exhaust stacks, square gray structures show the turbine air inlet filters. The turbines are also characterized by mushroom-cap-shaped features, visible north of the air inlet filters with appearances as white-blue horizontal structures with curved shadows to the north. These main power generators will also generally be grouped together and arranged in neat rows all facing the same direction.
Fig. 25. Gas turbine and backup power units visible in aerial imagery taken from the northeast looking southwest, image dated March 31, 2025. Image © Steve Jones Flight by Southwings for SELC.
In contrast, backup generators (outlined in purple in Figures 23 and 24) will show different visual indicators. They are smaller in size than the main power turbines, often rectangular in size and arranged in neat rows. They appear as pale-colored, container-like structures, and have no strong visual indicators. Long, rectangular white boxes in satellite imagery suggest the presence of backup generators such as these.
Identification of these gas turbines, over the course of 2025, was of particular use and interest to the Memphis local community. In 2024, xAI was offered permits for installing 15 temporary natural gas turbines at the site. However, by April 2025, 35 turbines were visible at the site, captured in aerial imagery. (Note that this was not satellite imagery, but imagery taken with an overhead airplane. This flyover was commissioned specifically for this purpose. However, satellite imagery can still be used to detect the quantity of turbines at the site.) These 35 turbines were in clear violation of the permitting rights that had been allotted to xAI. Following public complaints, xAI began dismantling the extra turbines. By July 2025, xAI was approved to have 15 permanent gas turbines on site. In Figure 21, dated October 19, 2025, 12 turbines are visible, implying that xAI is adhering to the legal allowance. This instance reveals how imagery can be utilized for ensuring that companies are complying with their legal commitments.
Cooling Infrastructure
Epoch AI’s recent work focuses particularly on using cooling infrastructure to make estimations on total cooling capacity and power consumption. This methodology has been demonstrated in the Frontier Data Centers Hub, which integrates satellite imagery, permitting documents, and company disclosures to model power, compute, and cost metrics across AI data centers.
Before calculations on energy output can be made, the cooling infrastructure must be identified and counted. Figure 21 shows the xAI Colossus cooling infrastructure outlined in blue. Close-ups are provided below in Figures 26 and 27.
Fig. 26. Close-up of cooling tower infrastructure annotations within the Epoch AI Satellite Explorer, image dated October 19, 2025. Coordinates: 35°3’32.30″N, 90°9’18.87″W. Image © Airbus via Apollo Mapping.
Fig. 27. Close-up of air-cooled chiller infrastructure annotations within the Epoch AI Satellite Explorer, image dated October 19, 2025. Coordinates: 35°3’26.41″N, 90°9’29.79″W. Image © Airbus via Apollo Mapping.
The Colossus data center facility is unique in that both cooling towers and air-cooled chillers are used. The cooling towers are more easily identifiable as large rectangular blocks with multiple circular fan features on top arranged in a row. The air-cooled chillers are often smaller in size than the cooling towers. They are also arranged in neat rows and also have many small fans on top (as opposed to the cooling towers, which show several much larger fans).
Counting the blocks of chillers and the cooling towers can then be used, as per the Epoch AI calculations, as inputs for quantitative power-capacity estimates. Epoch’s empirical models convert observable cooling infrastructure—such as the number and configuration of air-cooled chillers or cooling towers—into estimates of total cooling capacity and, from there, total facility power. For the Colossus facility, Epoch’s energy calculations and estimates range from approximately 200-500 MW, which is in line with the published expected capacity of 300 MW. While these estimations are broad and highly variable, Epoch’s work demonstrates how satellite imagery of cooling equipment can be used to infer broader operational characteristics of large AI data centers, even when direct disclosures are limited.
Support and Administrative Infrastructure
There is a large parking lot located to the west of the facility, visible as a large gray paved area with multicolored vehicles parked in lines. Just south of this parking lot, there is an additional unpaved open lot with white infrastructure, which are likely mobile container units. This area is likely being used for storage or other construction-related purposes. This is further evidenced by the open nature of the lot and the other mobile features parked in the area. This is notable due to the fact that, even as the building is operational, construction or upgrading activity may still be ongoing. Therefore, even after an official operational date is announced and comes to pass, this does not indicate the end of development activity at a site overall.
Key Takeaways & Analysis
Satellite imagery provides ground truth of construction progress, particularly when compared to official publications.
Commercial announcements cannot be solely relied upon for accuracy regarding the actual development of a data center site. Accuracy may be low due to changing construction plans, unscheduled delays, or intentional obfuscation. Early announcements can also function as publicity events. In the case of Ajman, illustrating that the “steel structures” claim is not consistent with ground truth helps to establish Khazna’s credibility and the reliability of future construction milestone announcements. In addition, these kinds of claims show that there could be a misalignment between understood definitions of “steel structures.” While they might refer to overall support structures, they might also be referring to other elements not yet visible at the site. This analysis ties the imagery observations to wider industry norms and shows that standard industry construction timelines are more realistic. Satellite imagery verification, therefore, reinforces these basic industry norms.
EO imagery is generally better for tracking construction (rather than ongoing activity) at data centers.
In the open source space, historical EO imagery is easier to obtain than new, up-to-date images, with providers making older imagery more readily available. Therefore, EO imagery is most easily integrated in the open source space to track changes over time. It is not well suited for identifying data centers from scratch or to track ongoing operations because A) the imagery is more difficult to access, and B) there are fewer large-scale infrastructural changes that occur at a data center facility once it has been completed. Construction changes take place over a wide stretch of land, making visible changes easy to track. Once a facility is operational, signs of ongoing activity might be limited to the presence of vehicles, which are by themselves not significant indicators of activity; they simply illustrate that staff is likely on site. Therefore, EO imagery is useful for tracking the construction and buildout of data centers over time.
Construction progress reveals internal building mechanisms and the schematics of data center halls.
In completed data center buildings, the internal arrangement of AI-related infrastructure is not visible. However, EO imagery can capture the foundations of the building in process of being built during construction. From this, it could be possible to estimate how many rows of racks will exist in the finished product. This could be used for estimations of total data center capacity.
This analytical process establishes a repeatable satellite imagery verification method for data center buildout.
Having satellite imagery for multiple construction stages shows, for future data center buildouts, that tracking the following infrastructural elements will give indicators as to the progress of the overall site. The established system for trackable developments is as follows:
- Site clearing;
- Data hall planning;
- Foundation work;
- Steel framing;
- Roofing;
- Cladding;
- HVAC, substations, and cooling; and
- Commissioning.
Analysts can use the tracking of these features to estimate how far along a construction project might be.
Satellite imagery should be analyzed in conjunction with other publicly-available information sources.
Satellite imagery does not exist in a vacuum; open sources can and should be utilized to support ground data. Satellite imagery, in the open source community, is the most useful when compared to other sources—particularly ground photos, videos, blueprints, maps, and other publicly-available satellite imagery. This is illustrated by the integration of open sources to the two case studies: for Ajman, the comparison of Khazna’s publications to satellite imagery exposed inconsistencies between commercial announcements and on-the-ground activity. In the Colossus case, publicly-available permit data outlined the legal limits of turbine installation, and Memphis Light Gas and Water’s open plans for development of the site, when compared to imagery, illustrated the potential size of the infrastructure and allowed for a general estimation of power capacity. Thus, open sources provide critical context and allow analysts to develop a deeper understanding of features and activities observed in satellite imagery.
Satellite imagery of data centers reveals some repeating patterns in expected infrastructure. Feature identification provides a means of independently verifying commercial activity.
EO satellite imagery is ideal for tracking large-scale construction and can be used to identify specific infrastructure features at data centers. These features—from gas turbines to switchyards to chillers—tend to have specific, visible patterns on satellite imagery that can be applied to the analysis of other data centers. While these features will not be the same at every data center, this guide serves as a starting point for analysis. The case of the xAI Colossus site is illustrative of the potential impact of imagery for tracking commercial activities. The unpermitted installation of gas turbines at the xAI site was discovered by overhead aerial imagery, prompting a public outcry and demands for the firm to adhere to their allotted quantity of turbines. The use of satellite imagery as recently as October 2025 shows that xAI has adhered to these restrictions and is operating below the allowed quantity of turbines. While overhead aerial imagery was used to initially discover this breach of contract, satellite imagery could have also been purchased and used for this purpose. Now, satellite imagery proves that xAI’s behaviors have altered and that they were responsive to complaints. Moving forward, satellite imagery can be used to continue to count turbines and identify other infrastructure, comparing permits to imagery to determine whether companies are adhering to their legal obligations.
Cooling infrastructure identification, measurement, and counting can be used for some energy calculations.
The identification of the natural gas turbines is particularly important for verification of commercial commitments to energy usage and environmental impact. A key takeaway, as discussed by Epoch AI, is the ability to use satellite imagery to calculate energy usage. The calculations are outside the scope of this report but can be explored through Epoch AI’s published methodology. This approach is analytically rigorous and innovative, and provides a useful starting point for understanding data center energy capabilities. However, as Epoch AI notes, there are limitations to this methodology. Firstly, it relies on the correct classification of cooling infrastructure, and these features can vary in appearance and visibility in satellite imagery. Some Amazon Web Services data center designs, for example, may not show clear cooling signatures. Additional uncertainty arises from various performance characteristics not being directly observable: while cooling towers themselves can be counted, their fan speed cannot be determined. To make calculations, Epoch AI selects specific parameters for cooling system performance based on known industry standards and infrastructure specifications. Thus, these measurements are best interpreted as approximations of capacity rather than exact measurements of operational energy use.
There are significant limitations to using EO satellite imagery for identification of data centers and their accompanying infrastructure.
One of the primary analytical challenges is determining whether an infrastructure feature is actually related to a data center facility as a whole. Not all of the aforementioned elements will necessarily be visible at a data center site. For example, not all facilities will have main power on site. Some will have power plants further away that supply the electricity, but they will not necessarily be in the vicinity of the data center itself. In addition, the construction of the infrastructure itself is variable; for example, fans and chillers may not look consistent across different companies or different countries.
Another challenge is the existence of infrastructure near a data center hall that is not related to the facility at all. There may be power supply and substations adjacent to a data center hall that has nothing to do with the data center itself and is being used for nearby unrelated facilities.
In addition, widely-available EO imagery, in general, shows only wavelengths of light visible to the naked eye. EO is unable to detect complex spectral signatures and can only view the outside of buildings. Therefore, it cannot provide information on internal infrastructure, including GPU clusters, racks, indoor chillers, or other equipment. This can sometimes be mitigated with historical construction imagery of a site: it can be possible to view, on the foundations of a main data hall, how racks are planned to be installed. Therefore, historical imagery might be used to predict future rack and cluster quantities.
Clouds, smoke, or other adverse weather conditions can also block infrastructure from being seen in EO imagery. The temporal frequency—or the time between image captures of the same location over Earth—is high enough that imagery could always be purchased, but this is a significant barrier to imagery being collected over particularly cloudy or rainy regions.
Satellite imagery is best suited for identifying activity at hyperscale data center sites. Smaller data centers can be housed in existing buildings in other corporate, commercial, or industrial sites, and may not show the distinctive features visible at the larger facilities. In addition, because of the variability of hyperscale data centers, it may not be feasible to identify hyperscale data centers from scratch without previously knowing their locations; while they have a large footprint, their buildings might be too generic to be identified with imagery.
In addition, the analysis of satellite imagery itself is a complicated art. The author is a trained imagery analyst and has years of professional experience with image interpretation. Understanding how to accurately analyze satellite imagery may be a barrier to entry for newcomers to the field, but this research is intended to serve as a starting point for those interested in developing these skills.
Finally, purchasing recent, high-resolution imagery is often prohibitively expensive, with each image costing anywhere from hundreds to several thousand dollars. This can be mitigated with freely available imagery online, but lacking the control over image capture dates and resolution is a significant drawback.
Recommendations
Recommendation 1. AI researchers and analysts can use this report as a guide for assessing data center buildout and operation.
Using both the established construction milestones and the identifiable features at completed data centers, the AI research community can reproduce these methodologies to compare construction milestones to commercial announcements, identify data center features, estimate buildout timelines, and evaluate potential energy usage requirements.
Recommendation 2. Policymakers should integrate satellite imagery monitoring into assessments of foreign and domestic hyperscale data center buildout to support energy planning, competition analysis, and strategic forecasting.
This is particularly useful for monitoring the progress of lower-transparency countries such as China, Saudi Arabia, and the UAE.
Recommendation 3. U.S. state governments and civil society actors should use EO imagery as a layer of independent verification to ensure that frontier AI data center developers are adhering to their legal obligations.
Comparing satellite imagery to permits, environmental claims, and commercial announcements can empower local communities to hold data center developers accountable.
Recommendation 4. International organizations should consider integrating satellite imagery as one element of verification for future international agreements.
As global discussions of international AI governance continue, satellite imagery should be included as a tool for independent verification of future agreements. The International Atomic Energy Agency has a satellite imagery analysis team for nuclear safeguards verification purposes, and this displays the efficacy of satellite imagery as one element of infrastructure development and operation evaluations.
Opportunities for Further Research
Future analysis could apply the principles developed in this paper to thermal infrared (TIR) imagery of operational data centers, as this type of sensor tracks and displays heat signatures. For example, the heat emitted by the walls and roofs of data halls, waste heat removed by chillers, and heat generated by electrical operation at substations and switchyards all are signatures that could potentially be visible in thermal infrared imagery. Commercial imagery provider SatVu in particular launched its second thermal infrared satellite, HotSat-2, on March 30, 2026. HotSat-2 has uniquely high-resolution capabilities for a thermal imaging satellite, providing up to 3.5-meter resolution. At this level of detail, useful building and other infrastructural signatures would likely be visible.
Additional research could also involve assessing the utility of other imagery sensors, such as synthetic aperture radar (SAR), multispectral, and hyperspectral imagery for tracking data center buildout and operation. On a larger scale, another avenue of research could explore how to potentially automate the detection of data centers; however, this would require high quantities of imagery for training data and a specifically-trained, accurate model. The difficulties of systematically distinguishing data centers from other industrial facilities is an additional challenge. Thus, while automated data center detection could be highly-desired as model capabilities develop, its practical implementation remains primarily conceptual at present.
Future research could also include comparing infrastructure of hyperscale data centers between different companies or different countries. For example, the infrastructural features at U.S.-based sites might have different overall patterns than those in the UAE or China. Satellite imagery provides a unique look into low-transparency countries, and understanding the infrastructure of international data center sites would provide insights as to the energy usage and overall capacities of existing and future facilities.
Conclusion
Data center facilities are critical to the successful operation of AI models, and their recent expansion reflects high interest in the continued development of these systems. However, despite public attention, key elements of data center construction and operation remain poorly understood due to a lack of publicly-available information, with corporate announcements being scarce and commercial timelines not necessarily reflecting true on-the-ground activity. This research demonstrates that EO satellite imagery can provide an independent, credible layer of insight regarding data center layouts, making it a valuable tool for verifying construction milestones and the presence of visible infrastructure rather than for measuring compute or internal activities.
The UAE Ajman Data Center case study exposes inconsistencies between construction milestone announcements and observable development progress, outlining the limitations of solely relying on corporate information for tracking data center buildout. At xAI’s Colossus facility in Memphis, EO imagery is used to identify specific features unique to data center facilities. In particular, the characteristics of power and cooling infrastructure support calculations of energy capacity and can be cross-referenced with public data such as permits and utility publications to ensure compliance.
While EO imagery is useful for tracking infrastructure-level developments, it is also limited by several constraints. It cannot be used to see the internal activity of data center facilities, and thus is ill-equipped to inform on GPU other compute activity. Current methodologies for estimating energy usage are still broad and exploratory, necessitating analytical inferences and requiring validation with other sources. Thus, satellite imagery should not be evaluated alone but should be treated as one of many analytical resources for assessing data center facilities.
These findings provide assessments to address global competition concerns, support evidence-based planning of industrial policies, and can contribute to longer-term international discussions of transparency and verification of AI-related infrastructure. From individual researchers to international organizations, satellite imagery can be integrated at many levels to support verification efforts as one facet of a larger toolkit of transparency-building measures.
Ultimately, this report illustrates where satellite imagery adds value—and where it does not—to inform future AI policy and governance efforts. As AI infrastructure expands, its physical signatures will continue to play a vital role in policy and regulatory discussions, and, when analyzed critically and integrated with other sources of data, it provides valuable, credible insight into the pace and scale of AI development.
Appendix
Face Recognition Performance, Bias, and the Limits of Technical Fixes
Christopher Gatlin was arrested for a brutal assault he didn’t commit after AI Face Recognition Technology (FRT) said he matched the suspect. He spent 17 months behind bars, and clearing his name took two years. As of March 2026, there were at least nine documented U.S. wrongful arrests tied to face recognition misidentification, mostly involving Black people. From 2012 to 2020 Rite Aid customers, disproportionately in non-white neighborhoods, were flagged by FRT as shoplifters, confronted, and sometimes expelled, including the searching of an 11 year old girl, all on the basis of bad matches.
Errors made by FRT are one cause of these harms, and these systems are known to make more errors on certain populations, including Black people, women, East Asians, and older people. But the way these systems are used by humans is a key component of these errors. Christopher Gatlin was identified based on a grainy photo of a hooded, partially obscured face, which could not be expected to lead to reliable identification. Moreover, police arrested him despite a lack of corroborating evidence. Harms caused by Rite Aid were due in part to a decision to mainly deploy face recognition in disproportionately non-white communities, as well as a lack of proper user training and the use of poor quality photos.
At the same time, face recognition does provide real benefits. In controlled, cooperative settings such as unlocking phones, banking apps, or passport verification, modern systems can be highly accurate. NIST evaluations show dramatic improvement over time, with errors occurring about one time in 1,000, depending on conditions. Millions of Americans use face recognition daily for convenience and security.
In tasks involving uncontrolled settings with uncooperative subjects however, such as identifying people from surveillance images, accuracy is much lower and more difficult to measure. Law enforcement and child-protection organizations have still used face recognition to identify suspects, locate missing children, and support trafficking investigations, but the potential from harms from inaccurate results in high stakes settings is much greater. Furthermore, the effect of biased performance is magnified in these uncontrolled settings, in which the number of errors seems to be much greater for some subpopulations. This report focuses on the causes of this bias, its potential harms and possible steps to reduce these harms. The use of face recognition in mass surveillance obviously raises other serious potential concerns, but these are outside the scope of this report.
Harms from FRT result both from technical errors and flaws in the ways humans use these systems. This suggests two parallel strategies for reducing the negative effects of biased face recognition. One approach is to reduce the bias in face recognition systems directly. Bias can occur due to training FRT using biased datasets that do not accurately reflect the demographics of the overall population. This can be difficult to eliminate due to the massive scale of data used to train FRT, which makes it difficult to control or even understand the demographics of the data. But further efforts can be made to reduce demographic bias in the data. Numerous other external factors that are more difficult to control may also create biased performance. Consequently, in the near term it may be practical to reduce, but not to completely eliminate biased performance.
A complementary approach to reducing harms from biased face recognition is to ensure that FRT are used appropriately by human operators. This solution is much easier to implement in the near term than the previous technical solution. It is not sufficient, however, simply to ensure there is a human in the loop confirming the results of FRT, since often FRT are more accurate than humans, their errors occur on challenging cases, and people may be unable to correct these errors. Behavioral policy interventions range from research aimed at better measuring bias and understanding when FRT results are not trustworthy to clear standards for how human operators use and interpret the results of FRT and restricting the use of FRT when potential harms outweigh the benefits.
In this report we provide an overview of face recognition performance and differential performance between different demographic groups. We summarize results from the National Institute of Standards and Technology assessing performance of numerous commercial face recognition systems. And we provide an overview of potential policies to reduce harms from face recognition bias.
Acknowledgements
Our understanding of this topic has benefitted greatly from conversations with Kevin Bowyer, Leah Frazier, Patrick Grother, Anil Jain, Brendan Klare, Alice O’Toole, Jonathan Phillips, Jay Stanley, and Nathan Wessler. We also received insightful comments and suggestions from Clara Langevin and Caroline Siegal Singh. Any failure in understanding is due to the authors.
Contents
Introduction
Face Recognition Technology Has Caused Significant Harms
Improper development or use of face recognition technology (FRT) can lead to serious harms. One such example occurred in 2020 when Christopher Gatlin was arrested for a brutal assault he didn’t commit after a face recognition system proposed him as a possible match for the suspect. He spent 17 months behind bars, and clearing his name took two years. Porcha Woodruff, eight months pregnant, spent 11 hours in detention for a carjacking after another bad match, even though surveillance footage showed the suspect was not pregnant. As of March 2026, there are at least nine documented U.S. wrongful arrests tied to face recognition misidentification.
In another example of this dynamic, Rite Aid, a major pharmacy chain, deployed face recognition technology widely in stores to spot alleged serial shoplifters. Impacted customers, disproportionately in non-white neighborhoods, were flagged, confronted, and sometimes banned from stores, including searching an 11 year old girl, all on the basis of bad facial recognition matches. Federal regulators later banned the company from deploying facial recognition technology in stores for five years, noting higher false-positive rates in stores serving predominantly Black and Asian communities and improper pre-deployment safeguards (more details here).
These instances of incorrect matching and arrests have mostly involved non-white people. But, while errors may be more prevalent among these populations, as FRT use grows it can increasingly affect all people. For example, police recently released a white Tennessee grandmother who had been wrongly jailed for nearly six months based on FRT results. She was arrested while babysitting four children, accused of committing bank fraud in North Dakota, although she had never been there. Unable to pay her bills, she lost her home.
Figure 1. On the left is a surveillance photo taken at a crime scene. On the right is the image of Robert Williams that was incorrectly matched to this photo by an automatic face recognition system.
The harms described above were instigated by flawed matches produced by FRT—computational models that perform face recognition. However, these models always form part of a larger system in which humans apply FRT to some task. The failures were not just the product of a bad model, but of human failure to follow effective procedures. In many cases, face recognition searches are performed using low resolution images, with faces partially obscured. Figure 1 shows the surveillance photo used to identify Robert Williams, who was wrongly arrested for theft on the basis of this image. He later stated, “My daughters can’t unsee me being handcuffed and put into a police car.” In some cases, police have violated accepted practice with suggestive remarks that prompt witnesses to confirm the results of automatic face recognition technology. In the Rite Aid case, poor employee training, the use of low quality images, and many other deployment decisions contributed to a large number of mistaken identifications.
Face Recognition Technology is Increasingly Widely Used
Face recognition technology has become increasingly accurate and widely adopted. It is estimated that 131 million Americans use face recognition on a daily basis for applications such as unlocking their phones or banking apps, providing convenience and improving security. FRT usage is especially prevalent in applications in which the person being recognized cooperates with the system. In controlled, cooperative settings, face recognition systems have improved rapidly, with error rates roughly halving every two years in some evaluations. Under ideal conditions, top-performing systems may make a mistake only once in several hundred attempts.
Face recognition is also increasingly used by law enforcement agencies to identify uncooperative subjects, identify criminal suspects, and find missing children. Its use in surveillance is also growing. For example, Immigration and Customs Enforcement (ICE) is using FRT to identify people and determine their immigration status. In these applications, FRT often successfully identifies individuals, but their accuracy is not as high, and the potential for harmful errors increases. An incorrect match in this instance can potentially result in wrongful detention or deportation of American citizens. As face recognition use grows, so will its benefits and harms, making it an urgent matter to understand its properties, impact, and effective policy interventions.
Figure 2. Each column shows a pair of images of the same person. Experimental subjects find the images on the left easiest to match, while it is most difficult to determine that the images on the right come from the same individual.
Face Recognition Difficulty Varies Significantly
The difficulty of face recognition problems varies tremendously depending on the setting. Figure 1 has already shown a difficult operational setting, in which a poor quality surveillance image must be matched. A human examining these images has a hard time telling whether they are of the same person. Figure 2 shows that even when images are of good quality, it is not always easy to tell whether they come from the same person, due to changes in things like hairstyle.
What Do We Mean by Bias in Face Recognition?
Bias in face recognition has been the subject of significant public concern and extensive research over the past decade, particularly as these systems have been deployed in high-stakes settings such as law enforcement and surveillance. This report examines the nature, causes, and consequences of this bias, and in this introduction we begin with a brief discussion of what we mean by “bias”.
Face recognition is meant to solve a problem that has an objectively correct solution; do these two images come from the same person? We say the system displays bias against certain demographic groups if it makes more errors on these groups than on the general population. We will use the terms “bias” and “differential performance” interchangeably.
FRT have consistently shown worse performance on women than men and worse performance on Black people than on white people, and many FRT display worse performance on East Asian people than white Americans. One way that bias can occur is through training FRT models using unbalanced data that better represents some groups. When this occurs, bias can be mitigated by augmenting the training set to represent different groups more equally.
However, defining demographic subgroups exactly can be difficult, making it hard to balance data. Studies that compare performance on men and women generally ignore subtleties of gender identity. Groups of Black or white people used in studies certainly contain many individuals of mixed race and, for example, Black people in the United States might have a different distribution of traits than Black people from East Africa. Different studies sample demographic subgroups in different ways, and therefore may not be evaluating exactly the same questions.
Moreover, it is unclear how best to define demographic subgroups. For example, is it more fruitful to measure differential performance between white and Black people, or between light-skinned and dark-skinned people? Black people can differ from white people not just in skin tone but also in structural properties of their face. At this time, it is unclear which aspects of appearance account for differential performance and how this would align with all possible subgroups. Most studies have been limited to a few broad demographic categories and it is not known, for example, whether performance would differ between specific nationality groups within a similar region such as Vietnamese and Korean people.
Outline of the Rest of the Report
This article aims to provide necessary background to assess the trajectory and risks of bias in face recognition technology. We do not address other important concerns about FRT, such as maintenance of privacy and the use of FRT in mass surveillance.
In the next section we will briefly describe how face recognition systems work. We will then discuss the world-wide scope of face recognition. Next we summarize the accuracy of FRT and how this has progressed. We then discuss the nature of bias in FRT, and consider the causes of this bias. Next we consider FRT as part of a socio-technical system, and the impact of human users on FRT harms. Finally, we suggest possible policy interventions to reduce these harms.
This report makes the following points:
1. Improvements in accuracy have not eliminated bias.
Face recognition systems have become significantly more accurate in recent years, but they continue to exhibit differential performance across demographic groups.
2. Bias is difficult to measure and difficult to fully eliminate.
In real-world, uncontrolled settings, bias is harder to quantify and may be larger than benchmark results suggest. While technical interventions can reduce disparities, there is no simple or complete solution.
3. Harms arise from both technical errors and how systems are used.
Errors in face recognition can lead to significant harms, including wrongful arrests and other adverse outcomes. These harms are often amplified by deployment decisions, such as where systems are used and how results are interpreted.
4. Face recognition should be understood as a sociotechnical system.
Bias and harm arise not only from the underlying models, but also from human judgment and organizational practices. Inappropriate use of face recognition results can be more significant than technical error.
5. Policy interventions can reduce harms even without perfect technical solutions.
Effective policies include improving transparency and evaluation, supporting research on real-world performance. Furthermore, just having humans check the results of FRT is not sufficient to avoid errors; this requires establishing clear, detailed protocols governing when and how face recognition may be used.
6. Governance of use is as important as improving the technology.
Auditing data and system outputs, developing tools that signal when results are unreliable, and enforcing strict use protocols can significantly reduce the risk that errors lead to harmful outcomes.
Glossary
How Face Recognition Works
Face recognition is based on machine learning, and highly dependent on the use of large-scale data sets. This data is difficult to carefully control or characterize.
Face Recognition refers to the process of automatically identifying a person from a photo. It is divided into two tasks. In verification (or one-to-one matching), two images of faces are compared to provide a yes/no answer to the question of whether they come from the same person. This is used, for example, in border control, when a live image of someone may be compared to their passport photo. In identification (or one-to-many matching), a single probe face image is compared to a potentially large gallery of images to determine which, if any faces in the gallery match the probe image. The gallery might contain, for example, mug shot images of people who have been arrested, driver’s license photos, images of people who have been barred from access to casinos, or a large collection of images scraped from the internet. A system performing identification might declare that it finds no match, return a single match, or return a potentially large collection of images that might resemble the probe image. In the latter case it is expected that these potential matches will be assessed by the user to identify valid matches. FRT may also return a confidence level about the correctness for each match, although these may not correspond to the true probability that the match is right.
A Brief History of Face Recognition
The first fully automatic face recognition system was developed 50 years ago as the subject of the PhD thesis of Takeo Kanade, who went on to become one of the pioneers in the field of computer vision. It identified landmarks on the face, such as the corner of the mouth, and used their position to compare images. Early methods like this, based on face geometry, had limited effectiveness. Scientists began to develop more useful and accurate face recognition systems through the growing use of machine learning, beginning in the late 1990s. These methods are trained with numerous face images, called a training set, to automatically extract representations of faces that can be used to compare them more robustly.
Progress accelerated rapidly as researchers began to appreciate the power of using an approach known as neural networks, which allowed them to leverage massive datasets of faces to “teach” the computer how to recognize new faces. While neural networks were used by FRT by the late ’90s, their use became dominant in the mid-2010s after further breakthroughs in machine learning with large neural networks, a technique known as deep learning. Since the mid-2010s, improvements in model architectures, training methods, and data scale have driven substantial gains in measured accuracy, especially on standardized benchmarks. At the same time, these advances have enabled rapid adoption of face recognition across a range of applications, from smartphone authentication to large-scale identification systems used by governments and private firms, even as performance in real-world settings remains highly dependent on context.
How Face Recognition Models Are Trained
To perform accurately, an FRT must be able to determine that two images of the same person are similar, even if the images are taken at different times, from different viewpoints, under different lighting conditions. This is done by training the machine learning model to extract a representation that captures facial properties that can distinguish one person from another, but that are not significantly affected by viewing conditions or even some aging. The similarity between two faces can be given a numerical score that represents the degree of difference between the representation of each face.
In its simplest form, training occurs by incrementally adjusting the parameters of a neural network. In most current publicly available systems these parameters consist of tens of millions of numbers that control the network’s behavior. If it is shown two images of the same person, the parameters are adjusted to increase the similarity score. If the images are of two different people, parameters are changed to lower the score. Once the model is trained, if two images produce a similarity score above a chosen number, known as the cutoff, the system declares the two images to be the same person; if it falls below that cutoff, the system says they are different.
Once the model has been trained, it can perform identification using a gallery of faces by comparing a representation of the probe to representations of the gallery images. That is, it can verify or identify images of people who were not in the training set, because it has learned a general representation that should apply to any faces.
The large data sets used in training are typically scraped from the internet. For example, one influential early data set, Labeled Faces in the Wild, made use of face images detected in Yahoo! news stories, with identifying captions. A number of large scale datasets containing millions of images have been developed using photos of celebrities available on the internet. Some companies, such as Meta and Google have made use of internal data that users have uploaded and labeled; these training data sets may contain more than 100 million images. Clearview, a face recognition company, claims to use data sets of more than 70 billion face images scraped from the internet. Given the high cost and diminishing returns of training with so many images it is unlikely that all of these images are used for training, and this large corpus is more likely to be used to form the gallery.
Academic FRT generally train on datasets of images of public figures, such as the MS-Celeb-1M dataset, which contains ten million images of about 100,000 individuals. These massive datasets capture how a person’s appearance can vary with age, lighting, viewpoint, expression, and other conditions, which helps improve accuracy of systems trained on the datasets. Commercial systems do not generally provide details of their training sets, but it is expected that they include similarly large sets of images scraped from the internet, or provided by users, as in the case of Google and Meta. However, because these data sets are assembled at enormous scale—often from uncontrolled sources—they are difficult to audit, regulate, or correct when they embed systematic biases.
Face Recognition in Use Today
Face recognition use is increasing rapidly, becoming more prevalent in numerous high-stakes applications.
The global face recognition market was almost nine billion dollars in 2025, with projected growth to over 30 billion by 2034. Over a third of this market is in the U.S., but there is wide adoption of FRT around the world. One of the primary applications of face recognition is to efficiently and reliably identify people. This can make access to financial systems more secure, potentially preventing identity theft. It can also make hospital admissions quicker and more accurate, and speed up passport verification. In these applications, a human subject opts-in to using the FRT, cooperating to allow consistency in viewpoint, avoiding unusual facial expressions, and enabling controlled lighting. This leads to highly accurate systems. In many cases, such as using FRT to unlock cell phones, users opt-in to the technology for added convenience and device security. When entering the country, U.S. citizens may opt-in to face recognition systems, and their photos are deleted after 12 hours, while non-citizens are required to participate, with photos retained for 75 years.
Face recognition is also widely used in surveillance and law enforcement. Ten percent of U.S. police departments use FRT. The NYPD made 2,878 arrests resulting from FRT in the first five years of its use. The Metropolitan Police in London report 100 arrests using FRT in conjunction with mounted security cameras, including a suspect accused of kidnapping. Police in New Delhi used FRT to identify almost 3,000 missing children, and FRT has been used to identify refugee children who have been separated from their family. The National Center for Missing & Exploited Children (NCMEC) has used a tool called Spotlight, which makes use of FRT, to identify children who are victims of sex trafficking. In 2023, the FBI worked with NCMEC to identify or arrest 68 suspects of trafficking. A large number of retail stores use FRT to track customers to understand traffic patterns, and despite the Rite Aid case, retailers such as Wegmans still use FRT to spot accused shoplifters. Immigration and Customs Enforcement (ICE) is using FRT to identify people and determine their immigration status.
Face recognition has been widely used for surveillance of the Uyghur population by the Chinese government., FRT are used by the Israeli government to track and surveil Palestinians.
These applications of face recognition can solve crimes, enhance security and make access more convenient, but also raise troubling concerns about mass surveillance, repression of civil liberties, and high-stakes errors which materially harm people. In surveillance and criminal investigations, subjects are not cooperative, and probe images used are often of poor quality, as illustrated in Figure 1, which produces much higher error rates. An awareness of mass surveillance can also have a chilling effect on people’s ability and willingness to participate in Constitutionally protected activities such as protest or dissent.
As face recognition has grown more practical, a large number of companies have developed and marketed FRT. This includes large tech companies such as Amazon, Microsoft, Toshiba, NEC and Apple, and smaller companies that focus more narrowly on face recognition, biometrics and security, such as Clearview, Idemia, and Rank One Computing. Clearview is one of the most widely used by federal and local law enforcement in the U.S.
Early in the development of face recognition technology, the best performing systems were produced by academics and used openly available architectures and data. However, with its rapid commercial growth, state of the art FRT are generally developed by companies that provide little transparency about how they work or what data they use. As we will discuss in more detail, the National Institute of Standards and Technology evaluates the performance of some of these systems, but this evaluation is voluntary and not all companies participate.
Face Recognition Performance Across Different Conditions
Face recognition performance has improved rapidly, but recognition can still be quite difficult in many settings.
Two types of errors can occur in face recognition. With false positives, a FRT incorrectly states that two images come from the same individual. With false negatives, the system incorrectly states that two images do not come from the same individual. The cutoff is what determines the balance between false positives and false negatives. Tightening it makes the system more cautious about declaring a match (reducing false positives) but also more likely to miss legitimate matches (increasing false negatives).
Figure 3. The ACLU found that Amazon’s face recognition system matched 28 members of Congress to mugshots of other people.
The significance of this cutoff is illustrated well by the American Civil Liberty Union’s (ACLU’s) evaluation of Amazon’s FR system, “Rekognition” and the subsequent controversy. The ACLU reported that they had tested Rekognition, and that it incorrectly identified 28 members of Congress with people who had committed crimes (Figure 3). A significantly disproportionate number of these false matches were people of color. Amazon responded by arguing that although the ACLU had used the default cutoff, or confidence threshold, of 80% for Rekognition, this was more appropriate for finding celebrities on social media, and that their documentation recommended a much more stringent cutoff of 99% for use in high stakes applications such as law enforcement. Amazon also pointed out that the bias in the results may have been due to bias in the gallery of images used by the ACLU. If the ACLU compared images to a gallery that disproportionately contained people of color it would be more likely to produce false matches for people of color in congress. The ACLU replied by stressing the dangers of a system that was inaccurate with default thresholds and a lack of guidance for the system’s use.
One lesson from the Amazon Rekognition controversy is that the potential harms of an FRT depend not just on its technical accuracy but also on how users apply these systems. It also provides some indication that Rekognition was more prone to false positive errors when applied to people of color, at least at one significant cutoff threshold.
Figure 4. Three images of a researcher at the National Institute of Standards and Technology. The left image simulates a passport or similar photo, the middle image simulates images that might be taken while going through immigration, the right image simulates an image taken by a kiosk.
Figure 5. Two pairs of images, each pair shows the same person under identical imaging conditions except for a change in lighting (images from the Multi-PIE dataset).
Challenges in Real-World Face Recognition
The most rigorous experiments measuring face recognition accuracy are conducted under tightly controlled conditions. As a result, reported performance often overstates how systems perform in real-world settings, where error rates can be much higher.
The difficulty of face recognition tasks can vary widely. Frequently, identification is performed by performing verification between the probe image and all gallery images. Identification becomes more difficult as the gallery size grows and the number of opportunities for false positive matches increases. The difficulty of face recognition tasks also depends very much on the conditions under which images were taken. For example, in border control, the subject can be required to face the camera with their face fully visible, lighting can be controlled, and camera quality can be ensured.
Figure 4 shows that even images taken at a kiosk can be much harder to match, due, for example, to changes in viewpoint. Figure 5 illustrates the effect that a change of lighting can have on the difficulty of matching faces. As previously shown in Figure 1, when images come from surveillance cameras, the subject may not be facing the camera, they may not be close to the camera, so image resolution can be low, and their hair or hand or another object may obscure part of the face. Identification with poor imaging conditions may have many orders of magnitude more errors than verification under tightly controlled conditions.
By all metrics, there seems to be little doubt that face recognition accuracy has been improving rapidly. The National Institute of Standards and Technology (NIST) Face Recognition Vendor Test (FRVT) evaluations illustrate this increase (most recent results here). NIST evaluates verification performance on two high quality images of frontal facing individuals. From 2020 to 2025 the error rate fell by a factor of three. (They set a threshold for matching to achieve a false positive rate of 0.003%, so about one false identification in 33,000 attempted matches. They then measure the false negative rate, the number of correct matches missed. The best performing system as of January 2025 achieved a false negative rate of 0.13%, a little more than one correct match missed in 800.) Similarly, the error rate on an identification task that matched a mug shot probe image to a large gallery of mugshots fell by a factor of 5 during the same period. (The best performing method, when using a threshold to produce a false positive identification rate of 0.3%, had a false negative error rate of 0.05%. This means that the system would falsely identify a probe image in the gallery (of 1,600,000 mugshots) one time in about 300, while missing a correct match about one time in 2,000.) Some results are shown in Figure 6, as of March 2025. Over a period of decades, NIST has found that errors have generally fallen by about a factor of two every two years. Under controlled conditions, FRT are now much more accurate. For example, on the best performer as of March 30, 2026, when performing verification on two mugshots, using a cutoff set to make a false positive match one time in a million, a false negative failure to find a match will occur one time in 500. This sharp increase in accuracy in a short period has happened alongside widespread adoption in applications like border control or unlocking a phone.
These experiments represent relatively ideal conditions. FRT in the real world may face much higher failure rates. This can occur due to more challenging imaging conditions, such as using a surveillance image as a probe, instead of a mugshot, or other factors such as changes in the subject’s appearance. For example, when the best performing system at mugshot identification is applied in a scenario in which the gallery contains visa images and the probe is taken from a kiosk, the error rate increases by a factor of about 18 with a false negative error about one time in 30 instead of one time in 500. This is a fairly typical increase, and still represents relatively idealized conditions compared to the most challenging ones.
Defining and Measuring Bias in Face Recognition
Face recognition performs with different levels of accuracy on different demographic groups. As face recognition becomes more accurate, this may limit the effects of this disparity in some applications, but it can still be quite significant in high-stakes applications.
Going back more than 30 years, researchers have observed different rates of accuracy in face recognition systems depending on demographic properties of the subject, including race, gender and age. For example, in 2011 a study showed that Western face recognition algorithms performed better on Caucasian faces than East Asian faces, while East Asian face recognition systems performed better on East Asian faces than Caucasian ones. In 2018, the influential Gender Shades paper examined differential performance not in face recognition, but in a related facial analysis problem of determining gender from a face, showing much poorer performance on images of dark skinned females than light skinned males.
Absolute vs. Relative Error
In considering differential performance, it is important to distinguish between absolute and relative differences in performance. We define the absolute difference in two error rates as the difference between the larger and smaller error. For example, if an FRT produces 2% error on male faces and 4% error on female faces, we would say that the absolute difference is 4% – 2% = 2%. We describe the relative error as the ratio between the larger and smaller value, which in this case would be 4%/2% = 2. As overall performance improves, the absolute error tends to decrease, while the relative error rate might or might not decrease. For example, if a new generation of FRT reduces error on male faces to 1% and reduces error on female faces to 2%, absolute error decreases from 2% to 1%, while relative error remains constant.
Whether absolute or relative error is more important depends on the operational considerations and use of the system. When performance is very high, absolute error will tend to shrink. If this translates into operational settings, then relative error may become unimportant. For example, if an FRT makes a mistake once in a billion queries on one population, and twice in a billion on another, errors for either population may be so rare that they are insignificant. In practice, the impact of absolute error also depends on how widely deployed a system is. As systems become more accurate, they may become more widely deployed, which can paradoxically result in more accurate systems producing more errors.
Even though current FRT achieve quite low error rates under ideal conditions, these error rates tend to grow much higher under more challenging conditions, and errors can be quite common. Although it is difficult to study error rates accurately under the most challenging conditions, high relative error under ideal conditions may predict relative error that is just as high or higher under challenging conditions that also have high absolute error. That is, while absolute error in operational contexts is of greatest importance, relative error in highly controlled conditions may predict high absolute error in less controlled conditions. Consequently, it is premature to think that FRT are so accurate that relative error is no longer important. A more nuanced view would hold that continuingly high relative error rates may be less important for some applications, such as unlocking phones, and still be quite important in other applications, such as criminal investigations.
NIST Experiments on Demographic Variation
Since 2019 NIST has performed extensive evaluations of demographic variations in performance on hundreds of face recognition systems. They have access to large collections of non-public images that they use to evaluate FRT submitted by companies. The large size and private nature of the dataset makes it especially unlikely that models are overfit to the data by, for example, selecting parameters that boost their performance on this particular data. NIST computes false negative rates using over a million pairs of images, comparing one high quality image of an individual to a medium quality image of the same person. False positive rates are computed using over a billion pairs of high quality images from different individuals. Image quality reflects applications such as passport checks at airports, but does not include more challenging problems such as police investigations using surveillance footage. All images come with demographic information, including the age, gender and country of origin of the subject. Country of origin is used as a proxy for race, focusing on countries that are less racially diverse, but this is not a perfect proxy.
NIST finds a relatively small demographic variation in false negative rates, in which a correct match is missed, and a much larger variation in false positive rates, in which an incorrect match is accepted. For example, the top performing FRT as of March 2025 produced 358 times as many false positives for West African females over 65 as for Eastern European males aged 35-50, with the false match rate increasing from about one in 15,000 to about one in 50. Among the top ten performing systems, the false positive rate for all West Africans was about 23 times higher, on average, than the rate for Eastern Europeans. The false positive rate for these performers on average is about 4.6 times higher for females than males, and about 2.9 times higher for people over 65 compared to people aged 20-35. The evaluations also show poorer performance on people from South or East Asia, relative to Eastern Europeans. Many additional studies have also found that FRT generally perform better on white people than people from other racial groups, and on males compared to females.
These studies do have important limitations. More narrowly defined groups (e.g. West African women over 65) will have less data, leading to noisy estimates, and when we take the ratio of two noisy estimates we amplify the noise. Also, images taken in different countries may differ in ways beyond the race of the subject, such as in the types of cameras or lighting used. Also, incorrect labels may have a significant effect on accuracy. If a visa photo is associated with the wrong name, this can lead to a false match, and these incorrect labels may be more prevalent in some countries than others. Finally, measures of bias may vary depending on the specific ways in which performance is measured. The chief scientist of a leading face recognition company has stated that in practice they find differential performance between racial groups of a factor of approximately 1.5, rather than the higher numbers found in NIST studies. (Brendan Klare, personal communication.)
Challenges in Measuring Bias in Face Recognition
There is decades of evidence of differential performance of face recognition between demographic groups, particularly affecting non-white people and females. However, these studies generally make use of relatively high quality images, and may not accurately reflect the degree of differential performance in challenging operational cases, such as the use of surveillance footage in criminal investigations or in identifying people on a watch list. This is due to the fact that it is quite difficult to accurately characterize and sample images from challenging environments. And while large scale photo collections with known identities and some demographic information exist, such as passport photos, we do not have large scale collections of photos taken in challenging conditions that have this information. While this problem is elusive, there is some evidence that differential performance increases with the difficulty of the recognition task.
Another limitation occurs because races are not well-defined biological categories but social constructs. It is not clear how to systematically divide a population into different races, especially in the case of multi-racial individuals. This is particularly challenging when images are scraped from the internet, and need to be labeled by race. Some studies have focused on skin darkness rather than race, but this is also difficult to determine accurately from photos due to the effect of unknown lighting conditions on apparent skin color. In spite of these limitations, there is a clear consensus among researchers that differences in FRT performance exist between racial groups.
An important question is how differential performance in face recognition is evolving over time. Is this a problem that was initially ignored, but is now being effectively addressed, or one that is recalcitrant? While there is no question that absolute differences in accuracy are shrinking over time, as FRT become more accurate, the behavior of relative differences is less clear. This is difficult to judge, since new test sets come out frequently, and experimental performance is generally measured over an ever changing landscape of conditions. Perhaps the most stable evaluation framework is NIST’s, which has consistently evaluated new FRT under the same conditions including systems developed from 2018 to 2026. Some of the top performing FRT have evolved, with multiple versions being released over this time period. When we examine these, we see that some have significantly reduced the amount of bias over time, while others have not, and have even seen increased bias. This suggests that it may be possible to reduce systematic bias through model design. More details can be found in the appendix.
Sources of Bias in Face Recognition Systems
Bias in face recognition systems arises from a combination of imbalanced training data, differences in image quality and gallery composition, and other technical and operational factors that are difficult to fully control or eliminate.
False negatives often arise when image quality is poor or facial features are obscured, while false positives are more likely when different individuals appear similar to the system, which can be exacerbated by limitations in training data or representation. For example, if we compare two images of the same person, and one of these images is blurry or has bad lighting or low resolution, the images may appear dissimilar due to these effects. FRT are trained to be somewhat robust to changes in viewing condition, but they are still likely to make errors when these changes are large. On the other hand, if a system is trained using few images of one demographic group, the system may not learn representations that distinguish between a wide range of appearances within that group. For example, if one trained an FRT using images of only one Black person, the system would likely learn to associate dark skin with that individual, and would not learn features that effectively distinguish between different Black people. This is an extreme example, but it is generally found that deep neural networks become more effective as the amount of relevant data increases.
We focus on false positive errors, as these show the greatest differences across demographic groups and are most closely associated with documented harms, such as wrongful arrests. In this section, we will discuss two key points. First, while it may be straightforward to improve demographic balance in datasets, completely eliminating demographic bias is complex and difficult. Second, while demographic bias in the data may be responsible for some bias in false positives, it is not necessarily the only source of these differences. Various research results present conflicting evidence of the importance of dataset bias in practice.
The Contribution of Dataset Bias
Face datasets collected in the last 15-20 years have generally consisted of images scraped from the internet. This enables the creation of large scale datasets that capture a wide range of variations in viewing conditions. These datasets often used well-known people with many online photos, without specific regard to accurately representing the distribution of people of different races or genders in the population as a whole. For example, an early and very influential dataset, Labeled Faces in the Wild (LFW), consisted of 77.5% images of men and 22.5% images of women. LFW was based on people who had appeared in Yahoo! news stories that were identified in captions, making it easier to build a large dataset of known people. However, these people were obviously not representative of the overall population.
Some more recent datasets pay closer attention to capturing the true distribution of people in the world. However, creating unbiased datasets can sometimes be a subtle and difficult problem. For example, the BUPT-Balancedface (BUPT) dataset was constructed to have equal numbers of images of Caucasian, Indian, Asian and African faces. However, subsequent analysis revealed that the Asian and Indian faces consistently appeared as a larger size in the dataset. So although the number of images was balanced, the viewing conditions of the images could still vary significantly. This discrepancy might, for example, lead to biased performance at test time.
The reason for systematic biases in datasets is often not well understood, but it is plausible that when scraping images from the internet, photos from different countries might follow different conventions, use different cameras, or differ in myriad other ways. Therefore, to judge whether a dataset is biased is not as simple as counting the number of images from each population.
A deeper difficulty is even defining what it means to have an unbiased dataset. BUPT represented four demographics equally. But it is unclear what should count as a racial category. For example, should Asian faces be counted as one category? Should Chinese and Japanese people be considered two separate racial categories? What about multiracial individuals? The concept of race is not biological, but a social construct that is not well defined. It is also problematic to correctly label the racial origins of large scale datasets, which may contain images of millions of people. It seems clear that paying attention to demographic diversity will produce less biased datasets than building datasets based on arbitrary selection of celebrities. However, it is also clear that creating completely unbiased datasets is an ill-defined problem. Even with a given definition of “unbiased” it remains very challenging and beyond current technology.
There is certainly strong evidence that dataset bias can produce differential performance, and bias can be reduced through improving the training data balance. It has been found that while Western face recognition algorithms perform better on Caucasian faces than on East Asian faces, algorithms developed in East Asia perform better on East Asian faces, a result that is likely due to dataset bias. After the Gender Shades paper demonstrated that Microsoft’s gender identification algorithm performed much more poorly on Black women than white men, Microsoft quickly improved performance dramatically on Black women by balancing its datasets.
Differential performance can also occur because of biases in the gallery data or probe data. When the gallery is formed from images scraped from the internet, the properties and number of these images may vary drastically from individual to individual, or even from group to group. It has been shown, for example, that if one group is more highly represented in the gallery, this will lead to more false positives among that group because there is greater potential for the gallery to contain faces similar to the probe. As another example, if one group, such as women, frequently have longer hair that covers more of their face in the probe image, this can also lead to higher error rates. Also, if a gallery image is of low quality, not showing a clear image of the face, it may be matched to a similar low quality probe image of a different person. Rite Aid’s use of low-quality images in its gallery is believed to have contributed to the large number of false matches it produced, which in turn led to customers—disproportionately in non-white neighborhoods—being wrongly flagged, confronted, and sometimes expelled from stores. When companies such as Clearview make use of billions of images scraped from the internet it is extremely challenging to balance these datasets or ensure uniformity in their quality.
Assessing dataset bias in commercial systems is complicated further by the fact that companies generally do not make their datasets publicly available or disclose many details about them. Moreover, NIST experiments on dataset bias do not make use of the galleries used by commercial systems. Therefore any bias due to galleries would not be detected.
Sources of Bias Beyond the Data
Other factors besides data may also significantly influence differential performance. Some experiments have shown that even balanced datasets do not produce equal performance on men and women, or between races, and that sometimes more biased datasets produce less biased and better results. Furthermore, demographic groups may have properties that make them easier or harder to recognize. For example, there may be greater variation in hairstyle in one gender than another, and males in different countries may have different trends in facial hair. If someone has an unusual beard, for example, this may make him easier to recognize, or harder to recognize if he shaves his beard. It is difficult to determine the effects on differential performance of social conventions affecting appearance. It has also been noted that darker skin may require different types of lighting to bring out the facial structure. This could result in more recognition errors for people with darker skin when lighting is not controlled.
In summary, it is clear that extreme dataset bias produces biased results. It is quite challenging to produce perfectly unbiased datasets, and less clear to what extent the differential performance observed in modern face recognition systems may be due to dataset bias, especially since these systems are built with proprietary data that is not open to public examination.
Reductions in Bias Over Time
From a policy perspective, perhaps the most important question is whether companies have the ability to produce less biased FRT. To address this question we examined NIST measurements of the performance of models produced by leading companies. NIST has assessed the degree of bias in multiple models produced over time by some companies, allowing us to see how their performance has evolved. Based on NIST reports, we find that some companies have significantly reduced the absolute and relative bias in their systems in two or three years after initial evaluation, while other companies have not reduced relative bias, and in some cases it has increased, even while absolute bias decreases due to improved overall accuracy. Details of this analysis may be found in the appendix.
These results suggest that companies are capable of reducing bias, although this is certainly not definitive. In a conversation with one of the authors, the chief scientist at a leading face recognition company confirmed that NIST evaluations have helped them identify certain variants of differential performance between racial groups, enabling them to take effective steps to proactively identify and reduce bias whenever the company becomes aware of it. (Brendan Klare, personal communication.)
The Human Factor: Face Recognition Systems as part of a Socio-Technical System
Many errors in face recognition are due not just to mistakes by the technology, but to the way in which people make use of it.
The preceding sections focused on the technical properties of face recognition systems. However, these systems do not operate in isolation. They are embedded in what researchers call a sociotechnical system, in which the technology interacts with human judgment and organizational practices. The real-world effects of face recognition therefore depend not only on technical FRT performance, but also on how human users interpret and act on its results. In practice, this interaction can create distinctive failure modes. For example, users may rely too heavily on algorithmic matches without considering other evidence or fail to appreciate how image quality and threshold choices affect reliability.
Limitations of Human Oversight
Some authors argue that these human factors can be structured to correct for technical weaknesses in face recognition systems. One commentator contends that: “it is stunningly easy to build protocols around face recognition that largely wash out the risk of discriminatory impacts…. A simple policy requiring additional confirmation before relying on algorithmic face matches would probably do the trick… one has to wonder why so few researchers who identify bias in artificial intelligence ever go on to ask whether the bias they’ve found could be controlled with such measures.”
However, empirical evidence suggests that this confidence in human oversight may be misplaced. First, FRT tends to make errors on difficult cases, in which humans also make errors. Studies show that humans are unable to identify many of the errors made by automatic systems. Furthermore, human performance on face recognition suffers from similar differential performance as machine learning systems. Dubbed the other-’race’ effect, it has long been known that humans are more accurate in recognizing faces from their own race than from others (it has been posited that this also stems from dataset bias, in that people encounter more individuals of their own race than of others). Some work indicates that current automated systems recognize faces more accurately than the typical person, and that in some cases, combining a less effective human judgement with an automatic system may actually lead to lower accuracy than simply using the results of the automatic system. Human judgements can in some cases be used to improve algorithmic accuracy but it may be difficult to determine when that is the case. In general, we cannot assume that human judgements will be accurate or that human oversight can be counted on to correct errors made by automatic systems.
Figure 7. Christopher Gaitlin, right, was identified using the security photo on the left.
User Errors
Consistent with these findings, many of the known cases of false arrests due to FRT errors involved questionable practices by investigators. Christopher Gatlin was arrested for the brutal assault of a security guard, after an FRT flagged him as a possible suspect, based on a low quality image (Figure 7). Police steered the security guard to identify Gatlin, in what they later admitted was improper behavior.
Robert Williams was arrested for burglary one year after the crime, based on applying FRT to a surveillance video. Lacking witnesses, police showed the surveillance video to an employee of the store’s insurance company, who identified Williams from a photo array, although the video was of poor quality and his face was obscured by a shadow (Figure 1). The police failed to take basic steps such as investigating Williams’ alibi. The police chief at the time, James Craig, said that “this was clearly sloppy, sloppy investigative work.” In other cases, police have shown a single suspect’s photo to a witness, violating best practices by being unduly suggestive. This led to an arrest despite the suspect’s convincing alibi.
In cases where FRT lead to false arrests, it seems that police may in fact give undue weight to the results of FRT, rather than catching their errors, an example of “automation bias”. In another case in which recommended procedures were not followed, police were unable to obtain face recognition results due to the low quality of the surveillance image. A detective felt that the surveillance image resembled the actor Woody Harrelson, and used a picture of him to search for matches, rather than the suspect’s photo.
Failures in the use of FRT occur not only in police investigations. In the Rite Aid case mentioned in the introduction, the FTC’s complaint highlighted not just algorithmic errors but significant governance failures in how the system was operated by store employees. The commission found that Rite Aid did not take reasonable steps to train or oversee store employees who were responsible for acting on match alerts, including failing to teach staff how to interpret alerts or warn them that false positives could occur. The company also failed to test or monitor the technology’s accuracy once deployed, enforce image-quality standards, or implement any procedure for tracking false positive alerts and employee responses. As a result, employees in hundreds of stores routinely followed, confronted, searched, or even called police on customers based solely on system alerts—actions taken without meaningful training on the system’s limitations or appropriate safeguards. These shortcomings in training, oversight, and procedural controls were central to the FTC’s determination that Rite Aid had failed to prevent foreseeable consumer harm from the technology’s use.
In summary, it may be difficult for humans to correct mistakes made by algorithms, and in some cases they may place undue confidence on FRT results that are questionable and based on low quality images. In many applications, such as drug stores that are looking for known shop lifters, the people making use of FRT may not be expert investigators or well trained in the appropriate use of these systems.
Policy Interventions to Address Bias in Face Recognition Systems
Many errors can be addressed by better understanding and regulation of the way in which the technology is used.
A wide variety of policy interventions are available to deal with potential harms caused by bias in FRT. These include research, transparency in documenting bias, voluntary or mandatory guidelines governing the use of face recognition, and outright bans on the use of face recognition in certain contexts. As noted above, FRT make positive contributions in law enforcement and other applications, and these positives must be weighed against potential harms in crafting policy. Numerous institutions have suggested policy changes to address bias in FRT, including a comprehensive set of proposals in a recent report from the National Academies.
Research
Federal agencies already support substantial research on face recognition. NIST conducts ongoing evaluations of performance and demographic disparities, and agencies such as the Office of the Director of National Intelligence (ODNI) and the Intelligence Advanced Research Projects Activity (IARPA) have funded foundational research in face recognition systems. However, important gaps remain, particularly in understanding how these systems perform under operational conditions and how human users interact with their outputs. Additional federal funding could expand independent research in these areas, either by strengthening NIST’s evaluation programs or by supporting academic and nonprofit research focused specifically on bias mitigation and real-world deployment risks.
Two research priorities are especially important. First, evaluation frameworks should better reflect real-world conditions. Current large-scale benchmarks often rely on relatively high-quality images, whereas many high-stakes uses—such as criminal investigations—depend on low-resolution or poorly lit surveillance images. While efforts such as the IARPA Janus Surveillance Video Benchmark (IJB-S) dataset have begun to address this issue, broader and more systematic testing under operational conditions would provide policymakers with a clearer understanding of real-world risk.
Second, research is needed to develop tools that help human operators interpret and appropriately limit their reliance on face recognition results. For example, systems could assess probe image quality, estimate the likelihood that a reliable match can be produced, and warn users when results are unlikely to be dependable. Such tools could reduce the risk that investigators or retail employees draw strong conclusions from low-quality, unreliable inputs.
Measure and Reduce Bias
A better understanding of the bias in FRT can inform the procurement decisions of potential customers and encourage companies to take steps to reduce bias. Transparency in bias can be promoted in a number of ways. NIST is already conducting regular and impactful evaluations of bias in FRT, which can be thought of as an application of the Common Task Method (such evaluations have long been common in the computer vision community). This can be continued and potentially expanded. Regulations or government procurement guidelines can be used to incentivize or require companies to participate in evaluations and make these results public. Since criminal investigations are conducted by the government, procurement guidelines are a strong potential lever in promoting transparency. In addition to transparency in performance, these approaches could also be used to promote transparency in the data used to train FR systems. Making training data public may raise significant privacy concerns, but the government could incentivize the release of information describing the data and the steps taken to enhance the demographic balance of these data sets.
Regulate Sociotechnical use of Face Recognition
If we view FR as part of a sociotechnical system, it makes sense also to govern the way in which face recognition is applied, not just the technical performance of the underlying algorithm. In practice, “responsible use” protocols need to specify who can run searches, what minimum image-quality standards apply, what form results can take, and what documentation and oversight are required. They should also define the permissible purposes for which searches may be conducted, restrict access to trained and certified personnel, require supervisory approval for high-stakes uses, and mandate that face recognition results be treated only as investigative leads rather than as dispositive evidence. Protocols can require minimum similarity thresholds below which no candidate match is returned, prohibit the use of face recognition on images that fall below objective quality metrics, and require contemporaneous documentation explaining why a search was initiated and how results were interpreted.
Additional safeguards could include audit trails of all searches and outcomes, periodic independent audits of performance and demographic disparities, disclosure requirements when face recognition contributed to an arrest or charging decision, and exclusionary consequences if required procedures are not followed. Agencies could also be required to collect and publish aggregate statistics on the number of searches conducted, the rate at which matches lead to arrests, and the frequency of erroneous identifications.
As an example of governance procedures, the FBI has established guidelines on the use of face recognition. These include limiting situations in which it can be used and the type of probe images used. They require that all face queries be evaluated by trained examiners and mandate that face recognition be used for investigative leads that must be corroborated.
As another example, the New York City police department (N.Y.P.D.) has spelled out a detailed protocol for the use of FRT. This requires investigators to submit face images to a special facial identification section of the department (the Real Time Crime Center, Facial Identification Section) that will, for example, ensure that image quality is sufficient and that use of FRT is warranted. The section can reject unsuitable probe images and reviews matches. Critically, a “possible match candidate” is meant to be “treated as an investigative lead only” and does not establish probable cause to make an arrest. The unit also retains records of searches and results. It has been reported that in other localities, investigating officers have accessed FRT directly, without supervision. Specific requirements could be mandated, with legal consequences if they are not followed, such as disallowing evidence produced in subsequent investigation.
However, in spite of N.Y.P.D. guidelines, FRT did lead to the false arrest of Trevis Williams. After FRT identified him as a suspect in a crime, the victim identified him from a photo lineup, although he was eight inches taller and 70 pounds heavier than her initial description of the suspect, in addition to other exculpatory evidence. This illustrates the difficulty of ensuring that guidelines effectively prevent errors and false arrests.
Regulation may be applied not only to government agencies, such as police departments, but also to private companies that are increasingly deploying face recognition systems in commercial settings. RiteAid’s use of face recognition illustrates how governance failures can arise outside of law enforcement. According to the FTC complaint, “Rite Aid failed to consider or address foreseeable harms to consumers flowing from its use of facial recognition technology, failed to test or assess the technology’s accuracy before or after deployment, failed to enforce image quality standards that were necessary for the technology to function accurately, and failed to take reasonable steps to train and oversee the employees charged with operating the technology in Rite Aid stores.” These deficiencies were not primarily algorithmic; they reflected a lack of risk assessment, testing, training, oversight, and ongoing monitoring.
The FTC’s enforcement action demonstrates that existing consumer protection laws can be applied to address some forms of misuse. However, as commercial deployment expands, more explicit regulatory standards may be necessary to prevent similar failures. Such standards could require companies to conduct pre-deployment accuracy and bias testing, implement image-quality controls, establish employee training and supervision protocols, monitor and document false positive rates, and assess foreseeable risks before using face recognition in customer-facing environments. Clear statutory or regulatory requirements would provide ex ante guardrails rather than relying solely on ex post enforcement after harms have occurred. Regulations could also require clear disclosure when face recognition is used—both to affected individuals and in aggregate public reporting—so that its role in decision-making can be scrutinized, evaluated, and corrected where harms emerge.
Policymakers should be willing to ask if using facial recognition is appropriate at all in certain circumstances. In higher-risk contexts, policymakers could impose outrights bans, limit use to specified categories of serious crimes, require a warrant, or mandate corroborating evidence before an individual identified through face recognition is included in a lineup or arrested.
As an example of use restrictions, the state of Maryland has limited the use of automatic face recognition to specific, serious crimes, and requires that defense attorneys be notified when it was used in a case. Montana and Utah require police to obtain warrants in the use of face recognition. In Detroit, police must obtain corroborating evidence before placing a suspect identified through face recognition in a line up. Several cities have banned the police use of face recognition, including San Francisco and Boston, while Portland has banned the use of face recognition by private entities in all public places.
At the federal level, members of Congress have introduced legislation that would impose a nationwide moratorium on government uses of face recognition technology absent explicit congressional authorization. Together, these restrictions illustrate a broader policy approach: limiting deployment in high-risk settings until adequate safeguards, transparency, and accountability mechanisms are in place.
Conclusions
Face recognition systems have improved dramatically in accuracy over the past decade, and in tightly controlled environments they now perform at very high levels. At the same time, substantial differences in performance across demographic groups persist, particularly in the false positive errors most closely associated with wrongful arrests and other harms. As overall error rates decline, these disparities may matter less in low-risk settings, but increasing deployment in high-stakes and uncontrolled contexts may lead to continued harms.
Technical improvements can reduce some sources of bias. Developers can improve dataset balance, adjust thresholds, and refine model design. However, eliminating differential performance entirely is beyond the current state of the art, particularly in operational environments involving low-quality images and large search databases. Policymakers should not assume that continued technical progress alone will resolve these disparities.
Perhaps most importantly, policymakers should view the regulation of face recognition through a sociotechnical lens, considering the interaction between the technical system and the humans who use it.
We cannot wait for perfect sociotechnical systems, but must govern the deployment of imperfect ones. Policymakers must decide where face recognition is not legitimate. If face recognition is used in high-stakes applications, it should be subject to clear limitations, transparency requirements, and enforceable protocols designed to prevent errors from cascading into wrongful arrests or other serious harms.
Appendix: Variations in Bias Over Time
We examined the performance of face recognition systems evaluated by NIST on different demographic groups. All results are based on data on a verification task, updated on March 5, 2025. More recent data on somewhat different tasks shows similar levels of bias. False positive matches are measured when comparing two high quality, visa-like images of two different people of the same sex, age group and region of birth. Demographic disparities are computed by taking the ratio of the false positive rate for two different demographic groups. For example, the ratio of the false positive rate on faces of people born in Western Africa to the false positive rate for people born in Eastern Europe for the highest performing FRT was 17.42, meaning that a false positive match was 17.42 times as likely for someone from Western Africa.
NIST has evaluated differential performance of commercial systems for over five years. Many companies have submitted multiple versions of their FRT over time, as the systems have improved. This allows us to determine how the bias in these systems has changed. We considered the 20 systems with best overall performance, which originated from 12 different companies. Eight of these companies had submitted at least four different versions of their FRT for evaluation, and so we focused on these eight systems.
Figure 8 shows the change in the ratio of differential performance for three pairs of demographic groups. For illustrative purposes, we show results from two different companies. The curves from Sensetime illustrate differential performance that has increased over time, while the curves from Rank One Computing (ROC) show differential performance that has decreased. Solid curves show the ratio of false positives for subjects of West African birth compared to Eastern Europeans. The dashed curves show performance on females compared to males. The dashed-dotted curves show an older age group (65+) compared to a younger cohort (20-35).
Table 1 shows the correlation between the passage of time and the ratio of differential performance for all eight companies. A negative correlation indicates that bias has dropped over time, while a positive correlation shows an overall increase in bias. If the correlation is close to 1 or -1, this means that the change in performance over time is highly consistent, while a correlation close to 0 means that there is no clear trend in the increase or reduction in bias. We can see that Toshiba, Idemia, and ROC have reduced biased performance over all three ratios, while Sensetime has increased bias, with other companies showing mixed performance.
Public Participation IS the Ingenuity We Need
Building Blocks To Make Public Participation Solutions Work
Participation is not a distraction from governing — it is how government governs well. When treated as compliance, it comes too late and excludes those most affected, weakening legitimacy. Designed as a strategic asset, it builds trust, eases implementation, and supports more durable decisions.
Implications for democratic governance
- Participation is how decisions are made. Engagement should focus on clearly defined choices, constraints, and tradeoffs so decisions can move forward.
- Who participates, and when, shapes what government hears. Participation must reflect the full scope of impacts, not just the most visible or organized voices.
- Legitimacy depends on follow-through. Agencies should explain how input was considered so people can make sense of the outcome, even when they disagree.
Capacity needs
- Learn and adapt throughout the process. Track who is participating, what input is generated, and how it is used — then adjust the approach as needed.
- Engage early enough to matter — and continuously where possible. Bring people in while options are still open, and stay engaged so feedback informs both decisions and implementation.
- Structure input for decisions. Ask targeted questions and use formats that help compare impacts, priorities, and implementation considerations.
- Train for real-world engagement. Equip staff to facilitate conversations, synthesize input into key themes, and navigate disagreement.
- Make participation feasible. Offer enough lead time, plain-language materials, and multiple ways to engage (e.g., virtual, in-person, asynchronous) that account for different access needs.
Government, at its best, is democracy’s promise made real: the mechanism through which a society turns values into action and public voice into policy. But that mechanism has corroded. Rebuilding it starts with something simple — treating the public not as a problem to manage, but as a source of ingenuity government cannot function without.
Public participation, as the federal government executes it today, rarely builds trust: the public hearing held after decisions are already made, the comment period that produces thousands of responses with no visible impact, the listening session where officials take notes but never engage. The other version, where a highly organized few monopolizes public ears and distorts public response for their niche interests, is equally demoralizing. Critics are right to call out this failure. Across the political spectrum, there is a shared diagnosis: the current system of public engagement too often functions as a series of veto points, rewarding obstruction over problem-solving and delay over delivery.
That erosion matters enormously right now. Americans have grown deeply skeptical that government can solve hard problems, and climate change, with its complexity and its demands on every sector of the economy, may be the hardest challenge it has ever faced. As the authors ask in the opening argument of the Center for Regulatory Ingenuity, if government can’t effectively address challenges it deems an “existential threat,” what good is it, and can democracy overcome this downward spiral of mutually reinforcing cynicism?
We believe it can. Today, we are living through a mid-transition moment in climate policy, in which the technologies we need exist, the economics are increasingly favorable, and many obstacles are governmental: slow processes, fragile coalitions, and policies that get built and then litigated into irrelevance. In that context, the instinct to streamline is understandable. But a government that privileges artificially weighted listening, or avoids listening because it didn’t plan well, doesn’t move faster. It moves blindly or with bias. It builds the wrong things, in the wrong places, for the wrong reasons, and then wonders why nothing sticks. The argument here is not for more or less participation but for better participation, treated as a strategic asset rather than a box to check, and designed with the same rigor as any other policy instrument. Done well, public engagement demonstrates that government can listen, adapt, and earn trust. At a moment when democratic institutions are fragile, that demonstration is not incidental to the work of governing, it’s central to it.
Climate change is not just a technical problem, it’s a governance problem, and increasingly, a democratic one. It reaches into every community, every economy, and every aspect of daily life, from how people power their homes to how they move through their cities to whether their communities remain livable at all. There is no technocratic solution that can bypass the public. The choices climate demands (where to build, what to prioritize, who bears costs, and who benefits) are fundamentally democratic choices. If democratic government cannot make those choices in ways that are effective and legitimate, it will not just fail on climate. It will fail at its most essential purpose: helping people shape the conditions of their shared future.
Definitions and the Role in Governance
Before making the case for better participation, it helps to be precise about what we mean in a space where a range of practices often get conflated.
In January 2025, the White House Office of Management and Budget issued a memorandum that the authors of this paper helped develop, laying out a federal framework for broadening public participation and community engagement. The memo itself was built through the practices it recommends: a public request for information drew input from hundreds of participants and nearly 300 written comments, documented in a public summary that informed the final guidance. It offers definitions worth building on. Public participation is any process that engages the public in government decision-making — helping shape policies, regulations, or research, or soliciting new ideas and innovations. It is inherently transactional: it seeks to inform and obtain input from those interested in or affected by agency action. Community engagement, by contrast, is primarily relational — the consistent building of relationships with communities over time, informed by the history those communities have with an agency, and transparent about the real opportunities and real limitations of that relationship.
Public participation without community engagement can produce processes that feel hollow — technically open, but not genuinely accessible to the people most affected. Community engagement without public participation can build trust and goodwill that never translates into actual influence over decisions. Together, they reinforce each other: relationships make meaningful input possible, and meaningful input, when reflected in decisions, strengthens those relationships over time. But they are not interchangeable, and not always needed in equal measure. One of the memo’s most important contributions is helping government actors think more carefully about which tool fits which moment; a public comment period serves a different purpose than a years-long relationship with a frontline community, and conflating them produces both bad process and bad outcomes. Resources like EPA’s Public Involvement Spectrum make this concrete, mapping participation options from basic outreach through information exchange, recommendations, agreements, and stakeholder action, each carrying a different promise to the public and requiring a different level of agency commitment and design.
What the OMB memo does is not add new mandates (federal statutes from the Administrative Procedure Act to the National Environmental Policy Act already require it across a wide range of agency functions). It strengthens and clarifies what good practice looks like within those existing requirements while making that guidance accessible to anyone in government who wants to apply it beyond those requirements.
That the memo exists at all reflects a recognition that the statutory floor was never enough and democratic practice does not end on election day. The case for meaningful participation rests on the idea that government derives its legitimacy from the people it serves, and that legitimacy has to be earned continuously.
Why Engagement Often Fails
If the case for participation is so strong, why does it so often fail to deliver? Not because the theory is wrong! But because the dominant models were built for a different era, and even well-intentioned efforts, when filtered through broken processes, can deepen the cynicism they were meant to address.
This matters because bad engagement actively fuels the cynicism it was supposed to address. Every bad process reinforces the conviction that participation is theater (or participation is illegitimate, or corruption sanitized) and makes it harder to do the real thing next time. These failures matter in any policy domain, but for climate they are something closer to existential — not just for policy outcomes, but for faith in democratic governance itself.
The Design Failures Are Structural, Not Incidental
The most common problems with public engagement are baked into the dominant models. Traditional notice-and-comment processes, for instance, were designed for transparency, judicial review, and weighting technical and legal expertise over lived experience — and they achieve those aims in a narrow procedural sense. But transparency is not the same as accessibility, and access is not the same as influence. As Nicholas Bagley has argued, procedural rules now actively exacerbate the very problems they were designed to solve. Notice-and-comment tends to reward the organized, the well-resourced, or the professionally represented, producing voluminous records that reflect the priorities of those who could afford to engage, not necessarily the communities most affected by the decision. A single well-funded group can submit thousands of pages of technical comments; a frontline community facing the same decision may have no idea the comment period exists, and no capacity to respond to it even if they did.
Formal public hearings share similar pathologies. They are designed to create a record, not a dialogue, and they tend to produce exactly what their format invites. Participants deliver prepared statements into a microphone. Agency officials listen without responding. The atmosphere is frequently adversarial, structured in ways that entrench “us versus them” dynamics rather than creating any genuine opportunity for exchange, learning, or compromise. People leave feeling unheard not because their words weren’t recorded, but because nothing about the process suggested anyone was actually listening.
One-way communication runs across many engagement formats. As the IAP2 Public Participation Toolbox makes clear, even well-designed information-sharing tools — fact sheets, websites, press releases — have significant limitations when they substitute for genuine dialogue rather than supporting it. Listening sessions, informational webinars, and town halls designed primarily to transmit information leave the public as a passive audience (or unheard correspondent) rather than active participants. When people cannot ask questions, push back, or engage in real exchange with the decision-makers who affect their lives, the engagement reinforces rather than reduces the distance between government and community.
Timing compounds all of these design problems. Engagement that happens too late in the decision-making process — after the key choices have been made, after alternatives have narrowed, after political and financial commitments are in place — is unlikely to meaningfully influence outcomes no matter how well it is conducted. Collaboration works best when it begins early, when there is still genuine room to shape the purpose, alternatives, and design of a proposed action. By the time a formal public comment period opens on many major decisions, the substantive work is essentially done. Community input at that stage can affect the margins but rarely the fundamentals and communities know it. When people show up, offer input, and watch nothing change, they stop showing up.
The scale and complexity of engagement materials create their own barrier. Technical documents running hundreds of pages, written in regulatory language accessible only to specialists, are not a neutral feature of the process — they are a filter. Effective strategic communication requires understanding who the audience is, what they already know, and how the issue connects to their lives, none of which is served by dense regulatory documents distributed through official channels. When the baseline requirement for participation is fluency in administrative law and agency-specific jargon, the people best positioned to engage are lawyers and lobbyists, not the residents of a community downstream from a proposed facility. Simplifying materials is not dumbing them down. It is recognizing that the expertise of affected communities is just as relevant as the expertise of credentialed professionals — and that accessing it requires meeting people where they are, not where the agency finds it convenient.
Perhaps most corrosively, some procedural tools originally designed to protect communities have been repurposed as obstruction. When participation requirements become veto points and when the primary function of an engagement process is to create grounds for litigation rather than to genuinely improve decisions, they undermine both the efficiency that critics of government rightly demand and the democratic accountability that participation is supposed to deliver. Small, organized, well-resourced groups can exploit procedural requirements to delay or block decisions that broader communities support, effectively capturing processes intended for the many and wielding them on behalf of the few. This is the participation failure that most directly drives the “just build it” impulse and it is a legitimate grievance. A system that was designed to give communities a voice has, in too many cases, been captured, producing neither democratic legitimacy nor efficient delivery. At the same time, when communities are left out of upstream planning, these veto points often become one of the only avenues available to influence decisions.
The Barriers to Inclusion Run Just as Deep
Even when processes are reasonably well-designed, they often fail to reach the communities that matter most. The gap between who participates and who is affected often follows the contours of existing inequality with uncomfortable precision.
Awareness is the first and most basic barrier. Many people simply don’t know that an opportunity to engage exists. Believe it or not, federal agency websites are not where most Americans spend their time, and the Federal Register is not how most communities learn about decisions that will affect them (even for experienced users). When engagement opportunities are announced through official channels that already skew toward the educated, the connected, and the English-proficient, the resulting participant pool reflects those skews. Compounding this is a lack of clarity about what participation is even for: many people who are aware of an opportunity don’t understand how their input could make a difference, or whether it ever has. That uncertainty is itself a barrier, and it is one that agencies rarely address directly.
The materials and communications that agencies use to invite and support participation often create their own exclusions. Technical documents written for specialists, notices distributed through unfamiliar channels, comment periods with deadlines that give working families no realistic time to respond, engagement formats that assume reliable broadband and digital literacy — none of these are neutral design choices. A sound public engagement plan starts by understanding audience needs and building authentic, reciprocal relationships — the opposite of defaulting to formats that are convenient for the agency. Each design choice narrows the pool of who can meaningfully participate, and the cumulative effect is systematic. When engagement materials are only available in English in communities where many residents speak other languages primarily, the process has already decided who counts. When comment periods close before community organizations have had time to mobilize their members, the timeline has already decided who counts. These are decisions that agencies make, often without fully recognizing them as decisions at all.
Physical access barriers operate similarly. Transportation costs, distance to venues, the inaccessibility of meeting spaces for people with disabilities, the difficulty of attending a weekday hearing while working multiple jobs — these are all practical obstacles that might seem minor in isolation but compound into systematic exclusion for communities that are often most directly exposed to the environmental and infrastructure decisions being made. Virtual engagement has opened some of these doors, but it has closed others: digital access gaps, limited bandwidth in rural and low-income communities, and the particular challenges of meaningful online participation for older adults or those with limited technology experience mean that remote options solve some access problems while creating new ones.
Trust may be the most intractable barrier of all, because it cannot be addressed by logistical improvements alone. Communities that have experienced past harms from government (broken promises, extractive processes, decisions that hurt them and were made without them) carry a rational skepticism about whether this time will be any different. That skepticism is not ignorance or apathy. It is an accurate reading of a track record. Overcoming it requires more than a well-designed meeting or a plain-language summary. It requires demonstrated consistency over time: showing up when there is no immediate decision to be made, acknowledging historical harms directly rather than implicitly, and following through on commitments in ways that are visible and verifiable. This is precisely what Hollie Gilman and Sabeel Rahman mean when they argue in Civic Power that meaningful participation isn’t ultimately about better meetings — it is about redistributing power so that those closest to the problem are genuinely part of the solution.
Privacy concerns add another dimension that is easy to overlook from a position of power or privilege. In communities with historically complex or adversarial relationships with government – such as those experiencing overpolicing, immigrant communities, or other underserved groups – the act of identifying oneself in public participation can seem risky. Showing up on the record or speaking at a public hearing can feel less like civic engagement and more like exposure. These are not irrational fears; they reflect lived experience and how government has used information, access, interests, and related levers against communities. Genuine participation in these contexts will not occur just by invitation. Agencies must actively and continuously address the conditions that make showing up feel unsafe (e.g., by using intermediaries, anonymous or aggregated input, or other protective measures) and acknowledge the prior harm.
One additional factor shapes who participates and whether their participation endures: organization. Even when agencies improve conditions for participation, durable community voice does not emerge automatically; it depends on collective capacity, especially in communities historically marginalized by race or class. Without organizational support, individuals often lack the connections, information, and trust to translate lived experience into effective engagement. Organization helps aggregate that experience, sustain engagement, and convert it into usable input for government. In its absence, participation remains fragmented, reinforcing the advantage of those already organized and resourced.
The Benefits of Effective Engagement as a Strategic Asset
In the early 2000s, a proposed bridge crossing the St. Croix River (a National Scenic Riverway on the Minnesota-Wisconsin border) had been stuck in gridlock for five decades. When serious planning had kicked off in the 1980s and 1990s, the 1931 structure was already fracture-critical. Stakeholder groups and a disparate range of public institutions, representing sharply conflicting interests and mutual distrust, had reliably blocked every attempt to move forward.
What finally broke the logjam wasn’t a better technical study or a more powerful agency directive, a least common denominator solution, or a decision to simply ignore the interests of a set of stakeholders. It was a deliberate shift to structured collaboration rather than the usual method of talking at one another with concerns, with no means of finding points of alignment, resolution, or tradeoffs. Twenty-eight stakeholder groups shaped the bridge’s location and design, with a comprehensive mitigation package addressing the natural, social, and cultural impacts of the new bridge. Moreover,
Relationships and communication among the stakeholders improved remarkably during the problem-solving process. In the words of one stakeholder, “We were able to spend the time necessary to get over our natural inclination to not trust people from the other side. […] We had enough time and enough space to come to a conclusion that everybody could feel comfortable with.”
That outcome wasn’t the result of exhaustion, but of deliberate design and an emphasis on negotiation rather than government serving as an answering machine that never calls back.
Before making that case, it is worth naming what good engagement is not. Smart, well-designed participation is not a mechanism for local communities to veto decisions that serve broader public interests. When a neighborhood is asked whether it wants new apartments, or a community facing a new transmission line is simply asked whether they approve, the answer is predictably no — and designing processes that guarantee that outcome is not good participation! It is a design failure that produces exactly the NIMBY dynamics that rightly frustrate those who want to build. The problem is not that people oppose things. It is that poorly designed processes over-sample the most proximate opposition, structure engagement queries without a sense of purpose or audience, start with technicalities rather than longstanding trust, avoid the potential for early negotiation, or systematically exclude the regional, national, and future stakeholders who have just as much at stake in the outcome. These failures aren’t inevitable.
Good engagement produces better decisions.
Agencies have technical expertise, legal authority, and institutional knowledge, but they routinely lack on-the-ground understanding of how policies will actually land in specific communities, what tradeoffs matter most to the people affected, and what solutions might work that no one in a headquarters office has thought of yet. Meaningful participation fills those gaps.Well-designed engagement generates solutions that are more effective, empowers people from different backgrounds, and builds the local networks that make implementation actually work. It brings in lived experience that data alone cannot capture, surfaces local knowledge that improves policy design, and produces decisions that are more responsive to the full range of affected interests rather than the loudest or most proximate ones.
Research on structured deliberation consistently shows that when people are given good information and genuine opportunity to reason together — rather than just reacting to proposals that feel threatening — they reach more nuanced, durable conclusions than either polarized public opinion or top-down expert judgment alone produces. Deliberation, in this sense, is not just a democratic value. It is a technical tool for overcoming the polarization that makes hard policy decisions feel impossible. It is also, critically, a tool for helping people reason past their immediate self-interest toward a broader understanding of tradeoffs, which is precisely what is missing when participation processes sample only those with the most to lose from a particular change, rather than those with the most at stake in the outcome. Done well, it changes which voices dominate and acknowledges the role of power. That shift is the difference between a participation process that ratifies the preferences of whoever showed up and one that actually informs a decision. This is not an argument for endless input or process without limits. As James Goodwin argues in this collection, the most effective participation is targeted rather than open-ended, focused on the core disputes that actually need resolving, rather than generating voluminous records that obscure more than they illuminate. The goal is engagement that is both more inclusive and more purposeful: asking the right questions of the right people at the right time (which means going well beyond the immediate neighborhood to find the people whose lives will be shaped by a decision).
But asking the right questions requires doing the work before the room fills up. Good engagement doesn’t begin with an open-ended invitation to say whatever comes to mind. It begins with agencies doing enough homework (with experts and communities) to frame the problem clearly: what is actually being decided, what constraints are real, what tradeoffs exist, and where there is genuine room for public input to influence the outcome. That frame is one of mutual respect.
This is where standards matter, as explored later in this essay. The difference between engagement that produces insight and engagement that produces noise is largely a design question.
Good engagement builds the trust that makes government function.
Scholars have considered the consequences of low trust through many lenses, but the legitimacy of democracies relies on trust. Lower trust means less engagement with functions that government performs uniquely or drives, whether disaster response, weather warnings, federal benefits, security functions, public health, or independent data collection and analysis — and that lesser engagement means those functions work less well for everyone else. This has a cascading impact as democratic institutions weaken when government cannot, does not, or is not believed to deliver on expectations of its citizens.
When people experience decision-making as transparent, accessible, and genuinely responsive to their input, trust builds. When they show up and feel unheard, or never show up at all because no one made it possible, trust erodes, most quickly and most consequentially for those communities that can least afford to lose it. This isn’t incidental — trust is the medium through which every other benefit of engagement operates. When it erodes, the downstream consequences reach far beyond any single process, a pattern worth examining directly when we turn to why engagement so often fails.
While study after study shows that less than half of Americans trust the federal government, far fewer (21%) believe it listens to the public — and just 15% believe it is transparent, both according to a 2024 national survey. This underscores a deep perception that government is neither responsive nor accountable to the people it serves. Every engagement process is a small test of whether democracy is meaningful — especially when government is asking people to accept changes as visible and consequential as those required by climate policy.
Good engagement reduces conflict and can prevent litigation.
As Andy Gordon has argued at FAS, listening is a prerequisite for discovery and a requirement for success on any ambitious public goal where stakes are high. The ARPA-I national listening tour demonstrated this concretely: starting with questions rather than answers, and drawing on distributed expertise from every layer of the transportation system, produced an agenda-setting process that no small group behind closed doors could have replicated. The same principle holds for climate and environmental policy. The instinct to streamline participation (to save time, to avoid change, to avoid the NIMBY trap) often backfires in the most concrete terms. Decisions made without adequate input tend to generate opposition downstream, when it is far more costly to address: through litigation, organized resistance, implementation failures, and the kind of sustained community distrust that shadows projects for years. The veto-point problem, in other words, is not a feature of too much participation. It is what participation looks like when it arrives too late, is structured too adversarially, and samples too narrowly. Fix the design (broaden who is heard, start earlier, frame the problem honestly) and participation stops functioning as a veto mechanism and starts functioning as the evidence base that allows government to make hard calls with confidence. An agency that has genuinely sought broad, representative input is in a far stronger position to defend a difficult decision. When that engagement runs through organized, representative groups, it also enables more effective deal-making — where tradeoffs can be negotiated and outcomes reflect a broader, more representative set of community interests.
Collaborative approaches improve the quality of decision-making and increase public trust precisely because they bring the right stakeholders in early (engaging more upstream), before positions have hardened and before the public record has closed. Agreements get built earlier, key voices feel heard rather than steamrolled, impacts are better understood across not only the loudest but those most widely impacted, and mitigation is transparent. And with that, the potential litigation drops, compliance improves, and implementation becomes something communities feel invested in rather than something done to them.
A Framework for Doing It Right
Moving with the public is how government earns trust and makes decisions people can understand, accept, and stand behind.
Participation doesn’t eliminate conflict, and isn’t meant to. The challenge is structuring disagreement so decisions can still be made — and hold up to scrutiny — by clarifying tradeoffs, surfacing impacts, and narrowing options. Climate policy makes this especially clear; decisions about energy, land use, and infrastructure must move quickly while navigating disagreement about costs, risks, and local impacts.
Participation must reflect not just who shows up, but the full set of people affected, including those whose interests are less visible but equally consequential. Engagement can overrepresent highly organized or locally affected groups, even when decisions carry broader regional or national benefits. Designing participation to reflect those broader impacts — and the perspectives not in the room — helps avoid decisions that are responsive but not well-balanced.
The OMB memo advances a shift from participation as a single event (a hearing or listening session) to a practice agencies must design intentionally, tailor to context, and improve over time. The memo’s five principles — Purposeful, Respectful, Transparent and Accountable, Accessible, and Learning-Focused — offer a practical framework aligned with the decision, the stakes, and the people affected.
At their best, these principles help government:
- reach the right people (not just the easiest to reach)
- ask questions tied to real choices and constraints (not just general opinions)
- use input to shape decisions and follow-through (not just record it)
The goal isn’t consensus — it’s decisions that can move forward.Climate urgency makes getting this right non-negotiable. This framework is designed for exactly the conditions climate policy creates: high stakes, contested choices, deep skepticism, and no margin for processes that consume time without building trust.
Five Guiding Principles for Meaningful Engagement
Is it purposeful?
Purposeful engagement starts with clarity. What decision is being made? What is open to input at this stage? What is constrained by law, budget, or timing? Who is affected, and when can input still affect outcomes? It also means asking participants to respond to specific, decision-relevant questions. Engagement begins early enough for communities to help shape options, not simply react to them.
Why it matters
Without a clear purpose tied to a decision, engagement captures reactions to incomplete information or comes after key choices are already set. When choices are complex, participants may rely on partial or misleading assumptions rather than the factors shaping them. Framing questions around real constraints and tradeoffs produces more informed, actionable input. Inviting input outside an agency’s authority or capacity can also overwhelm staff and create unmet expectations.
What it looks like
- Engage before major choices are locked in.
- Present a limited set of realistic options (e.g., 2–3 alternatives).
- Ask targeted questions tied to specific considerations (e.g., cost, siting, mitigation, local impacts), not general preferences (e.g., “do you support this project?”).
- Define what input is in scope, and how out-of-scope input will be routed (e.g., partners, interagency coordination, future tracking).
Is it respectful?
Respectful engagement treats communities as knowledgeable partners and acknowledges the costs of participation. It reduces barriers where possible and ensures participation is worthwhile and relevant.
Why it matters
When engagement feels extractive, one-sided, or not worth the time required, participation drops and trust erodes, especially in communities already bearing environmental and infrastructure burdens. This affects not just who participates, but the relevance of the input received. People are more likely to engage — and stay engaged — when the process is clearly connected to decisions, provides enough context, and is worth their effort.
What it looks like
- Partner with trusted intermediaries to co-design or co-host engagement.
- Provide support (e.g., stipends, childcare, travel reimbursement) where feasible (see, for example, community compensation guidelines from the Colorado Department of Human Services and Washington State Office of Equity).
- Set clear norms for dialogue and how input will be documented.
- Train staff in facilitation, cultural competence, and navigating disagreement.
- Explain how input will be considered by decision-makers.
Is it transparent and accountable?
Transparent and accountable engagement sets clear expectations about what is being decided, how public input will be used, and how decisions will be communicated.
Why it matters
People don’t need to agree with outcomes to see them as legitimate, but they do need to understand how decisions were made. Transparency clarifies how input connects to decisions — including what can and cannot change — and helps maintain trust even when alignment is difficult. Accountability comes from closing the loop: showing what was heard, and what followed. Not all input will change outcomes, but its role should be visible.
What it looks like
- Communicate decision criteria, constraints, and timelines up front.
- Explain how input will be used before engagement begins.
- Share examples of how prior feedback influenced agency decisions or plans.
- After the engagement phase, show how input was considered by decision-makers (e.g., agency response to comments, draft revisions).
Is it accessible?
Accessible engagement removes practical barriers and proactively invites participation from those most affected but least likely to show up by default.
Why it matters
Barriers determine who participates and whose perspectives are heard. When participation depends on time, resources, technical fluency, or familiarity with government, input skews toward those advantages and may not reflect the full scope of public impacts. As a result, decisions rely on a narrower — and potentially distorted — set of inputs. Broadening access improves both representation and the quality of information decisions rely on.
What it looks like
- Offer multiple ways to participate (e.g., in-person, virtual, evenings, weekends) and provide input (e.g., written, audio, mapping tools).
- Use accessibility-by-default design principles (e.g., plain language, compatibility with assistive technologies, screen reader-friendly materials).
- Include stakeholders beyond those most visible or organized (e.g., future residents, regional beneficiaries).
- Conduct outreach through trusted community channels (e.g., local organizations, faith groups, libraries, community centers).
- Provide captioning, translation, interpretation, and other accessibility supports as standard practice — not only upon request.
Is it learning-focused?
Learning-focused engagement is iterative and adaptive. Agencies assess whether engagement is reaching the right people, producing helpful input, and informing decisions — and adjust accordingly.
Why it matters
Without iteration, engagement repeats the same gaps in participation and input. Learning in real time allows agencies to adjust how engagement is designed and delivered. It also prevents wasted effort by enabling course correction, saving resources and community goodwill — essential in fast-moving environmental and infrastructure contexts.
What it looks like
- Collect participant feedback (e.g., on clarity, accessibility, usefulness).
- Conduct internal debriefs to identify what worked and what didn’t.
- Monitor who is participating and who is missing.
- Adjust outreach, format, or framing mid-process.
- Use clear measures to assess the effectiveness of participation.
Is this engagement tied to a decision that is still open — and are participants being asked to respond to clearly defined choices, constraints, or tradeoffs?
Is this engagement worth people’s time — and are participants equipped to provide informed, relevant input?
Is it clear how input will be used — and how participants will see how it was considered?
Are we reaching and enabling participation beyond those most visible or organized — and who may still be missing?
Are we using what we learn to adjust this process in real time — and to improve future engagement?
Matching Methods to the Moment
No single method fits every situation. Effective participation matches the approach to the decision, based on who is affected, what is open to input, and what participation is feasible. Different stages of the policy lifecycle call for different forms of engagement:
- Early stages: define problems and surface lived experience.
- Mid-stage decisions: compare options and test tradeoffs.
- Implementation: troubleshoot and adapt.
The five principles set the standard for how to engage. The next step is choosing methods that apply them — fitting the decision, the audience, and the agency’s constraints (e.g., timeline, resources, legal requirements). Frameworks like the IAP2 Spectrum of Participation, the T.I.E.R.S. Public Engagement Framework, and the National Coalition for Dialogue and Deliberation’s Engagement Streams Framework can help avoid a common trap: defaulting to the same level or type of participation regardless of context.
Effective participation isn’t about asking more people more questions. It’s about selecting approaches that produce usable input.
That alignment also applies to who is included. Decisions with regional or national consequences require engagement that includes those who will benefit or bear indirect impacts. For example, housing or transmission projects often draw input primarily from current residents, even when benefits accrue to future residents or regional users.
This is not about giving any group veto power, but ensuring decision-making reflects the full distribution of impacts and interests.
Evidence from the Field
Well-designed participation isn’t just a process — it’s a governance tool. Examples from environmental and infrastructure policy show it can inform decisions, improve design, clarify contested evidence, and build the capacity for better engagement over time.
Participation that changes decisions
A 2025 study analyzing 108 Environmental Impact Statements under the National Environmental Policy Act (NEPA) found that public comments led to substantive changes in agency decisions in the majority of cases:
- 62% involved meaningful changes to decisions,
- 64% modified project alternatives,
- 42% changed mitigation plans, and
- when preferred alternatives shifted, agencies directly credited public input as the reason.
While the study did not assess outcome quality, longstanding NEPA success stories suggest these changes often strengthen project design, e.g., by identifying overlooked impacts, informing mitigation strategies, or incorporating local knowledge into technical analysis.
Place-sensitive design
Infrastructure decisions highlight the importance of place-sensitive engagement because impacts vary by location, history, and lived experience.
The Bipartisan Policy Center’s examination of a U.S. Department of Energy-funded carbon storage demonstration project in Illinois shows how early, sustained engagement helped build understanding and trust around geologic carbon storage. Engagement began years before site selection and relied on trusted local experts, multiple outreach strategies, and two-way communication to familiarize communities with the technology and its potential impacts. These efforts contributed to broad-based support and community willingness to host the project, illustrating how early engagement can shape perceptions of risk and benefit and improve the conditions under which projects move forward.
Similarly, analysis by Acadia Center and Clean Air Task Force found that opposition and delays were reduced, and public support for infrastructure grew, when clean energy planners took local siting and environmental concerns seriously and equipped communities to participate meaningfully.
These examples underscore an important balance. Place-sensitive engagement works best when local input is considered alongside broader system needs, so place-based concerns inform — but do not override — decisions with wider public benefits.
Joint fact-finding and shared inquiry
In disputes involving scientific uncertainty and contested values, joint fact-finding — where agencies, experts, and stakeholders collaboratively define questions, gather evidence, and interpret findings — produces more credible, usable information. Rather than positioning agencies and communities as adversaries, these approaches shift focus from competing claims to shared inquiry, helping participants develop a common understanding of facts and tradeoffs.
Environmental dispute cases show that joint fact-finding can narrow disagreements and reduce mistrust even when consensus is not possible. In practice, these processes help participants clarify what is known, what remains uncertain, and where value-based disagreements persist — allowing decisions to move forward on a more transparent and informed basis.
Environmental justice case studies documented by the U.S. Environmental Protection Agency (EPA) likewise illustrate how collaborative inquiry can enhance participation and buy-in from affected communities. In several cases, involving community members directly in data collection and interpretation improved the relevance of findings, increased confidence in the results, and fostered more constructive dialogue between agencies and communities — strengthening both the substance of decisions and their implementation.
Community-led research and data governance
Community-based participatory research offers another pathway to stronger decisions by changing who controls the production of knowledge. Analysis from the Brookings Institution shows that when communities help set research priorities and interpret findings, the results better reflect local context and needs. Traditional research models often reflect externally defined agendas that lack community-specific knowledge, limiting their usefulness for decision-making.
Community-led approaches, by contrast, redistribute control over research and data governance, enabling communities to shape how information is generated and used. While joint fact-finding focuses on how agencies, experts, and stakeholders collaboratively interpret evidence in decision-making contexts, community-led research changes who sets the agenda in the first place. In practice, this can produce more relevant inputs for policy and planning, strengthen the connection between data and lived experience, and support ongoing partnerships that extend beyond a single engagement process.
Evaluation
Effective engagement improves through feedback. For example, the U.S. Army Corps of Engineers’ evaluation of public involvement in flood risk management pilots found that engagement tied to clear decision points and structured activities (e.g., working groups, facilitated discussions) strengthened agency capacity for public involvement and improved two-way dialogue with communities. Participating staff reported that these efforts helped teams better understand community concerns, identify information gaps, and structure engagement more systematically.
Internal debriefs and participant feedback informed adjustments across project phases, helping teams refine outreach, coordination, and how input is organized and applied.
These findings spotlight another benefit of well-designed engagement: not just contributing to individual decisions, but building the knowledge, relationships, and processes that make more informed and collaborative decision-making possible.
What These Examples Show
Taken together, these examples point to a clear pattern: engagement works best when tied to decisions still being shaped and structured to produce usable input from those affected. In these conditions, participation does more than gather input — it improves decisions and delivery.
As noted earlier, participation doesn’t remove disagreement. It makes it more manageable by clarifying what is known and uncertain, surfacing tradeoffs, and reflecting a more balanced set of perspectives. This better equips decision-makers to explain and defend their choices.
Building Capacity to Deliver
Well-designed participation doesn’t substitute for agency capacity — it sharpens it, especially at the state and local levels where timelines are tight, staff are limited, and decisions are high stakes.
Participation is often treated as an added burden on already stretched institutions. But when targeted and structured, it helps agencies use existing capacity more effectively by identifying concerns early and reducing downstream conflict, redesign, and delay. This isn’t just an equity argument; it’s a speed and delivery argument.
What matters isn’t whether agencies “have capacity,” but whether participation is designed to support decisions that can be explained and sustained.
Strengthening State Capacity
Next, let’s look at the specific elements necessary to improve public participation.
Leadership and Governance
Why it matters
Engagement succeeds when it is treated as core governance, not a communications add-on. When leaders treat participation as part of decision-making, it affects how processes are designed, how staff are incentivized, and how tradeoffs are handled.
What it looks like
- Designate engagement leads to coordinate across programs and align input with decisions that cut across issues or policy areas.
- Embed clear ownership within program teams.
- Build engagement milestones into project timelines.
- Have senior leaders review engagement summaries alongside legal, technical, and budget analyses.
- Reflect participation goals in agency strategic plans, implementation plans, and performance reviews.
Skills and Culture
Why it matters
Engagement failures are often organizational, not technical. Without the right skills and norms, staff may struggle to use public input or navigate conflict, and even well-intentioned engagement can break down.
What it looks like
- Train staff to interpret and weigh public input alongside technical and operational constraints.
- Synthesize input into themes and areas of agreement or tension so it can be compared and shared without revisiting individual comments.
- Design engagement plans across functions (policy, legal, technical, communications).
- Set expectations that engagement is part of policy development and implementation, not a parallel process.
- Reinforce these norms through staffing, timelines, and accountability.
Tools and Resources
Why it matters
Without the right tools and resources, engagement can generate more input than agencies can realistically analyze or respond to. The issue isn’t just volume, but whether input can be organized, interpreted, and applied to decisions.
What it looks like
- Use practical checklists and templates for planning, documentation, and follow-up.
- Use formats that produce comparable, decision-relevant input (e.g., facilitated discussions, guided prompts, prioritization exercises).
- Use analysis approaches that account for different types of input and perspectives (e.g., written comments vs. oral input, surveys vs. community discussions), not just the easiest responses to process.
- Establish pre-approved contracting mechanisms for facilitators, translators, and interpreters.
- Track commitments and follow-through internally (e.g., via dashboards).
Learning and Improvement Systems
Why it matters
Adaptation requires more than intent. Learning at the project level isn’t sufficient — without shared systems, agencies tend to apply the same approaches across teams and over time, regardless of effectiveness.
What it looks like
- Standardize internal reporting that connects public input to decisions and next steps (e.g., “what we heard / what we’re doing” summaries).
- Establish criteria for adjusting outreach or formats based on participation patterns (e.g., extend timelines if turnout is low, add targeted outreach to missing groups).
- Create shared repositories so lessons learned carry across teams and inform future projects.
- Incorporate outcome checks over time (e.g., whether engagement reduced conflict or improved implementation).
Strengthening Public Capacity
This challenge extends beyond agencies themselves. Participation design should account for both agency capacity and who can realistically participate. When engagement skews toward people who face fewer barriers to participation, it raises equity concerns and weakens the quality of information and problem-solving.
Reducing Participation Barriers
Why it matters
Reducing barriers isn’t about paying for feedback. It’s about making participation feasible, informed, and reflective of those most affected, including those who will experience the long-term impacts.
What it looks like
- Provide context in advance so participants can engage without technical expertise.
- Provide information in multiple languages and accessible formats.
- Share clear examples of useful input (e.g., specific impacts, leading practices).
- Present side-by-side comparisons of relevant options and tradeoffs (e.g., anticipated impacts, costs, timelines).
- Offer opportunities for the public to ask clarifying questions before providing input (e.g., Q&A sessions, office hours).
Equipping Communities to Engage
Why it matters
Meaningful engagement often requires skills, time, and capacity that some communities — especially smaller or under-resourced ones — may not have. Without intentional outreach and resourcing, agencies hear repeatedly from the same well-resourced groups. Just as important, without support for organized participation, engagement struggles to translate into representative voice, particularly for communities historically marginalized by race or class.
What it looks like:
- Provide practical tools to support effective participation (e.g., how-to videos, comment templates, guides to agency processes).
- Share meeting materials early so communities have time to review, discuss, and respond.
- Provide small grants or stipends so trusted partners can convene discussions, and synthesize input, and support representative participation.
- Designate clear agency points of contact to help communities navigate participation processes.
Developing Long-Term Relationships
Why it matters
Engagement is faster and more collaborative when relationships already exist. When agencies engage only during moments of controversy, participation is more likely to feel reactive and transactional.
What it looks like
- Maintain engagement beyond one-off interactions, including through light-touch ways (e.g., periodic newsletters, community listservs).
- Maintain continuity in points of contact through clear handoffs and shared records so relationships and institutional knowledge persist as staff change.
- Communicate regularly and follow through outside active decision processes.
- Acknowledge past harms or broken trust.
- Return to communities after implementation to share outcomes.
Return on Investment
Trust, as this paper has argued, is the load-bearing wall. Investing in state and public capacity builds it — so participation helps decisions progress rather than stall. Well-designed engagement enables agencies to use input and respond credibly. When communities can participate fully, decisions are better grounded, easier to explain, and more likely to hold.
Looking Ahead
As engagement tools evolve, the same principles apply. Digital participation, civic technology, and AI-assisted analysis are already being used to help governments reach more people and make sense of large volumes of input. But without deliberate design, they risk introducing new exclusions and harms, such as unequal access, privacy concerns, and bias.
Internationally, several examples illustrate both the promise and the tradeoffs. France used AI tools to analyze millions of contributions submitted during the Grand Débat National, grouping input by theme to make citizen insights more accessible to policymakers; the process highlighted tensions between scale and nuance, and raised questions about how input is aggregated and whose perspectives are preserved. The UK government uses an AI consultation analysis tool to analyze consultation responses at scale, reducing manual work and improving the speed and consistency of analysis. Its evaluation documents practical challenges, including the need for human review and concerns regarding bias, accuracy across groups, public trust, and transparency in interpretation.
These examples point to real potential: making participation more scalable, accessible, and usable in decision-making. But they also reinforce the same lesson as the rest of this paper — the impact of these tools depends less on the technology itself than on how they are designed and governed.
The path forward is not to adopt emerging tools wholesale, but to hold them to the same standards.
- Purposeful: Use tools to help answer specific questions or synthesize input, not simply because they are new or project “modernization.”
- Respectful: Handle community data, stories, and participation with care, including clear consent and appropriate privacy protections, especially when considering AI tools beholden to terms of use agreements.
- Transparent and Accountable: Explain how input is collected, analyzed, and used, especially when automation or AI is involved.
- Accessible: Design for inclusion rather than assume technology access or digital fluency.
- Learning-Focused: Assess whether tools improve engagement and decision-making, and adjust or retire tools that do not.
Used this way, technology doesn’t replace judgment or governance; it strengthens both.
As agencies adopt new tools, they can also build on existing human infrastructure — such as promotores and other community-based messengers — that has long been used in public health and service delivery to support outreach, build trust, and connect agencies with communities.
When participation is designed to clarify tradeoffs, surface real impacts, and support accountable decisions, it becomes what democratic governance needs most right now, and what this collection argues is still within reach: a government that listens well enough to be worth believing in.
A Plan for Revitalizing the U.S. Auto Industry
It’s January 20, 2029. The new President, committed to revitalizing America’s once-dominant car and truck industry, requests a briefing on the current state of affairs. The news isn’t great. Global market share of the big American automakers has bottomed out at single digits, well below the nearly 25% it enjoyed in the 1990s. Factories that made cars for export are either closed or closing.
That’s because the rest of the world doesn’t really want American cars any more. Continued tension and uncertainty in the Middle East has kept gas prices high and underscored the security and economic vulnerabilities of overreliance on oil. Outside the United States, consumers are snapping up Chinese-made EVs that go hundreds of miles, charge in minutes, and sell for less than $10,000. Industry is similarly pivoting to electric freight. On the home front, Americans are angry that U.S. policy is preventing them from buying exciting cars hitting the streets in neighboring Canada and Mexico. Workers aren’t much happier, as trade disputes and tariff barriers cause the once-integrated North American auto industry to come apart at the seams. And the U.S. truck oligopoly still runs on diesel, adding cost to almost every consumer good.
The President sighs and asks their team for a plan. Fortunately, they have one.
The World is Jumping Ahead on Electric Vehicles, Whether the U.S. Likes It Or Not
The writing is on the wall: electric vehicles (EVs) are the future. Industry knows it. Workers know it. Investors know it. Politicians on the left and right know it. EVs are simply getting too good, too quickly, for even the most pro-fossil policy agenda to stop. That’s good news for our climate, and for consumers globally, who deserve access to the latest, greatest, and cheapest cars. But it’s bad news for Americans, who are going to suffer the long-term effects of short-sighted transportation policies that fail to acknowledge economic and technological reality.
Eventually, this reality is going to become impossible to ignore. EVs are already dominating in countries as diverse as China, Norway, Vietnam, and India, and the geographic spread is only accelerating. Sooner or later, the United States is going to need a plan to catch up. It’s clear that what the plan won’t be is a simple return to the outdated regulatory frameworks we used for decades to regulate vehicle emissions and fuel economy. We need to think more broadly and more creatively.
Why Nixon-Era Statutes Won’t Catch Us Up On Their Own
Two statutes have historically governed vehicles. The 1970 Clean Air Act (CAA) structures the tailpipe emissions standards that the EPA and California enforce, while the 1975 Energy Policy and Conservation Act (EPCA), which charges DOT with setting separate (albeit overlapping) fuel economy standards. These laws, as we’ve previously written, were helpful for making gas-powered cars cleaner and more efficient, but they weren’t designed to guide a wholesale transition from fossil to electric technology with the pace and comprehensive scope needed.
There’s a case to be made for doing your best with old tools if those are the tools available. Indeed, that’s why the Biden administration’s vehicle regulatory regime was largely predicated on the CAA and EPCA. But these tools have become far less available in light of the second Trump administration’s actions. The Trump administration, in partnership with Congress, has dismantled the legal basis for regulating greenhouse gas emissions from vehicles under the CAA, gutted the fuel economy program that EPCA sets up, and rolled back California’s EV rules (which traditionally set the most stringent, future-forward vehicle standards). It also fired lots of people from federal agencies who understood how federal vehicle regulations worked. Thus from both a legal and a capacity angle, CAA and EPCA have become significantly weaker policy levers. Compounding these issues is the fact that the Supreme Court (now and likely into the near future) is profoundly hostile to regulatory ambition, limiting the probable reach of new rules set under these statutes.
Even if one could magically snap Biden-era vehicle rules back in place, those rules would only cover one player in the auto market – automakers themselves. Telling automakers that they have to develop cleaner, more innovative vehicles is an important tactic, but it is not a comprehensive policy agenda. CAA and EPCA don’t cover worker pensions, factory refits, trade policy, or continental industrial strategy. They don’t deploy charging networks or build out the grid. They don’t help fossil-fuel-producing counties diversify their economies, don’t structure smart spending priorities for the transportation system, don’t thoughtfully manage refinery closures and other giant infrastructure shifts. And as we’ve seen, CAA and EPCA don’t put real pressure on giant incumbent companies to change, and to make affordable new products, rather than to lobby hard against rules they don’t like.
By 2029, many of these core economic and industrial policy areas will also be in shambles. And as we’ve learned from the partial repeal of the Inflation Reduction Act, a party-line reconciliation bill won’t be a silver-bullet fix. Reconciliation bills are, as the Trump administration has made clear, no basis for stable investments. Fiscal policy is useful. Unstable budget policy will not draw in the billions in investment we’ll need to catch America back up.
These aren’t ancillary problems. Failing to manage the EV transition as the giant economic shift it actually is makes the policy and politics of the transition brittle and subject to reversal.
And Neither Will Ad Hoc Dealmaking
We need approaches that both compel and encourage giant incumbent companies to change and make affordable new products, rather than lobby hard against rules they don’t like. We need approaches that take the lived experiences of people who can’t afford to finance a new car, who don’t have access to charging at home, or whose livelihoods depend on the fossil economy seriously.
Yet the United States has never had a clear, cohesive, and comprehensive electrification and innovation strategy for the auto sector. In the absence of a clear national program, we’ve cut a series of tenuous deals to structure nearly twenty years of U.S. vehicle policy.
President Obama, during the 2009 auto bailout, was able to use fiscal recovery funds to steer the auto industry into accepting a first round of emissions rules from the feds and California. This “car deal” set the template for several more. The first Trump administration (with tacit support from automakers looking to slow EV investment) blew it up. California stepped into the breach with another set of deals, persuading companies to keep making EVs in exchange for regulatory relief and by threatening to stop buying government cars from holdouts. Team Biden then used Inflation Reduction Act funds to provide EV rebates and support EV manufacturing, as a spoonful of sugar to help a new round of federal and California emissions rules go down. And then the second Trump administration blew up those deals, again with industry support as American companies pivoted back toward short-term profitable gas SUVs.
Same story with trucks. California made a deal with the truck makers to continue progress on clean trucks. Post-Trump 2, those same truckmakers sued California to blow up their own deal. They are now backing Trump on his effort to destroy the entire CAA greenhouse gas vehicle program.
These deals were crucial progress at the time – and it would have been rational for the industry to keep its word. But it didn’t – policymakers shouldn’t be fooled again, or lulled into thinking that even good deals are a long-term substitute for durable policy. Relying on ad hoc dealmaking is an absolutely wild way to run a major industrial sector.
No wonder Chinese companies, able to rely on consistent government policy, have surged ahead while American consumers struggle to navigate the increasingly byzantine and expensive world that is the American auto market. While American workers move their families to be near good-paying auto jobs only to see investment pulled back and factories shutter. And while auto companies, in the absence of a stable and dependable policy environment, almost inevitably engage in short-term profit chasing rather than longer-term innovation.
A Path Back to the Future for U.S. Auto Policy
What breaks this bad pattern? Green industrial policy, embedded in a bipartisan statute that can pass a filibuster-proof majority of the Senate, and that blends competition and innovation.
Because we need both. The U.S. auto industry is highly concentrated, with only a few companies dominating domestic sales of both cars and trucks. These companies are used to calling the shots and setting prices, which is terrible for everyone but CEOs. American auto CEOs are deathly afraid of real competition – witness how Tesla’s new affordable electric semi is sending major ripples through the entire freight sector. But it’s been decades since the United States seriously enforced antitrust laws. Rather, politicians from both parties have raced to erect steep tariffs shielding the American auto sector from essentially all real global competition. There’s a role for any government to play in supporting its domestic industries. But that role isn’t propping up failing legacy companies indefinitely, at the public’s expense.
A well-designed statute could strengthen market oversight and competition while also providing American companies with reasonable support. Don’t like the Chinese-backed EVs that are undercutting your market? Start making and selling better, cleaner, and cheaper vehicles – and we’ll help. This is the logic that motivated Canadian Prime Minister Mark Carney and Mexican President Claudia Sheinbaum to welcome Chinese competition while bolstering their domestic industries. The U.S. can catch up with the same approach. For instance, government can readily provide loan funds and policy support for overseas companies that want to scale up in the United States, in partnership with domestic manufacturers. Such joint ventures are win-wins: the foreign company gets a new market and learns how to build for the U.S. market; the domestic company gains process knowledge and the IP it needs to modernize its products. That is exactly what Toyota and GM did a few decades ago, when the U.S. industry faced major competition from overseas autos (then, from Japan). The strategy can work again, for cars and trucks and batteries. Throw in better federal loan terms for union facilities, and you have a worker-friendly, industry-friendly, pro-competition, pro-consumer package. And this time the government can even explore taking an equity stake. Because why not re-shore profits, too?
Is comprehensive statutory reform ambitious? Certainly. But anything less consigns core national economic and industrial strategy to the vagaries of judicial interpretations of decades-old pollution statutes. Half-measures will, given the context above, be essentially useless. As with riding a bike, we can’t afford to pedal slow. The fate of a vast national industry warrants real national vision, rooted in a democratic vision for ordinary people. Democracy repair is climate repair is industrial reconstruction.
To deliver that vision when the policy window emerges – when the next President asks for it – we’ve got to start structuring it now. We suspect it will include some or all of the following elements:
- A national electrification trajectory, established in law. No more trying to retrofit EV rules from fossil-era statutes. We need a clear, negotiated path for electrifying America’s auto sector and restoring our global competitiveness, with related provisions on battery life and durability. This path must be binding on manufacturers.
- An accelerated charging network. There’s truth to the industry argument that they can’t sell EVs without chargers. We need federal funding to fill gaps in the charging network that private capital won’t (taking inspiration and precedent for federal funding for rural electrification), while incentivizing additional private-sector buildout guided by federal reliability standards and provisions for incorporating chargers in building codes and urban planning requirements.
- A continental industrial strategy. North American EV manufacturing should be baked into the USMCA and trade policy. Do we focus on advanced manufacturing of batteries? On assembly? On new high-tech models? We don’t have to do it all, but we do have to decide and put real resources and legal muscle behind it.
- Trade policy that lowers prices and creates new manufacturing. Our current system both blocks affordable Chinese EVs with prohibitive tariffs and fails to transform domestic industry. Explore options like joint ventures that on-shore competition through domestic-foreign partnerships, paired with tariff adjustments tied to domestic policy supports.
- A real free market for vehicles. Today’s American vehicle market isn’t free. The few companies that dominate vehicle sales in America serve consumers poorly yet get bailed out when they fail to compete. Enough. It’s time to provide Americans with real choices. That means, at minimum, statutory terms ensuring that prices are transparent and fair (as California is currently considering) and resources for anti-trust enforcement and consumer protection.
- Genuine support for workers. Federal support (including funding, loan guarantees, land use and permitting policy, and offtake opportunities like fleet procurement) should be conditioned on high-road job standards (e.g., labor neutrality agreements, training programs (including apprenticeships and new career opportunities), community benefit agreements, and pension protections. This policy should move in tandem with pro-worker reforms to labor law, such as the PRO Act.
- Genuine support for communities. Local governments and communities seeking to attract new manufacturing, manage plant closures and repurposing, or diversify revenue streams will need support. Federal government can provide funds to patch short-term budget holes as well as technical assistance and capacity-building grants to manage these issues across silos – all while deeply engaging state and local leaders to co-create and iterate national auto strategy.
- Innovative fiscal mechanisms. Public money should be used to align and leverage private capital into serious modernization of the U.S. auto sector. The federal government can provide loans for new auto facilities and market hubs, make loans for EVs (especially expensive medium- and heavy-duty vehicles) as standard and easy to secure as loans for other vehicles, push prices lower by tying federal incentives to reduced MSRPs, and even consider funding some of these measures by taking equity stakes in firms it funds. There is plenty of private capital in the United States to harness towards a globally competitive auto sector. The role of public funding is to push it in the right direction.
- Durable implementation support and education. We’ve seen how the IRA failed to build a political constituency because it rolled out too slowly, and with too little political cover. This time, the national program needs to come with durable support to increase implementation capacity at all levels, for speed, and a clear political plan to keep the American people engaged on this crucial national project, sharing its rationale and working through both concerns and victories as they come.
What Needs to Happen Now
The comprehensive, statutorily based plan we envision is only so much wishful thinking today. But we can actually create the conditions to realize this plan in the future. That’s because states have very substantial economic development authority, complemented by the substantial financial muscle of large state-based green banks and infrastructure banks.
What if the next governor of California provided state-backed loans for any company seeking to expand an EV or battery firm with pro-union policies in the state? The governor might particularly encourage joint ventures, perhaps looking to the Chinese firm BYD’s existing facility outside of Los Angeles (which doesn’t make cars and semis now, but could be expanded to do so). This isn’t wishful thinking. California more or less created Tesla with over $3.2 billion in subsidies. It can now create a host of new competitors, but this time it can also be taking equity to ensure profits flow back into a “transition fund” in the budget that supports economic growth in counties facing declining oil production. It can simultaneously pass state laws strengthening oversight of the U.S. auto industry, forcing transparent prices and pro-competition market structures.
If California leads, it’s unlikely to stand alone. States like Michigan might well jump in, using its existing Office of Future Mobility and Electrification to draw in billions in investment in existing facilities and pairing public investment with scaled-up private capital in powerful capital stacks. So might Texas, Illinois, Ohio, and Georgia – states that have made significant investments in EVs and batteries that they, and their constituents, are loath to lose. After a year or so, we might reasonably expect an MOU linking a growing state coalition that is working to develop the capital instruments and policies needed to grow industry.
With this policy certainty, investments start flowing at scale and a positive cycle begins. States continue to implement the panoply of policies in their control that otherwise support economic modernization, from clean air planning to freight system electrification rules and programs to policies that help vehicle batteries profitably support the grid by storing energy. They update building codes, deploy chargers, insist on durable and quality products using their consumer protection laws, and steer public money towards necessary infrastructure. Critically, they root all that work in expanded capacity to navigate the “messy middle” of the ongoing clean-technology transition, ensuring that neither workers nor communities will be left in the lurch.
At the same time, far-sighted public officials, backed up by civil society and foundation support, start bringing together the interests the new President will need for a major statutory play. They bring in counterparts from Canada and Mexico, recognizing that the continent has to either work together or lag together. Soon, the collaboration starts looking like an actual industrial strategy at the scale of East Asia’s behemoths.
In this scenario, when the next President calls for an auto modernization plan, Governors across the country back her. So do many major investors, donors, local officials, and federal electeds. After all, they’ve already seen the future in their states, and it is in their interest to scale up the national program. When the gavel falls in the first year of the 121st Congress, the conditions for change are in place, and twenty years of shaky rules and ad hoc deals give way to a new approach – a lasting one.
Four Innovations Driving Climate Progress in State Government
Subnational governments—cities, states, and counties—took some of the earliest steps in the United States to protect air quality, water quality, and public health. Over the last several decades, as the federal government has wavered in its commitment or failed to take a leadership role on climate change, states and cities have continued to act, both by bold pronouncements and aggressive targets and in more quiet, subtle ways that have eased the path for clean energy and other investments. This has resulted in a diverse cohort of cities and states that have made great progress in advancing clean energy and other climate solutions and are well-positioned to continue this leadership and innovation.
However, the federal government is taking aim at subnational authority and leadership, including threats to states’ actions deemed “overreach.” This threat from the federal government not only limits the ability of subnational governments to develop innovative tools to address climate challenges, but can hinder opportunities for economic development by limiting access to lower cost energy solutions, stifling innovation in emerging industries, and hampering global competitiveness.
Subnational Governments as Innovators and Leaders
Subnational governments, in liberal and conservative states, have long been the drivers of clean energy and climate progress in the United States.
Iowa adopted the first renewable portfolio standard in the United States in 1983, which laid the foundation for the state to remain one of the top renewable energy producing states in the country. George W. Bush and Christine Todd Whitman took aggressive actions as governors to reduce greenhouse gas emissions. Like Iowa, this early commitment in Texas has proven durable as the state continues to lead on renewable energy and energy storage deployment – due in large part to the state’s work to accelerate deployment of renewable energy projects, including permit streamlining and speeding interconnection for new projects. Renewable electricity generation reached a new high in the United States in 2024 and four of the top five producing states were “red” states .
U.S. cities and states have also taken a lead to affirm global commitments to achieve greenhouse gas emission reductions. In 2005, the U.S. Conference of Mayors launched its climate commitment when 141 mayors committed to meeting the emission targets included in the Kyoto Protocol. That same year, Governor Arnold Schwarzenegger issued an executive order in California, establishing economy-wide greenhouse gas emission reduction targets for 2010, 2020, and 2050.
Subnational actions are far more than symbolic. As of today, 33 states have climate action plans, 24 states have economy-wide greenhouse gas emission reduction targets, and 36 states have renewable energy or clean electricity standards. These commitments spotlight state leadership and serve as a reminder of the role of states’ authority and commitment to environmental leadership and their ability to serve as a backstop to federal inaction. They are important in several important ways. These commitments to reduced emissions and clean energy targets send a signal to developers, investors, and other governments that they have customers, markets, and willing partners. This has driven investment in companies and created opportunities for workers and industries in these states.
Subnational leadership is equally important for reimagining the systems, governance, and institutions to make these targets, goals, and investments a reality. For example, cities and states have developed streamlined permitting and inspecting processes for residential solar and energy storage installations (167 jurisdictions in 47 states), identified least conflict areas for large scale renewable energy projects (California and Washington), and developed innovative finance structures and institutions to support clean energy investment and development (e.g., Green Banks in 29 states, the District of Columbia, and Puerto Rico). Scaling these innovations can accelerate the energy transition and boost economic development, workforce development, and local communities.
The Need for Subnational Leadership
Subnational leadership on decarbonization is a practical necessity. Modeling shows that meeting global emission reduction commitments requires actions by subnational governments and businesses. In the United States, existing commitments by subnational governments and businesses could reduce national emissions 25% below 2005 levels by 2030. State level action can be approximately cost-comparable to federal (top-down) action and it tends to focus more on electrification of energy end uses, clean energy, and direct air capture, all solutions that fall under the authority of subnational.
Subnational governments hold primary authority over infrastructure development, siting and development of energy generation and transmission, energy efficiency in buildings, and land use decisions that determine development and conservation patterns. For example, scaling electrification to levels needed to achieve climate goals requires massive build out of clean energy resources and installation of heat pumps and other electric appliances at the household level. At the same time, state and local economies and communities are intimately connected to legacy industries that have shaped local fiscal structures, workforce, and cultures. Therefore, realizing the transformation necessary to decarbonize requires strategic action by state and local government.
Current regulatory structures are not optimized to navigate the challenges of decarbonization. Decarbonization is a systems challenge that depends on successful transformation of technical, social, and economic systems. Decarbonizing the energy system requires building new, clean energy and transportation systems, while at the same time making the investments needed for an orderly transition away from carbon in legacy industries like oil and gas. This dual approach enables environmental progress, while also protecting the workers and communities reliant on legacy industry. This requires linking environmental goals and policy in alignment with economic development, industrial policy, and environmental justice and equity goals. Failing to take an integrated approach repeating patterns of earlier transitions that concentrated pollution and contributed to growing inequity and environmental justice problems. Subnational governments are well positioned to take this integrated approach.
Working at a state or local level means that policies can be tailored to meet local contexts (e.g., economic, political, or social) in a way that top-down federal policies cannot. This is especially important because “successful implementation” means more than reducing emissions. Successful implementation means meeting climate and environmental goals while promoting prosperity and equitable opportunities for all residents, businesses, and communities. Success depends on accelerating project implementation while also addressing affordability issues and promoting society-wide benefits. Successful implementation stories are needed across a diversity of places to demonstrate approaches that are easy for others to follow and will resonate with cities and states that face different political, economic, and social situations. State and local government provide the right scale for building these solutions.
Challenges to Subnational Leadership and Innovation
It must be noted, especially at this point in time, that while the United States has a strong tradition of subnational leadership, it is not guaranteed. Subnational governments are experiencing many challenges that threaten to erode their leadership position—some inherent in the nature of the energy transition and other external factors.
A challenge of the energy transition is the need to build up new, clean energy systems while also carefully phasing out old, polluting systems. Governments need to accelerate investment in new clean energy, while continuing to invest in legacy industries. Failing to do both can introduce threats that can erode or even stall progress on decarbonization. California is experiencing this challenge now as the state navigates the impact of reduced demand for gasoline, a success of the state’s clean transportation policies, and what that means for the state’s oil and gas industry. The instability for oil and gas has resulted in the planned closure of two of the state’s refineries. While these closures will reduce pollution in host communities, they will also result in lost jobs and revenue and have the potential to increase the price of gasoline for California consumers. These dynamics have required the state to take actions to stabilize the oil and gas industry, while also accelerating its clean energy investments.
These challenges, inherent to the transition, are being exacerbated by other factors that threaten to erode subnational leadership and innovation, including:
- Rapid policy shifts at the federal level including termination of federal funding for clean energy projects, permitting uncertainty, and other disruptions that create significant uncertainty for governments, project developers, and labor.
- Growing inequity resulting from systematic racism, economic structures, and lack of opportunity poses a risk that some communities and populations will be left behind in the energy transition.
- Attacks on subnational authority to establish and implement climate, energy, and environmental policies.
While daunting, these challenges make the need for subnational innovation more important than ever.
Current Opportunities for Subnational Leadership and Innovation
Now, more than ever, subnational governments need to be places for regulatory ingenuity and action in the face of strong headwinds. The urgency of climate change requires immediate acceleration of the implementation of climate solutions—to reduce greenhouse gas emissions, protect public health and wellbeing, align economic development with environmental goals, and build resilience to changing climate and extreme events. Cities and states are best positioned to design policies to accelerate clean energy, innovation, and economic development because they can design approaches that work in different social, political, and economic contexts.
Innovation lies in the “how”—how to scope the challenges and design solutions that recognize the complexity of the decarbonization challenge. Subnational governments have already demonstrated several the power of regulatory innovation in several areas that show how rethinking regulatory and governance systems can accelerate progress:
- Innovation 1: Least conflict siting
- Innovation 2: Automated permitting
- Innovation 3: Making projects work for all
- Innovation 4: Deploying innovative approaches to funding and finance
These innovations by subnational governments show how government can accelerate the deployment of clean energy and other climate projects, implement projects that work in specific contexts, and deliver benefits to people and local economies.
Innovation 1. Least Conflict Siting
Uncertainty, conflict, and complex permitting and siting processes are major impediments to project implementation. Developing new approaches to siting and permitting can reduce uncertainty and delays that can increase project costs and diminish developer confidence. State and local governments can develop and deploy several innovative tools that can make it easier and less expensive to implement climate solutions, while also increasing transparency and engagement.
Cities, states, and counties can improve the siting and permitting processes by removing barriers to implementation, engaging with diverse stakeholders, and reducing costs for developers, government, and residents and businesses. For siting, this can include thinking at a regional or multi-project scale to identify priority areas for development. Engaging stakeholders early in the process can reduce objections and challenges later in the project process. Least-conflict siting processes provide one approach to innovate in the siting process.
Siting Innovation: Least Conflict Siting Process
Least conflict siting is a data-driven, participatory siting process guided by stakeholder priorities. A least conflict siting process uses spatial data that reflect stakeholder priorities (e.g., prime agricultural lands, sensitive habitat, etc.) to identify areas for infrastructure siting that avoid areas of high conflict. Using a participatory, stakeholder-driven process can avoid conflict at later stages in the development process, reduce uncertainty for developers, and provide a more transparent process for local residents, business, and other stakeholders. By identifying least conflict lands, stakeholders can then focus on removing obstacles to development on those lands (e.g., access to transmission). The process is non-binding, so can be adjusted as conditions or priorities change.
A least conflict siting process has been used in two regional contexts. The Center for Law, Energy, and the Environment and the Conservation Biology Institute piloted a least conflict siting approach for solar energy development in the western San Joaquin Valley in California in 2016. The project aimed to identify areas with least conflict for renewable energy development in a six-month process. A least conflict siting process has also been piloted for the Columbia Plateau in Washington. The project spanned eight months and concluded in 2023.
Both processes used geospatial data made accessible to all stakeholders through a collaborative gateway. Following this pilot program, Washington State passed a law in 2023 to improve project siting and permitting includes this least-conflict approach as a tool to be referenced for large scale renewable projects.
While this approach was developed in the context of large-scale renewable development, a least conflict-type process can be applied to a range of project types. This could include minerals mining, carbon removal, or transmission projects. The least conflict approach surfaces priorities and concerns in each region and the results of the process can be applied to different types of climate and energy projects, providing an opportunity to provide efficiency. A least conflict siting process requires commitment from a local permitting authority (e.g., local government), project developers, and stakeholders, including environmental, business, economic development, and other groups. The process also requires some investment to establish a robust process, including: spatial data showing energy, environmental, and other characteristics on the landscape, robust engagement, and strong facilitation.
Innovation 2. Automated Permitting
Delays in permitting increase project costs for developers and households. Developing transparent and simpler approaches to permitting can reduce delays, errors, and costs. Analysis of rooftop solar permitting in the United States shows that nearly 80% of a system’s cost is attributable to “soft costs.” These include design, project management, permitting, inspections, and interconnection. Reducing these soft costs through streamlined and/or automated processes can significantly reduce the costs of these projects. Streamlining can address many stages of the permitting process, including application, evaluation, and inspection – saving time and money for applicants, installers, and permitting agencies. For large-scale projects, mapping the permitting process to identify opportunities for creativity, flexibility, and efficiency can improve the permitting process.
Permitting Innovation: Automated Permitting and Inspection
Cities and counties are the permitting authorities for many clean energy projects, including rooftop solar and storage systems. Permit Power is a U.S.-based non-profit organization focused on reducing the bureaucratic impediments to deploying residential solar and battery storage. The organization’s research finds that a typical residential solar installation in the United States is up to seven times more expensive than a similar installation in Australia or Germany. To address this disparity, Permit Power is working at the state and local level to advance permitting and interconnection reform to reduce barriers to residential solar and battery storage projects.
Developers have identified permitting and inspection as major barriers to residential solar and storage project deployment—increasing both the cost and timeline for projects. Several states have adopted laws allowing or requiring cities and counties implement automated or streamlined permitting processes, while others have opted in on their own. The state laws that encourage or require automated permitting for solar projects have taken different approaches, generally with a nod to maintaining flexibility. New Jersey’s law directed a state agency to develop an automated permitting platform, but also to allow local jurisdictions to adopt an alternate platform with oversight from the state agency. Texas and Florida passed laws allowing the use of automated platforms but not requiring them. Maryland passed a law requiring local jurisdictions to use an automated permitting platform.
The Solar Automated Permit Processing Plus (SolarAPP+) provides a free, widely applicable platform to streamline permitting for rooftop solar and solar plus storage projects. SolarAPP+ is available free to jurisdictions designed to make the installation process easier for contractors and permitters, reducing needed staff resources, project timelines, and permitting delays. The National Renewable Energy Laboratory (NREL) developed SolarAPP+ in collaboration with industry and building inspectors. The platform streamlines permit approval for installations that meet specific requirements.
SolarAPP+ was launched in 2021. At the end of 2023, 167 permitting jurisdictions have adopted or piloted use of the SolarAPP+, and close to 600 additional jurisdictions have expressed interest in the application. Annual evaluations of the platform’s use show that permits for code-compliant systems is nearly instantaneous and that SolarAPP+ projects complete the full permitting process faster than installations that use the traditional permitting process. The evaluations also document savings in staff time and reductions in project delays.
SolarAPP+ is now managed by an independent foundation and is available to all jurisdictions free of charge.
Innovation 3. Making Projects Work for All By Using Community Benefit Tools
A risk of moving projects at a more rapid pace is the potential for harmful, unintended impacts on communities and the environment, including concentration of industrial activities, damage to habitat and natural systems, and reductions local quality of life. At the same time, these projects can bring economic development, workforce, and associated benefits to host communities. Community benefits tools can reduce conflict and improve project delivery by minimizing harms and harnessing benefits from these investments.
Community benefit tools can provide a mechanism for ongoing accountability and transparency for host communities, businesses, residents, and other stakeholder groups. These tools include community benefits agreements, community and cooperative ownership structures, and community oversight structures. If carefully crafted, community benefits tools have the potential to deliver meaningful benefits to infrastructure host communities, provide opportunities for community oversight and shared governance of projects, and reduce friction between developers and communities. Including community organizations and stakeholders as planning and implementation partners is vital to the success of these tools.
Subnational governments can require or create incentives for the development and use of community benefits tools for project deployment and they also have an important role to play in building community capacity to engage in the development and deployment of community benefits tools. They are also well-positioned to create the guidance and accountability tools to create the conditions for more effective community benefits structures. However, the existence of a community benefits agreement or related tool is not sufficient in and of itself, these tools need to be developed and designed well to deliver benefits to communities.
Importantly, linking community benefits to siting and permitting innovations can provide durable assurances of project performance and that a project will deliver benefits to a host community.
While requirements and incentive structures have promise, it is critical that tools are available to ensure that community benefits agreements are done well. These include guidelines for agreement development and technical and legal assistance for communities to ensure that agreements deliver real benefits to communities. It is worth noting that developing project by project community benefits agreements could result in two unintended and undesirable outcomes: the added process could slow down or discourage wanted projects and host communities end up with piecemeal benefits that cannot deliver meaningful and transformative investments in a community (e.g., comprehensive workforce development, integrated infrastructure investments, etc). Imagining scaled approaches (e.g., across a city or county scale) to deliver community benefits in a holistic manner across multiple projects in that area could increase efficiency for developers and support transformative investments in places hosting multiple projects or project elements.
Community Innovation: Models to Deliver Community Benefits
Community benefits agreements associated with development projects can deliver meaningful benefits to host communities, including investments in infrastructure, workforce development, and other community investments. State or local governments can require community benefits frameworks for projects in a specific geography (e.g., a community benefit ordinance) or create incentives for the development community benefits agreements or other structures to streamline project development (e.g., California Assembly Bill 205).
Detroit adopted a Community Benefit Ordinance in 2016. The ordinance requires that projects valued above $75 million or that receive significant subsidies from the city provide additional benefits to the community where a project is sited. When the ordinance is triggered, a Neighborhood Advisory Council from the project’s impact area is formed to work directly with the developer. The City of Detroit tracks progress on the commitments made through the agreements developed under the ordinance. Since its passage, eleven projects have finalized agreements under the ordinance and four more are in progress. Regular review of the ordinance’s performance has identified areas for improvement but monitoring shows that it has delivered measurable benefits in Detroit communities.
In accordance with California Assembly Bill 205 (2022), the California Energy Commission (CEC) has developed an Opt-In Certification program for clean energy projects including large-scale renewable energy (i.e., greater than 50MW), energy storage, and some clean technology industrial facilities. Through the opt-in certification program, the CEC can issue a permit for the project and enable it to forego permitting by local land use authorities and most, but not all, state permits. To qualify for the Opt-In Program, a project must meet a set of requirements, including entering into at least one legally-binding and enforceable agreement that benefits one of more community-based organizations in the project area (e.g., a community benefits agreement). For most qualifying projects, the Opt-in Program provides a faster timeline for environmental review (within 270 days of a project’s complete application, in most cases). However, implementation of the Opt-In Program is limited to date.
Innovation 4. Develop Innovative Financing Tools
Funding and finance for climate actions is a major barrier to advancing action, especially as the federal government claws back and reduces federal funding for energy, environmental, emergency response, and other programs that states, cities, and project developers have depended on. Now more than ever, developers, cities, and states need to be innovative in how they access and deploy funding and finance tools to support project development.
Subnational jurisdictions can initiate various revenue generation strategies to support project development. They can also access or establish different funding (i.e., grants) and financing strategies (i.e., loans) to support public and private project development. Subnational revenue generation tools include bonding authority, taxing structures, credits programs, and implementation of pricing programs (e.g., congestion pricing in New York City). They can also establish and/or access financing institutions like green banks and public-private structures to finance project development.
Subnational governments need to deploy a suite of revenue generation, funding, and financing strategies to support implementation and unlock access to private capital and investment. This is especially true given current threats to municipal finance, including withdrawal of federal funds, increasing climate risks and disasters, and the fiscal dimensions of the energy transition that affect local revenue structures. An analysis of options to support implementation of San Francisco’s Climate Action Plan found that the City needed to access all tools available to it to achieve the levels of investment necessary to implement the plan. This included development of new tax and fee structures, use of financing districts, integrating climate actions in the City’s schedule of general obligation bonds, establishment of a green bank, and implementation of pricing policies, including congestion pricing. Since the time of the analysis, the City of San Francisco passed Proposition A in March 2024, the city integrated climate-related actions into a scheduled general obligation bond to support affordable housing.
Financing Innovation: Public Finance for Transmission Infrastructure
Thirty states either have or are considering development of a green bank. A green bank provides access to capital for clean energy and other sustainable projects by issuing loans to projects that might otherwise have difficulty accessing capital. Programs within the California’s Infrastructure and Economic Development Bank (I-Bank), the State’s financing institution to support public infrastructure and private development projects that benefit California’s economy and quality of life, serve as the state’s green bank. Recent legislation established a program within the I-Bank to support investment in transmission infrastructure.
Transmission infrastructure is needed in many regions to distribute clean, renewable energy from where it is generated to load centers. California anticipates a quadrupling of in-state renewable energy generation by 2045, which will include offshore wind generation off the northern and central coasts, large scale solar generation in the Central Valley, and geothermal energy from the inland south regions of the State. Currently new transmission is funded though investor-owned utilities that pass costs on to ratepayers or by private developers.
To provide an alternative model, California recently established a public financing mechanism for new transmission infrastructure, the California Transmission Accelerator Revolving Loan Fund Program. The Accelerator will exist in the State’s I-Bank, the State’s financing institution to support public infrastructure and private development projects that will benefit California’s economy, jobs, and quality of life. The Governor’s Office of Business and Economic Development will develop a financing and development strategy in coordination with the State’s energy agencies to guide the implementation of the Transmission Accelerator. The program is designed to reduce burdens on ratepayers by using State funds to support needed transmission infrastructure development.
Closing Thoughts
We are at a critical moment for climate progress—given both the urgency of climate change and political polarization in the United States. However, this combination provides an opportunity for creative thinking and innovation at the subnational level. State and local governments have an opportunity, and perhaps obligation, to reimagine the regulatory, institutional, and governance structures to design decarbonization strategies that work for state and local economies, communities, and the environment. If undertaken at scale and implemented quickly, these subnational actions can have a significant impact on carbon emission reductions.
Some ways to help realize this innovation include:
- Build Subnational Capacity. Cities and states need additional resources and capacity to develop policies and programs, access funding and financing to implement projects, and partner with industry, communities, and labor to build durable solutions—especially in rural and under-resources areas. Building additional capacity requires creative use of existing public resources; partnership with the private sector, non-governmental organizations, research institutions, and philanthropy; and building networks for peer to peer learning, sharing, and collaboration.
- Leverage Networks to Replicate and Scale Successes. Subnational governments have strong platforms to share and scale successful policy actions. When President Trump withdrew the United States from the Paris Agreement in 2017, a bipartisan group of states established the U.S. Climate Alliance. The members of the U.S. Climate Alliance have committed, collectively, to achieve net-zero emissions as soon as possible and by 2050 at the latest. Climate Mayors, established in 2014, is a network of almost 350 mayors from nearly all fifty states that are committed to upholding the Paris Agreement. Together, these networks—and others—represent a significant share of the U.S. population and economy and provide a community of practice to support policy development and implementation. Subnational collaboration extends beyond the United States, with subnational networks like the Under2 Coalition that includes over 180 subnational governments across six continents committed to achieving deep GHG emission reductions. These networks can provide durable platforms for replication and scaling of successful policy innovations.
- Respect the Power and Potential of Subnational Governments. Subnational governments have a long legacy of leadership and innovation, due in large part to the country’s history of cooperative federalism. This system has provided room for creative solutions and policies tailored to each state’s unique economy, environment, and culture. Congress and the White House need to respect this and build capacity in cities and states to enable them to align economic and environmental goals.
A People-centered, Power-conscious Regulatory Democracy Balancing Distributive Justice and Delivery Efficacy
Building Blocks to Make Solutions Stick
People-centered, power-conscious rulemaking, using deliberate stakeholder engagement strategies, produces faster and better results.
Implications for democratic governance:
- Stakeholder engagement strategies must understand and design for power disparities.
- Public participation in policy processes should be non-performative, while also avoiding meaninglessly chasing consensus.
- Legitimacy through visible reciprocity – show the public how their input shapes choices.
Capacity needs:
- Stop treating engagement as read only/listen only; set participation expectations to respond, ask, shape, and negotiate.
- Build the institutional muscle to make stakeholder engagement a light lift (e.g., through reusable engagement templates and tools).
- Enable sequential and and tailored participation, and incentivize early steers that flag risk and opportunity.
- Train staff in facilitation, organizing, translation, conflict navigation, and other engagement skills or hire staff that can bring these skills.
One of the biggest obstacles confronting meaningful action to address climate action are power disparities. Among our governing institutions, the federal administrative state is unique in its potential for overcoming these power disparities by offering an effective mechanism for redistributing political power from corporate interests committed to maintaining the status quo to the general public who are already bearing the costs of global climate disruption. The key to realizing this potential is “regulatory democracy.”
At present, though, the means for conducting public engagement in the administrative state generally fail to meaningfully engage the public, but instead have the perverse effect of reinforcing power disparities, as has been ably documented by the burgeoning Abundance movement. What is needed, instead, is a better approach to regulatory democracy – one that is people-centered and power-conscious.
This paper sketches out what a people-centered and power-conscious approach for improving regulatory democracy in the rulemaking process: a reform called “Public Participation Planning.” This reform consists of two major procedural components. First, it calls upon agencies to develop “public engagement strategy blueprints” as a mechanism for deliberately creating a tailored public engagement strategy for each rulemaking. The key innovation here is a recognition that different kinds of “expertise” (democratic vs. technocratic) are required at different stages of a regulatory development – a concept referred to here as “sequential participation.” Second, it calls upon agencies to document the actual performance of that strategy, by including a group of documents called the Initial and Final Public Participation Plan Statements. These statements would capture the impacts that the public engagement actions have on the development of the rule. They would be included in the rulemaking document along with the notice of proposed rulemaking and final rules, respectively, where they can help inform judicial review of the rule. In theory, a president could implement a rigorous version of Public Participation Planning without new statutory authority. Still, the full potential of this reform could be enhanced with additional actions by Congress and the judiciary. Properly implemented, Public Participation Planning would both improve the quality of agency decision-making and permit for more expeditious policy implementation by reducing the ability of powerful interests to use the rulemaking process to reinforce a bias toward maintaining the status quo.
Addressing the Climate Crisis Through Better Regulatory Democracy
One of the underappreciated features of the federal administrative state – at least within our contemporary context – is its potential capacity to prevent politically and economically destabilizing concentrations of power from taking hold by continually redistributing it to the general public. That the architects of the modern administrative state – turn-of-the-20th century policymakers, thinkers, and movement leaders alike – designed it with this particular goal in mind seems to have been lost to history, however.
The key to their radical vision of the administrative state was “regulatory democracy” – that is, the notion that administrative agencies would work cooperatively (and sometimes competitively) with the public to shape policy priorities, design, and implementation. At its best, regulatory democracy would take the form of a working relationship that was ongoing and durable. The dynamic it was meant to yield would be a much thicker form of engagement in our governing institutions than ordinary Americans would experience through the episodic opportunities of casting a ballot and the often-binary choices they would be presented with during those opportunities.
This vision should be particularly resonant today, though it may sound esoteric and peripheral at first blush. For the reformers of the early progressive era, creating a new venue for translating public power into policy change was essential for effectively meeting the then-emerging challenges that many Americans faced due to such societal changes as industrialization and urbanization. Today, we face cascading challenges – climate change, globalization, and rapidly evolving forms of computational technology such as Artificial Intelligence and quantum computing – which have similarly exposed the limits of our governing institutions. Then, as now, society was characterized by vast disparities of economic and political power that further threatened effective policy implementation. The administrative state’s comparatively decentralized and democratized design – relative to Congress – was meant to mitigate these effects.
More to the point, if we are to avert the worst consequences of the climate crisis, we will need to quickly revive these robust democratic traditions of the administrative state. After all, as the Green New Deal movement correctly taught us, power disparities are a root cause of this crisis; effectively decarbonizing our economy and investing in infrastructure that is hardened to withstand the unavoidable impacts of climate change will require confronting these same power disparities. Among our governing institutions, the administrative state is best equipped to meet this kind of challenge under these kinds of circumstances.
Achieving the full democratic potential of the administrative state will require some important reforms, however. As the primary law governing the operations of the administrative state, the Administrative Procedure Act (APA) establishes many of the mechanisms that agencies use for democratic engagement. For informal rulemaking, which has become a leading vehicle for administrative policymaking, the APA creates the notice-and-comment procedures. The benefit of decades of experience has revealed that these procedures not only fail to meaningfully engage large segments of the population; they also reinforce status quo inaction and the underlying power disparities that benefit from such inaction.
This white paper argues that what is needed instead is an approach to public engagement that distinguishes between the different kinds of stakeholders implicated by a given policy action and accounts for the underlying power disparities that define their relationships to the policy problem that the action is intended to solve. As discussed below, the passive, power-agnostic posture of notice and comment fails to accomplish either of these objectives. Among other things, a people-centered, power-conscious approach would require carefully cataloging the universe of relevant stakeholders, the barriers they face to meaningful engagement, and the kind of input those individuals would likely bring to inform a policy decision.
A people-centered, power-conscious regulatory democracy would also require deliberate attention to how best to obtain public input. It would demand that agencies have a ready toolbox of engagement tactics and solutions tailored to effectively obtain different kinds of input from different kinds of stakeholders. It would also require agencies to plot out in advance the different stages in their rule development process for deploying those tactics and solutions – a concept this white paper refers to as sequential participation.
This approach would depend for its success upon a sincere commitment to transparency and reciprocity with stakeholders. Agencies would need to continually communicate with stakeholders and be completely forthright in those communications – even when the news is not what those stakeholders will want to hear. They will also need to carefully document how public engagement was conducted throughout the rulemaking process and the role it had, if any, on the progression of decision-making at the different stages of that process.
Lastly, and perhaps most controversially, this approach should draw on the agonistic model of democracy. Practically speaking, that means the goal of regulatory democracy should be to surface and channel productive disagreement, rather than embark on a quixotic search for consensus or near-consensus on controversial policy matters. As explained below, powerful interests have used consensus-based approaches to decision-making as a kind of veto-gate to defend their preference for the status quo. Reorienting our expectations for regulatory democracy in this manner will thus permit meaningful engagement without unduly sacrificing timely policy implementation – a concern that has achieved greater prominence due to the Abundance movement.
These lessons could, of course, just as readily apply to reforming state- and local-level administrative procedures as well. Indeed, subnational governments are already playing a pivotal role in addressing the climate crisis, particularly while steadfast Republican obstruction has left Congress incapacitated on this issue. The state rulemaking procedures or public utility commission proceedings that are responsible for implementing these policies could be strengthened through a people-centered, power-conscious regulatory democracy program. For simplicity, however, this white paper will focus on developing a version of this approach that applies to the federal rulemaking process.
While there may be several different methods for institutionalizing a people-centered, power-conscious regulatory democracy, this white paper proposes the use of what it calls Public Participation Planning. Under this reform, agencies would develop tailored plans called public engagement strategy blueprints. The purpose of these blueprints is to ensure meaningful engagement by relevant stakeholder groups – especially those representing communities that are structurally marginalized or that historically have been excluded from democratic processes. Critically, these blueprints would account for the entire rulemaking process with the aim of proactively engaging particular stakeholders at stages where their input is most likely to be relevant and useful. This proposal would also call on agencies to document their outreach and engagement actions and any impacts they had on the proposed and final rule in a special report called a Public Participation Planning Statement, which would be made part of the rulemaking record.
The practical advantage of Public Participation Planning is that it could be instituted by a president with existing legal authority. Still, the proposal also outlines how the other federal governing institutions – including Congress and the judiciary – can help ensure that the benefits of public participation plans achieve their full potential. One important task for the coordinate branches would be to address whether and to what extent existing administrative law doctrines, such as Vermont Yankee, present barriers to achieving the full potential of Public Participation Planning for advancing regulatory democracy. It is also worth emphasizing that parallel efforts to reinvigorate the most important democratic institution in our constitutional framework – Congress – will also be necessary to realize the full potential of regulatory democracy. After all, the administrative state can only implement the laws that Congress passes. It will thus be more effective for the administrative state to leverage regulatory democracy to tackle something like the climate crisis if Congress were to pass legislation explicitly directed at that issue.
If properly implemented, a comprehensive reform program to accomplish regulatory democracy that is people-centered and power-conscious could be essential for addressing complex policy changes such as the climate challenge. As Public Participation Planning demonstrates, this approach would both improve the quality of agency decision-making and permit expeditious policy implementation.
Background: Regulatory Democracy at a Crossroads
Regulatory democracy – represented not just by the APA’s notice-and-comment procedures but also the National Environmental Policy Act’s analytical requirements and their state– and local-level analogs – has come under increasing criticism from several directions in recent decades. Concerns that the regulatory system is undermined by too much public participation stretch back to at least the Obama administration. At the behest of Cass Sunstein, then Office of Information and Regulatory Affairs (OIRA) Administrator, agencies during this era strongly embraced cost-benefit analysis and other technocratic decision-making tools as an apparent antidote to the irrationality and “mistakes” that ordinary lay people make. Under this approach, public participation was to be viewed with extreme skepticism – if not outright hostility – and thus minimized as much as possible.
More recently, the Abundance movement that has emerged in recent years has come to single out for criticism many of the existing public participation requirements in administrative law. According to this criticism, powerful entities often abuse such requirements as a means for delaying policies that they oppose. Conservatives have also begun to reject public participation, with President Trump having directed agencies to evade notice-and-comment procedures whenever possible during this second term. This move seems in keeping with his overall crusade to centralize administrative power in the White House and build something akin to an authoritarian administrative state.
At the same time, we have seen a growing movement among policymakers and advocates focused on expanding public participation opportunities. Notably, as part of his administration’s larger Modernizing Regulatory Review project, President Biden issued a memorandum providing agencies with guidance on how to strengthen public engagement in the rulemaking process – with a particular focus on marginalized communities. Among other things, this memo encouraged agencies to deploy various strategies for engaging members of these communities at the earliest stages of the rulemaking process. Separately, a group of progressive members of Congress have promoted a comprehensive regulatory reform bill called the EXPERTS Act. One of its provisions would create the Office of Public Advocate, which would be charged with supporting individuals and other underrepresented groups in the notice-and-comment process.
What these competing movements reveal is that almost no one is satisfied with the current approaches to regulatory democracy. This dissatisfaction, in turn, arises from both practical flaws and theoretical disagreements associated with these current approaches.
Practical Flaws of Regulatory Democracy
Due to poor design, the prevailing approaches to regulatory democracy generally fail to effectively engage most members of the public in critical administrative state tasks of decisionmaking and implementation. Worse still, in many cases, these design flaws can combine in ways that function to systematically exclude members of many structurally marginalized communities, thereby reinforcing the very power disparities that are often at the root of the policy problems that regulations are often designed to address.
Many of these approaches follow a rigid, one-size-fits-all design that prevents their implementation from being adapted to meet the unique demands that can arise in different policymaking contexts. This can be seen in the APA’s informal rulemaking context. The same basic notice-and-comment process applies for policies as varied as setting Medicaid reimbursement rates and regulating the use of non-compete clauses in employee contracts. Yet, each of these policymaking contexts involve very different kinds of stakeholders whose relationships are characterized by different kinds of power structures. As such, effectively obtaining input from these different kinds of stakeholders is likely to require different, tailored engagement tactics.
To be sure, the procedural requirements setting out the public participation mechanisms set a legal floor; a president generally has the authority under Article II of the Constitution to go above and beyond by adding new public participation strategies aimed at alleviating the flaws that arise from the mandatory approaches. Indeed, as noted above, the Biden administration undertook some steps along these lines. In general, presidents are unlikely to undertake such steps without a clear strategy for doing so due to budget constraints and the growing criticism of public participation in the regulatory system noted above.
One important adjustment a future president could take to help make regulatory democracy opportunities more inclusive for structurally marginalized communities is to introduce them earlier in the policy development process. Presently, most public engagement opportunities tend to occur relatively late in policy development, as the APA notice-and-comment process illustrates. By this point in regulatory development, many of the foundational decisions leading up to the regulatory proposal have been resolved, including problem definition and solution scoping (not to mention the decision to prioritize this particular rulemaking at all). The remaining issues left open for public input are the kind of esoteric or technical details that are typically well beyond the knowledge or expertise of ordinary people.
Put differently, the manner in which regulatory democracy is currently conducted fails to account for the “sequential logic” of the policymaking process – a logic that necessarily draws on different kinds of expertise at different steps. An agency’s authorizing legislation, of course, sets the key parameters for what regulatory actions an agency might undertake and how it might design those actions. From there, though, important decisions remain such as which “public problems” are worthy of priority attention and how best to begin constructing the scope of policy solutions to meet those problems. These earliest stages in the policy development process call for a more democratic form of expertise that finds its source in stakeholder’s lived experience and situated knowledge. This kind of input might, for instance, spur the EPA to prioritize tackling pollution from industrialized agriculture or determine how stringently the Consumer Financial Protection Bureau regulates the use of forced arbitration clauses in consumer contracts.
In contrast, the more conventional understanding of technocratic expertise tends to become more relevant at later stages, such as when decision-makers must refine policies to account for such complex questions as applicable legal constraints, economic factors, the state of technology, or the body’s toxicological mechanisms. These issues might include what control equipment should be required to limit emissions of a toxic air pollutant or the potential energy use impacts of strengthened appliance efficiency standards. Not incidentally, social philosopher and an early intellectual force behind the modern administrative state John Dewey memorably captured the ordinal nature of the policymaking process with his observation that “the man who wears the shoe knows best that it pinches and where it pinches, even if the expert shoemaker is the best judge of how the trouble is to be remedied.”
The upshot of this failure is that the regulatory democracy currently privileges the kind of technocratic input that well-resourced interests committed to maintaining the status quo are uniquely well positioned to provide. Those who lack access to this expertise – including the resources and training to obtain it – are thus effectively prevented from meaningful participation.
Another important adjustment that agencies could make to improve public engagement is to conduct more affirmative outreach to specific stakeholders. Instead, current regulatory democracy approaches adhere to an “open door” model by which agencies invite input on equal terms from all interested stakeholders and then passively wait to receive whatever input is provided. The notice-and-comment procedures, of course, best illustrate this model, and it has been replicated in other regulatory forums as well, such as the “lobbying meetings” during OIRA’s centralized regulatory review process. At best, this model favors stakeholders with the resources to monitor and answer open door invitations for participation. In contrast, individuals and smaller community-based organizations are unlikely to consult the Federal Register on a daily basis or to have the technical capacity to parse a large rulemaking proposal to identify whether and how it implicates their unique interests.
In the worst cases, entrenched interests can abuse the open-door model by leveraging their vast superior resources to excessively voluminous comments containing information of only marginal utility or relevance – a practice known as “packing the record.” As legal scholar Wendy Wagner noted, it is not uncommon for industry interests to submit hundreds of pages worth of highly technical comments, resulting in rulemaking records that are more than 10,000 pages in length. These stakeholders engage in this kind of gamesmanship because they treat the notice-and-comment process more as a prelude to litigation over the final rule rather than as good faith attempt to improve the quality of the rule. Significantly, from a power perspective, this record-packing scheme only works if the party engaged is able to back it up with a credible threat of bringing litigation – something individuals and community-based organizations are unable to do.
In effect, this gamesmanship has enabled powerful interests to install the courts as the primary locus of regulatory decision-making, on the apparent presumption that they will afford a more sympathetic audience for their policy arguments – usually in favor of maintaining the status quo – than what they may find at the agencies. This tactic has the additional benefit of contributing to regulatory ossification, as agencies seek to “bulletproof” their rules as much as possible to avoid adverse results on subsequent judicial review. The advent of Artificial Intelligence suggests that this problem could grow even worse in the future.
Significantly, various studies on the practice of regulatory democracy have documented the vast quantitative and qualitative disparities in participation that exists between unaffiliated individuals and organized interest groups. Empirical research on the APA’s notice-and-comment process in particular seems to confirm that the design of those procedures has the effect of systematically excluding most members of the public, particularly those from structurally marginalized communities. These results suggest that the notice-and-comment process is failing at its purported task of assembling something approximating comprehensive policy-relevant information for agency decision-makers by systematically depriving them of critical forms of “expertise” that tend to be in the exclusive possession of the individuals and communities who live closest to the problems that the policies are meant to solve. They are left to fix shoes without knowing where exactly they pinch their wearers.
The results of these studies also suggest that these kinds of shortcomings in the notice-and-comment process do not merely limit effective engagement; they also serve to exacerbate underlying power disparities. That is because the skewed perspective that agency decision-makers obtain through these procedures necessarily favors entrenched sites of political or economic power. In terms of substantive results, the public input obtained through the notice-and-comment process often translate into a strong status quo bias toward inaction – or, at best, towards actions that only minimally inconvenience empowered stakeholders. Thus, for example, the public comments for a rulemaking to address the climate crisis are likely to be dominated by fossil fuel interests – as opposed to those who are disproportionately harmed. To the extent that this public input creates a skewed picture for agencies of the harms that significant disruption to our climate systems will create, it risks militating against the kind of aggressive climate policies needed to avert these harms.
Lack of a Coherent Theoretical Basis for Regulatory Democracy
One of the major sticking points is that students of the administrative state have never really achieved something like a real universal agreement on why regulatory democracy is important in the first place. The business community and conservatives have questioned the legitimacy of the modern administrative state within the U.S. tripartite constitutional governing framework at least since the advent of the Great Society programs. In response, scholars have invoked a variety of democratic theories to salvage the constitutional legitimacy of the administrative state. (Though, for the most ardent of conservative critics, such as Philip Hamburger, no theory is likely to suffice.) In turn, the lack of a clear theoretical basis has hampered efforts to design effective public participation mechanisms – contributing to many of the practical flaws described above.
According to the pluralist model, the administrative state’s democratic features provided a forum in which competing interests could shape policy. As long as the forum provides a reasonably fair opportunity for engagement for all interested stakeholders, this theory assumes that the substantive results that emerge roughly approximate the common good.
Another theory is the civic republican model. Unlike the pluralist model, which holds that the administrative state becomes imbued with democratic legitimacy through the balancing of competing interests, civic republicanism starts from the position that some idealized notion of the common good exists externally from the administrative state. Revealing this concept of the common good – which, presumably, all stakeholders would recognize as worthy of their consent – can only be achieved through a process of careful reason-giving and deliberation. The administrative state’s democratic legitimacy thus hinges on its ability to enable such a process to take place.
The influence of these competing schools of thought can be seen in institutional and legal reforms over the last several decades. For instance, the Regulatory Flexibility Act’s procedural requirements aimed at ensuring that regulators account for small business concerns reflects the pluralist model, while the embrace of cost-benefit analysis sounds more in the key of civic republicanism. Regardless of the theoretical grounding, both have tended to promote the addition of new procedural embellishments to the existing notice-and-comment framework that produce suboptimal results. Specifically, they slow down the rulemaking process without improving the quality of regulatory decisionmaking, and they have reinforced power inequality by giving entrenched interests new tools for blocking or weakening new policies they oppose – that is, maintaining a status quo bias towards inaction. In turn, these consequences seem to be artifacts of a characteristic that pluralism and civic republicanism both share: an abiding belief that consensus can be achieved and should be the goal of regulatory decision-making.
Third Theory of Regulatory Democracy: Agonism
Yet, this shared belief has come under increasing criticism in recent years, leading scholars to entertain a third theory of regulatory democracy: agonism. This model begins from the premise that consensus is impossible to achieve in many policy contexts, particularly during times such as now marked by polarized discord and seemingly incommensurable division over values and worldviews. If conflict is inevitable, agonism posits, then the appropriate function for our democratic institutions is to channel that conflict so that it is as productive as possible. Under this view, the legitimacy of policy outcomes comes not from how they are ultimately resolved, since they will not be accepted as such by many stakeholders. Instead, legitimacy flows from affording dissatisfied stakeholders with an ongoing realistic opportunity to contest and displace those outcomes with those that more closely align with their preferences. Administrative law already contains agonistic features. Adherents of this model envision other institutional and legal reforms that would infuse a more agonistic orientation to regulatory democracy. For instance, they would require more frequent use of retrospective review for regulations and greater use of adjudication for policymaking in place of rulemaking where possible.
One of the practical benefits of redesigning regulatory democracy mechanisms consistent with agonism is that it could help prevent some of the abusive gamesmanship practices in the notice-and-comment process described above. By demanding consensus, prevailing theories of pluralism of civic republicanism create perverse incentives for entrenched interests to manipulate public participation mechanisms in order to prevent consensus from ever being achieved. For instance, such interests might seek to delay final action on a rule by packing the rulemaking record with the intent to manufacture uncertainty or continually raise new issues – rather than productively inform a regulatory decision. In this way, public participation becomes a tool for translating power into paralysis. This concern is well worth considering given, as adherents of Abundance liberalism have noted, such paralysis can be exploited by would-be authoritarians to support an agenda of democratic backsliding.
Modernizing Regulatory Democracy Through Public Participation Planning
In light of these challenges, it is time to consider not just incremental changes, but a fundamental rethink of how public engagement is conducted within the administrative state. In contrast to current approaches, effective regulatory democracy must be both people-centered and power-conscious.
This paper will concentrate on applying this reform framework to the rulemaking process, though it could also be applied to other aspects of the federal administrative state that involve public participation, such as NEPA and permitting. Likewise, they could be applied to state-level administrative analogs. While there might be several ways to build a people-centered, power-conscious, rulemaking process, this paper outlines what it refers to as Public Participation Planning. This would involve agencies:
- developing and executing a strategy that is
- tailored to each of their planned regulatory actions in order to
- engage relevant stakeholders throughout each step of the rulemaking process
- with the explicit purpose of building a reasonably comprehensive record of the public’s views on the action for the rulemaking record.
The validation of Public Participation Planning as a viable mechanism for achieving truly meaningful engagement comes through its embrace of four cross-cutting principles: proportionality transparency, communication, and pragmatic learning. First, the rigor of a particular plan should be fairly proportional to the significance of the rule under development. Second, strenuous adherence to transparency is essential for achieving the agonistic goal of productive disagreement. In particular, agencies must always be completely forthright with all stakeholders about how decisions were reached and what evidence and arguments proved determinative. Third, and related to transparency, agencies should strive to maintain open lines of communication with stakeholders throughout the entire rulemaking process. This will enable agencies to serve as the effective mediators of productive disagreement through the rule’s development and beyond.
Fourth, it requires agencies to commit to an ethic of pragmatic learning. Implementing Public Participation Planning is not as simple as plugging a few numbers into an equation and expecting an optimal result to emerge; it is impossible to predict ex ante what will work in any given situation. Instead, to make the most of Public Participation Planning, agencies will need to become adept at building and rebuilding the proverbial plane, even as they are flying it. Moreover, what has worked in the past will have to be continually reassessed in light of underlying power inequities. If history is any guide, powerful incumbents will eventually devise ways to use their resource advantage to corrupt even the best public participation mechanisms for engaging structurally marginalized communities.
As discussed in greater detail, one of the obvious objections to Public Participation Planning is that its apparent emphasis on proceduralism will exhaust scarce agency resources and contribute to excessive delays in a rulemaking process that has already become too bogged down to permit for effective and timely policy implementation. These four cross-cutting principles, however, are intended to alleviate those concerns – adherence to them will help to strike the needed balance between public engagement and expeditious policymaking.
Putting Public Participation Planning into Action
Public Engagement Strategy Blueprints
One of the distinguishing features of Public Participation Planning is the requirement that agencies assemble a public engagement strategy blueprint for engaging stakeholders for each planned regulation at the time that the action is initiated that is tailored to the unique circumstances of the rule. This step provides agencies with a mechanism to anticipate potential challenges and think through and identify effective solutions that are calculated to enable them to build a reasonably comprehensive record of the stakeholders’ views on the rule.
It is worth emphasizing at the outset that the development of a public engagement strategy blueprints need not be a resource-intensive and time-consuming task. As noted above, these should be tailored to match the significance and controversy level of the rule. In addition, agencies will learn by doing, resulting in increased efficiencies over time. For instance, analyses used for past rules will often be readily usable for future rules addressing similar subjects. Another crucial source of efficiencies will be for agencies to use their existing institutional resources to perform these tasks. Most rulemaking agencies already have various forms of public engagement offices and regional and local offices that can and should be tapped. Public affairs offices can also be brought into the rule development process earlier, rather than announcing decisions – such as proposals and final rules – after the fact. Lastly, agencies may build new institutions that increase efficiencies for implementing public engagement strategy blueprints. For instance, agencies may consider creating a standing process for conducting periodic townhalls or other listening sessions. This relatively modest investment could in turn yield significant value for informing the development of public engagement strategy blueprints for future rulemakings covering a wide variety of issues.
More broadly, as explained in greater detail below, the implementation of all aspects of Public Participation Planning, including the public engagement strategy blueprints, should be thought of in terms as an investment. There is no denying they will involve the dedication of resources and time – mostly at the front end of the rulemaking process. But, by directly addressing power disparities and by more constructively channeling irreconcilable disagreements over the competing values implicated by a given rulemaking, Public Participation Planning will mitigate the sources of delay that crop up later in the regulatory process, including, most notably, litigation and the problem of ossification it creates. And to be perfectly frank, much of this work involves things agencies should be doing anyway. The failure to do so often helps to explain why many agency rules have fallen short of accomplishing their stated goals. In short, more “process” at the beginning will lead to less process and less delay overall.
Step 1: Public Engagement Strategy Blueprint
The first step in assembling a public engagement strategy blueprint is to conduct a thorough and deliberate stakeholder mapping exercise. By the time an agency has completed this exercise, they should be in a position to identify the relevant range of stakeholders for a given rulemaking and the anticipated role or roles they might be expected to play in the rulemaking process. For each category of stakeholders, agencies should also perform a general capacity assessment. Specifically, they should seek to answer such questions as what kind of input or expertise members of a given stakeholder group are likely to bring, whether and how those individuals have participated in similar rulemaking processes in the past, and what barriers might prevent them from participating effectively in the current rulemaking process.
Step 2: Examine Power Disparities and Structural Injustice
The second step is to assess the role, if any, that underlying power disparities or other forms of structural injustice (e.g., racism or patriarchy) play in contributing to the problem that the regulation is meant to solve. To be sure, agencies should be performing such assessments anyway since a failure to do so could yield policy responses that either fail or have other unintended perverse effects, including making the underlying problem worse. With this background in place, the agency should then give careful consideration to how stakeholders identified through the mapping exercise might help them to better understand these underlying power disparities.
Step 3: Stakeholder Engagement to Inform Rulemaking
The third step is to use any learnings from the stakeholder mapping exercise and power disparities assessment to construct a strategy for conducting stakeholder engagement to inform the development of the rulemaking proposal before it is formally published. Most agencies already have an established “action development process” they follow for drafting proposals and building a supporting evidentiary record. Agencies can build off the procedural framework that process creates when designing this engagement strategy. They can ask which types of stakeholders are likely to have input that would help with the successful completion of each stage of the policy development process and what specific engagement tactics would likely be most successful in obtaining that input at a reasonable cost in terms of time and resources. For example, agencies might use more time-consuming and resource-intensive focus groups for particularly weighty matters such as scoping out alternative regulatory designs. In contrast, they might employ informal remote public hearings to quickly gather ideas for sourcing certain kinds of evidence related to the rulemaking. (A bonus of this process is that it might reveal ways in which an agency policy development process could be strengthened, by adding, removing, or combining steps, or by altering their order.)
In preparation for this step, agencies will likely also want to have created a general library or “menu” of engagement tactics, with a brief assessment of their strengths and weaknesses. This will enable agency staff to quickly pull tactics “off the shelf” and insert them into the individual public engagement strategy blueprint. Indeed, this is an example of how moving along the learning curve will help agencies to implement Public Participation Planning more quickly and at reduced cost.
The capacity assessment performed during the stakeholder mapping exercise will especially be important for successfully implementing this step of plan development. For instance, if that exercise revealed that an important group of stakeholders is unlikely to have reliable access to high-speed internet, then the agency should refrain from relying on something like a remote public hearing to obtain input from those stakeholders. This assessment will also help agencies to identify potential affirmative steps they can take to eliminate barriers to public participation. For example, agencies can take steps to provide translation services if a large number of crucial stakeholders do not speak English as a first language. At the same time, the capacity assessment might reveal that a particular stakeholder group exercises an unusually high degree of dominance over a particular issue. In such cases, the agency may find that imposing certain constraints on their participation during the pre-proposal time period. These might include limiting or barring ex parte contacts or placing reasonable page limits on documentary submissions. (Such actions will also have the advantage of expediting the rulemaking process by preventing well-resourced contacts abuse these contacts as a means for delay.)
Step 4: Identify Mechanisms to Include Marginalized Communities, Including Storytelling
The fourth step in building a public engagement strategy blueprint is to identify mechanisms for ensuring that even stakeholders from structurally marginalized communities are able to participate in the notice-and-comment process as effectively as possible. As noted above, the notice-and-comment procedures often systematically exclude such individuals. While agencies cannot completely obviate this dynamic, they should still strive to sand off its worst effects – especially as these procedures are likely to remain part of the rulemaking process for the foreseeable future. Somewhat regrettably, the general consequence of these auxiliary mechanisms would be to get ordinary individuals to more effectively behave like sophisticated lobbyists instead of their true, authentic selves. This means providing various kinds of educational resources and specialized training to individuals so that their input can fit the “technocratic” mold, much as the Federal Energy Regulatory Commission’s (FERC) Office of Public Participation (OPP) undertakes now. It might also involve creating institutional mechanisms to serve as representatives or ombudsmen on behalf of unaffiliated individuals, though this would likely require significant additional resources and perhaps even legislative change to effectuate.
A more radical option would be to undertake institutional reforms that make notice-and-comment procedures more amenable to obtaining and utilizing non-technocratic forms of input, such as storytelling. This approach has the advantage of permitting individuals to share their more authentic expertise – including their situated knowledge and lived experiences – though such input may be of limited relevance at this later stage in the rulemaking process. The degree of institutional reforms required to fully realize this procedure – ranging from changes in agency hiring practices to modifications of administrative law doctrines to recognize these different kinds of “expertise” – makes it unlikely that this approach will bear fruit any time soon.
Step 5: A Plan for Public Engagement After Rulemaking
The fifth and final step in building a public engagement strategy blueprint is to create a plan for how the public might remain engaged after the rule is finalized – that is, to identify opportunities, if relevant and possible, for the public to participate in the rule’s implementation and ensure those are reflected in the rule’s final design. Examples of such engagement include the public’s role in monitoring compliance, measuring the rule’s impacts through citizen science activities, and holding regulated entities accountable for violations of the rule’s requirements through citizen suits when legally available. The final rule may also seek to explicitly incorporate opportunities for the public to participate in any future retrospective review actions for the rule, though Congress will need to ensure agencies receive sufficient budgetary resources to carry out such reviews. Similarly, many statutes authorize agencies to grant individual businesses different kinds of compliance relief, such as deadline extensions, variances, waivers, and exceptions. The final rule could provide the public with a meaningful role in considering and awarding these grants of relief.
As noted above, the rigor and detail of the blueprint should be roughly proportional to the rules’ economic and social consequences as well as to the level of controversy it is anticipated to engender. As with other aspects of implementing Public Participation Planning, accomplishing this proportionality goal in practice will improve with practical experience.
Resource constraints and political pressure for expeditious policy implementation are likely to provide strong incentives for agencies to fall short of the desirable level of rigor and detail. Consequently, countervailing incentive structures will be necessary to offset that tendency. Perhaps that could be accomplished through well designed judicial review standards, as noted below. Political leadership – including from the White House agency appointees – could also signal the importance of careful implementation of Public Participation Planning. For instance, this could be institutionalized through agency strategic planning exercises or encouraged as part of performance review and promotion decisions for career staff. Of course, Congress can do its part by fully funding agency implementation. And over time, as agencies advance along the learning curve for implementation, they will achieve increased efficiencies that will alleviate some of the incentives to do insufficiently rigorous Public Participation Planning.
Strategy Blueprint Implementation, Tracking, and Public Participation Plan Statements
As indicated above, each public engagement strategy blueprint that an agency develops should focus on creating meaningful participation opportunities for members of structurally marginalized communities early in the pre-proposal process, since that is when their input is likely to be of greatest relevance and utility for agency decision-makers. Rather than be a mere “check the box” exercise, the execution of these early participation mechanisms (informal hearings, focus groups, etc.) should have a discernible impact on the structure and substance of the proposal. Thus, as agencies turn to implementation of these mechanisms, they should carefully track whether and to what extent the public engagement strategy blueprint is accomplishing what they expected it would.
This, of course, is not to say that the agency should use these engagement activities to build evidence for decisions that were already made by other means – much as occurs with cost-benefit analysis now. Rather, it means that agencies should base their monitoring on other more objective benchmarks. One question agencies should ask is whether the quantity of participants matches the predicted expectations. (Again, agencies will likely struggle at first to make these kinds of predictions with much accuracy – there will be a learning curve. But, as noted above, a crucial ethic for Public Participation Planning is a commitment to learning by doing.) Similarly, agencies should find that the input they are receiving through these early engagement mechanisms are providing answers to the questions they need to answer to develop the proposal – whatever those answers happen to be. Another good indicator that the early engagement mechanisms are working well is that they are uncovering important “unknown unknowns” – things that the agency did not realize it did not know when it launched the rulemaking.
If, on the other hand, agencies are not finding that the early engagement mechanisms are working as expected – that they are not helping to build a reasonably complete record of public input on important policy-relevant questions undergirding the proposed rule – then they should make adjustments to the engagement strategy. This goal, of course, does not mean agencies should strive to accomplish something akin to comprehensive accounting for all relevant views from the impacted public. Instead, the goal should be to obtain a reasonably representative level of input from each of the major stakeholders included in the agency’s initial mapping exercise. What constitutes a reasonable level of input will necessarily be a subjective determination, and one that agencies will improve on as they learn through implementation of the Public Participation Planning scheme over time. In making this determination, though, agencies will want to be especially attentive to the concern that they have not adequately engaged members of stakeholder groups they have initially identified as being structurally marginalized or as facing particularly high barriers to participation. When in doubt, an agency may wish to attempt other forms of engagement for these groups. As in other forms of research, if the input they obtain sounds repetitive, that would indicate a good stopping point has been reached.
Drawing on lessons learned from actual practice, agencies may want to consider employing different forms of affirmative outreach to targeted stakeholder groups, undertaking alternative engagement tactics, or finding other creative ways to minimize barriers that might be preventing effective engagement. For example, if an important stakeholder category is young families, then the agency may consider securing resources to provide childcare during in-person hearings. To be sure, agencies may encounter legal constraints that may prevent them from instituting strategies like this. One suspects that these constraints are not as significant as feared, however, and that agency counsel have been overly cautious in interpreting these constraints. Nevertheless, clarifying legal authority from Congress on these matters would be welcome.
As it carries out the specific components of its public engagement strategy blueprint, the agency should begin assembling a comprehensive Initial Public Participation Plan Statement, which documents its outreach and engagement activities, carefully summarizes the input that was received through each component, and briefly explains what impact, if any, that input had on the agency’s proposal. Consistent with the principles of transparency and communication noted above, it is particularly important that the agency use this document to identify instances when a stakeholders’ input did not influence a particular outcome and explain why that was the case.
Explaining the Democratic Basis for a Rule
The agency should include the completed Initial Public Participation Plan Statement in the rulemaking docket when the rule proposal is formally published so that it is available to the public when they are developing the comments. In this way, the Initial Public Participation Plan Statement will function similarly to an Initial Regulatory Impact Analysis (i.e., the initial cost-benefit analysis), only it explains the “democratic” basis for the rule instead of its “economic” basis. Ideally, as the Initial Public Participation Plan Statement becomes more institutionalized, it can even replace the Initial Regulatory Impact Analysis that agencies now perform as the most prominent supporting document for a proposal. This would conserve agency resources and symbolize that democracy has replaced technocracy as the key driver of regulatory decision-making.
After the proposal is published, agencies should likewise carefully monitor the implementation of any specific components from the public engagement strategy blueprint for supporting public participation while the public comment period is open. Again, they should strive to make appropriate adjustments whenever they discover that these mechanisms are not producing expected or helpful results. During this period, agencies should continue documenting their progress in implementing the public engagement strategy blueprint by updating Initial Public Participation Plan Statement.
In conjunction with releasing the final rule, agencies should then include in the rulemaking docket a Final Public Participation Plan Statement. (Again, this final statement would be democratic analog to the Final Regulatory Impact Analysis.) This document should describe the public engagement strategy blueprint that was originally created, any changes that were made during the rulemaking process, what input was received through the agencies’ engagement mechanisms, and what impact they had on the proposed and final rules, if any. Again, agencies should be forthright in identifying the input that did not impact the rule and briefly explaining why.
Lastly, after the final rule has been published, agencies should dedicate resources and time to reflecting on lessons learned from the implementation of public engagement strategy blueprint. They should be prepared to incorporate these lessons into the design and implementation of future public engagement strategy blueprints. This will lead to implementation of Public Participation Planning that is more effective, less expensive, and quicker. In addition, agencies will also need to be prepared to track the implementation of any public participation mechanisms related to implementation incorporated into the final rule design. As noted above, these mechanisms might relate to compliance monitoring and enforcement, retrospective review, and grants of compliance relief.
Advantages of Public Participation Planning
Public Participation Planning stands in stark contrast to the largely one-sized-fits-all approach to public engagement – basic notice-and-comment procedures with occasional public hearings – that characterize the current rulemaking process. As noted above, essentially no deliberation goes into the creation of this engagement strategy – it is effectively reflexive – nor does it recognize, much less attempt to address, realistic concerns that important categories of stakeholders may not be accounted for in its strategy or that such incomplete input risks aggravating the very power disparities and social inequities that gave rise to the problem that the rule is meant to address in the first place.
In addition, successful implementation of Public Participation Planning will promote better regulatory democracy in the following ways. First, it will provide agencies with a mechanism for systematically identifying all the relevant stakeholders for a given policy, particularly members of communities who might otherwise be systematically excluded from such decision-making processes by structural or other barriers. Second, it will ensure that input is elicited from these stakeholders consistent with the sequential logic of the rulemaking process, providing agency decision-makers with the information they need when it is most useful.
Third, it will empower agencies to tailor their outreach and engagement strategies to the unique policymaking context implicated by the rule under development. Fourth, it will enable agencies to use public engagement to surface and account for any underlying power disparities that contribute to the policy problems a rule is meant to address, leading to more effective and durable policies. Fifth, it will highlight productive disagreement among stakeholders rather than engage in a quixotic pursuit of consensus – that is, it seeks to move regulatory democracy in a more agonistic direction. This is essential to recalibrate public engagement so that it is more attentive to power disparities and to avoid being a source of excessive delay in the policy development process.
How Other Federal Institutions Can Support the Successful Implementation of Public Participation Planning
The White House
With the advent of presidential administration under Reagan, the White House has played an increasingly active role in coordinating and steering the actions of the administrative state. The White House would thus be well-positioned to support the effective implementation of public participation planning. Indeed, as noted above, the Biden administration took some important initial steps on strengthening public participation in the rulemaking process as part of its broader Modernizing Regulatory Review initiative.
A logical place to start would be for staff at the White House Office of Management and Budget (OMB) to produce a comprehensive list of public outreach and engagement tactics for agencies to use to inform their own public engagement strategy blueprints. They could create this list by surveying the relevant academic literature, reviewing agencies’ past experiments with innovative approaches, and even looking at examples offered by peer democratic states abroad.
To support ongoing agency learning, OMB could also convene a standing working group composed of representatives from the public engagement offices at the various agencies. This working group could provide a forum in which these offices regularly share their best practices and lessons learned. Just as significantly, by signaling that public engagement is a priority of administration leadership, the working group would also by its mere presence help to reinforce a broader ethic and commitment to democratic inclusiveness across the administrative state.
Inviting OMB support in the implementation of Public Participation Planning is not without risk, given its historic role of interfering with and unduly politicizing the rulemaking process. It would certainly be preferable if Congress created a new standalone office outside of the White House that is explicitly charged with these tasks, as suggested below. But short of that, OMB is institutionally best positioned to play this role – provided that it does so in a strictly auxiliary fashion, leaving individual agencies the ultimate discretion on how to implement Public Participation Planning. In addition, carrying out such an auxiliary role would be a far better use of OMB’s resources than its current practice of superintending agency decisionmaking through the centralized regulatory review process.
The implementation of Public Participation Planning would also benefit greatly from having staff with different kinds of skillsets and life experiences. For instance, staff with backgrounds in social work or community organizing and specialized training in sociology might be particularly valuable. The White House Office of Personnel Management (OPM), the main human resources agency for the administrative state, could be instrumental in helping agencies to identify and hire such individuals. OPM could also help make necessary revisions to hiring standards and practices to make it easier and quicker to bring them on board.
Congress
Agencies have adequate legal authority to undertake Public Participation Planning. Still, Congress can ensure that even future administrations that might be hostile to the goals of regulatory democracy will implement this reform, even if reluctantly, by codifying this procedure into law through an amendment to the APA.
Similarly, implementation of Public Participation Planning likely would not require a significant commitment of agency resources – especially, if agencies are able to redirect resources to it from other rulemaking requirements, such as cost-benefit analysis or the myriad energy-related analyses that agencies must conduct pursuant to various executive orders. Ideally, Public Participation Planning will also reduce the incidence of legal challenges against final rules, which would promise to save on direct litigation costs. With these reduced litigation risks, agencies may also find that they are no longer compelled to “bulletproof” their rules through elaborate rulemaking records and gargantuan preambles to their final rules. This resulting streamlining effect of Public Participation Planning could also yield significant cost savings for agencies over the long run.
Nevertheless, Congress should still commit adequate appropriations for agencies to launch this reform, especially while they are still overcoming the incremental additional costs required to move through the early stages of the learning curve. With increased experience and specialization, agencies will likely be able to implement Public Participation Planning in an increasingly cost-effective manner.
Congress can take other steps to affirmatively support Public Participation Planning. For instance, they can authorize and fully fund a new institution that affirmatively supports public participation in the notice-and-comment process. The EXPERTS Act, a comprehensive progressive regulatory reform bill now pending in Congress, offers one potential model. Specifically, it would create something called the Office of the Public Advocate, which would be charged with this responsibility.
In addition, Congress can tap the Administrative Conference of the United States (ACUS) – which is effectively the federal government’s in-house “think tank” on administrative law – to study existing administrative law doctrines that might present a barrier to effective implementation of Public Participation Planning (to the extent that the doctrines arise from statutory, as opposed to constitutional, law). Such doctrines might include Vermont Yankee’s bar on judicially created administrative procedures (the codification of public participation planning, recommended above, would obviously address this), Loper Bright (which gives the judiciary, instead of agencies, the primary responsibility in interpreting agencies’ statutory authority), and the “logical outgrowth” test (which constrains how significantly a final rule’s substance can deviate from what’s contained in the proposal). ACUS could develop recommendations for how agencies and reviewing courts can avoid running afoul of these doctrines or propose legislative fixes for Congress to adopt.
Lastly, it goes without saying that Public Participation Planning would benefit immeasurably from having a functional Congress in place. According to the common conservative myth, an empowered Congress is necessary to restrain the administrative state. Just the opposite is true, however. By recommitting to doing the people’s business and passing public interest legislation again, Congress would provide agencies with fresh opportunities to put Public Participation Planning into action with up-to-date legal authority to tackle the new and emergent problems that pose a great threat of harm to structurally marginalized populations.
The Judiciary
The most obvious way that the judiciary could support the implementation of Public Participation Planning is by devising new judicial review doctrines that reward rulemakings with exceptional democratic pedigrees with enhanced levels of deference. This, of course, would require conservative judges to apply an even hand to all regulations challenged on judicial review before them. That means they would have to deploy enhanced deference consistently and in the service of promoting regulatory democracy, rather than wield it as a weapon to justify striking down rules they oppose on policy grounds. Under present circumstances, one might be forgiven for doubting this will take place. Yet, since we cannot avoid this institution either, we must still do the best we can with our politicized judiciary until such time as we are able to accomplish significant judicial reform – a topic beyond the scope of this paper.
Such enhanced deference might have a role to play in assessing the statutory authority for the agency’s rule, even under the new Loper Bright review framework. For instance, reviewing courts might modify the application of Skidmore to apply a special degree of “respect” for agency interpretations that rest on public input received during the rulemaking process.
More likely, though, the influential weight of public input would be greatest during the review of agency policy decisions under the arbitrary-and-capricious standard. In theory, courts already employ a “super deference” for agencies’ determinations based upon science and other forms of technocratic expertise, though courts rarely follow this approach in practice. Courts could easily create (and actually follow) an analogous super-deference doctrine for agency determinations based on “democratic expertise.” The inclusion of the Final Public Participation Plan Statements in the rulemaking record, as outlined above, would provide the essential informational foundation for the application of such a doctrine. In developing this doctrine, courts would have to account for applicable doctrinal constraints, including, most notably, Vermont Yankee.
Short of that, though, much of the analysis involved in assessing Final Public Participation Plan Statements would fit comfortably within the “hard look review” that courts already perform as part of the arbitrary-and-capricious standard of review. In other words, implementation of Public Participation Planning would not create any insurmountable barriers for courts conducting judicial review pursuant to the APA.
Courts have long recognized that the APA’s arbitrary-and-capricious judicial review standard implies a general duty for agencies to build a sufficiently complete rulemaking record to enable such review. Contained within this broader duty is a more specific responsibility to have procedures or mechanisms in place for ensuring that the information before the agency meets some minimal level of quality. While this responsibility might traditionally be thought of as applying to more technocratic inputs, there is no reason why it should apply equally to the unique on-the-ground expertise of individuals and community-based organizations. Similarly, this general duty also implies a more specific requirement that agencies ensure that the scope of information available to them be sufficiently broad to permit for evidence-based, reasoned decision-making required by arbitrary-and-capricious review. Again, this concern should apply equally to all forms of expertise, not just those regarded as technical or scientific in nature.
Importantly, this same judicial review standard would also guard against attempts by any president who is hostile to regulatory democracy to implement Public Participation Planning with insufficient rigor. Just as the Trump administration is now seeking to bypass the notice-and-comment process altogether, a future administration might reduce this process to a mere check-the-box exercise or conduct woefully inadequate outreach and consideration of input. The Final Public Participation Plan Statements would afford a reviewing court with a basis for applying the arbitrary-and-capricious standard to the agency’s public engagement efforts and, ultimately, to remand an agency rulemaking to correct this aspect of the record where any flaws or gaps are identified.
The notion of courts policing Public Participation Planning raises a separate concern that this aspect of arbitrary-and-capricious review could be abused by conservative activist judges who are opposed to climate policy or other aspects of the progressive policy agenda. The Supreme Court’s recent extreme application of the arbitrary-and-capricious standard in Ohio v. EPA confirms that this is not an idle concern. Still, it seems clear that activist judges will find plenty of opportunities for abusing arbitrary-and-capricious review even in the absence of Public Participation Planning. On balance, then, the benefits of this reform would still seem to outweigh these risks.
Public Participation Planning as Part of a Broader Agenda to Increase Administrative Effectiveness
The past year has seen the Abundance Liberal movement spark a robust debate within the broader liberal community over the appropriate role of legal procedure in our governing institutions. Under the circumstances, then, it may seem like an unusually inopportune time to champion something like Public Participation Planning – an essentially proceduralist reform. As explained below, though, this reform strives to take seriously Abundance’s critiques and is consciously predicated on the recognized imperative to strike an appropriate balance between, on the one hand, public engagement and, on the other, effective, responsive administrative action that delivers concrete results.
The Abundance Liberal movement, as best captured in the recent book by its most prominent advocates Ezra Klein and Derek Thompson, argues that the Left’s reflexive embrace of proceduralism and litigation is an antiquated relic from a bygone era and is already becoming a political liability. That is because many of the progressive movement’s policy priorities – from addressing the climate crisis to promoting affordable housing – requires quick policy implementation, a goal that is ultimately defeated by excessive proceduralism. Instead, the book argues, the Left should dispense with most procedures as a mechanism for legitimizing government action and instead let the popular results of those actions (e.g., affordable housing, cheap clean energy, etc.) serve that legitimizing function after the fact.
Significantly, Klein and Thompson’s book singles out public participation in the policymaking process as emblematic of the broader problem they are trying to solve. Indeed, their specific critiques of notice-and-comment procedures largely tracks with those that have motivated the proposal Public Participation Planning, as detailed above. In particular, they correctly identify these procedures as excluding structurally marginalized communities and reinforcing broader power disparities in our society.
Where the Public Participation Planning proposal departs from Abundance adherents such as Klein and Thompson is its core claim that the problem is not procedure per se, but power disparities. More specifically, it posits that the dysfunctional procedures that cause excessive and unnecessary delays in policy implementation are better understood as a symptom of the deeper problem of power disparities in our society. After all, even if we were to remove all existing procedural requirements – that is, to take Abundance to its logical extreme – it is by no means clear that we would see more expeditious policy implementation, particularly where those policies go against the preferences of entrenched elites. Such interests would simply find other, non-procedural mechanisms for blocking policies they opposed.
Given that we are unlikely to eliminate the structural sources of power disparities in our society any time soon, it is worth exploring more practicable near-term mechanisms for alleviating the worst consequences of those power disparities. What Abundance potentially misses with its black-and-white diagnosis of the procedure problem is that procedure actually holds a lot of promise for accomplishing this goal. After all, if procedures can aggravate power disparities, as Abundance Liberals would freely stipulate, then it also follows that well-designed procedures can do the opposite as well. Legal scholar Nicholas Bagley, who has provided part of the intellectual foundation for Abundance, highlighted this intrinsic feature of procedure when he wrote: “government action — whether it involves dispensing public benefits or regulating private conduct — allocates resources, risk, and power within the United States.”
To put Bagley’s point differently, procedure can never be neutral in its effects on underlying power dynamics; it will tend to cut in the favor of one set of stakeholders or another. Taking this reality seriously means that policymakers have a lot of tools at their disposal to shape and reshape those power dynamics in more productive ways through carefully designed procedures. More to the point, Abundance overlooks the tantalizing possibility that, by effectively redistributing power within the administrative state, well-designed procedures can actually expedite policy implementation – or at least add value, such as improving the quality of decision-making, without causing new or undue delay. This is precisely the project that Public Participation Planning sets out.
As described above, Public Participation Planning illustrates how procedures might be designed with the goal of affirmatively redistributing power from entrenched interests committed to maintaining a suboptimal, unjust status quo to members of structurally marginalized communities that are both underrepresented in our political processes and disproportionately burdened by the harms that arise from status quo economic, social, and political arrangements. Specifically, it seeks to carve out new spaces in the rulemaking process for bringing in the unique expertise of historically underrepresented populations at precisely the points in that process when their expertise is most germane. At the same time, it contemplates placing reasonable constraints on the participation opportunities available to entrenched powers such that their input is channeled to maximize its utility for agency decision-making. This would also have the additional benefit of preventing these interests from abusing their resource advantages to overwhelm agency decision-makers with extraneous information for the purpose of causing unnecessary delay.
Public Participation Planning’s commitment to transparency and ongoing communication with stakeholders plays an important contributing role in this power shifting dynamic. For structurally marginalized stakeholders, it will be essential for them to know concretely how their input is materially shaping the policy and factual determinations that undergird agency decision-makers. Otherwise, it will be entirely rational for members of these communities to assume that their participation is little more than a check-the-box exercise, rather than a genuine effort at promoting regulatory democracy. Understandably, members of such communities will still harbor some degree of skepticism even despite agencies’ good faith commitment to transparency and communication. Trust will take time to build, but it can be built. Somewhat counterintuitively, one way that agencies can do this is by always honestly explaining to these stakeholders when their input did not substantively influence a particular decision – that is, by delivering the bad news as well as the good.
Equally as important, Public Participation Planning demonstrates how these kinds of investments in well-designed procedures can actually pay off in terms of more expeditious policy implementation. In other words, it helps to refute the commonly held belief – one clearly embraced by the Abundance Liberal movement – that public engagement and expeditious policy action are fundamentally at odds. Instead, these tradeoffs can effectively be mitigated when such public engagement procedures are designed to correct power disparities that are implicated by a given rulemaking.
The key to accomplishing this seeming procedural alchemy is by addressing the primary driver of rulemaking delay: litigation over final rules. (To be sure, Public Participation Planning would also endorse Abundance’s more general call to clear out unnecessary procedural requirements that have accreted over the years, such as cost-benefit analysis, the Regulatory Flexibility Act, and the Unfunded Mandates Reform Act, to name a few. Doing so would be entirely consistent with its analytical framework. These procedures tend to aggravate power disparities – often by design – and thus contribute to delays in the rulemaking process.) Litigation is time-consuming in its own right, often taking several years to reach a definitive conclusion. During this time, agencies are increasingly subject to court orders barring them from implementing the rule’s provisions.
In addition, as noted above, the near certainty of legal challenges creates strong incentives for agencies to “bulletproof” their rules to reduce the chances they will be struck down on judicial review. While some degree of deliberative and evidence-based rigor underlying rules is desirable, of course, the perverse result of this incentive structure is that agencies go far beyond what the law reasonably requires for substantiating rules, an increasingly resource-intensive undertaking that significantly delays the completion of new actions.
As noted above, many of Public Participation Planning’s distinguishing features are designed to reduce the incidence of litigation. Particularly significant in this regard is its emphasis on early engagement. Not only does engagement provide agencies with better and more timely input; it also has the additional benefit of promoting greater buy-in from stakeholders, which in turn defuses litigation risk down the road. In other words, frontloading public engagement seems to shorten the length of the rulemaking process even if agencies are technically conducting a greater number of procedural steps overall.
Several empirical studies of the implementation of the National Environmental Policy Act’s (NEPA) analytical requirements appear to confirm this effect, finding that NEPA processes that involve early public engagement are statistically shorter than those that do not. Reduced litigation and, by extension, the reduced incentives for bulletproofing NEPA analyses seem to explain this discrepancy. (Not incidentally, extensive NEPA-related delays are also the subject of extensive criticism by the Abundance movement) It is reasonable to expect that this dynamic would translate to the functionally similar rulemaking context as well.
Also important, Public Participation Planning seeks to institutionalize a more agonistic orientation into the regulatory process. As explained above, agonism seeks to avoid the pursuit of near-universal consensus over policy outcomes, which is typically impossible to achieve in practice anyway, and instead create conditions for productive disagreement. Similar to early engagement, one of the desired effects of administrative agonism is to reduce litigation risk and the perverse effects such risks create. Instead, final regulations would be treated as more contingent and subject to realistic mechanisms for ongoing revision and refinement or to various forms of implementation flexibility. In this way, agonism enabled by Public Participation Planning would attempt to lower the stakes on final rules, such that stakeholder focus could be gradually shifted from post-finalization litigation to implementation where it can be put to more productive use.
To be sure, achieving the full agonistic potential of Public Participation Planning would require other legislative changes. For instance, Congress could amend agencies’ authorizing statutes to give them greater authority to deploy back-end implementation adjustments and flexibilities, such as waivers, exemptions, and compliance extension deadlines (all subject to vigorous public participation mechanisms, as noted above). For agencies that already enjoy these kinds of authorities, implementing this general approach could serve as a valuable proof-of-concept demonstration that could help catalyze this legislative action in the near future.
Another option would be for Congress to incorporate more rigorous schedules for reviewing and updating regulations into their statutory design, such as those that exist for appliance energy efficiency standards. To ensure that these reviews are carried out expeditiously, Congress could also experiment with different kinds of “hammer provisions” that would kick in automatically – setting default regulations or standards, for instance – if an agency is unable to work with relevant stakeholders to adopt the update according to the statutory schedule. Ultimately, the goal of these reforms would be to rebuild a rulemaking process in which stakeholder contestation is increasingly shifted to the implementation phase, where dynamism and flexibility can be permitted to flourish.
Even better, the judiciary could further reinforce this litigation-dampening effect of Public Participation Planning by adopting a strong deference doctrine based upon a rule’s democratic pedigree, as noted above. Such doctrines have the potential to substantially alter the calculus for stakeholders who are contemplating a challenge against the rule. If a rule’s legal and policy bases are strongly supported by public input such that it is likely to earn some measure of enhanced judicial deference, stakeholders may refrain from undertaking the expense of bringing a legal challenge, given the reduced likelihood that the challenge would succeed.
Even with all this careful attention to the design and implementation of Public Participation Planning, ongoing vigilance from policymakers will still be required to ensure that its performance delivers on its promises of better-informed decision-making and more expeditious policymaking. Regrettably, while effective use of this tool can help to alleviate the consequences of power disparities, those underlying power disparities will still remain in place, absent other more radical power interventions. The practical upshot is that over time entrenched interests may learn to “capture” parts of the Public Participation Planning program and deploy them to advance their narrow policy preferences at the expense of the broader public. (Both the APA notice-and-comment process and NEPA’s analytical requirements appear to illustrate this general dynamic.) This is why Public Participation Planning demands that agencies continually reevaluate and update their public engagement and outreach actions. Ideally, though, the general Public Participation Planning framework will remain resistant to such capture risks, even if more specific tactics and strategies for instituting that framework do not.
Conclusion
An effective response to the climate crisis faces numerous obstacles. One of these is a regulatory system that inadvertently reinforces underlying power disparities that help to maintain status quo conditions on how we obtain and use energy. A better approach to the practice of regulatory democracy – one that takes seriously and affirmatively addresses power disparities – will be essential for overcoming this obstacle.
This paper proposes a comprehensive reform to how agencies engage the public during the rulemaking process called Public Participation Planning. The distinguishing feature of this reform program is that it would require agencies to develop tailored public engagement strategy blueprints for each of their planned rulemakings. The purpose of these blueprints is to enable agencies to draw on a variety of engagement tactics that are calculated to build a reasonably comprehensive record of the stakeholders’ views on the rule. Importantly, the blueprint would also enable agencies to determine when best to engage different stakeholders throughout the various stages of the rulemaking process – a concept known as sequential participation. This reform would also require agencies to assemble Initial and Final Public Participation Planning Statements, which document the implementation of the blueprints and the ultimate impact that the resulting public input had on the substance of the rule. These Statements would be made part of the rulemaking record where they could be considered by judges, if necessary, during judicial review of the rule.
Public Participation Planning offers two major advantages. First, and most directly, by improving the quality of public input, it will lead to better decision-making, and thus better policy outcomes. Second, its implementation is likely to expedite rulemakings in many instances. Public Participation Planning would help to accomplish this outcome by using procedures to alleviate power disparities among relevant stakeholders. Properly understood, such power disparities appear to be the root cause of delayed rulemakings in the past. Taking these two advantages together, Public Participation Planning would offer a crucial piece in the larger puzzle of addressing the climate crisis effectively and with the urgency needed to avoid its worst consequences.
The Civic Research Agenda
The Civic Research Agenda is a culmination of several years of study, partnerships, and intelligence gathering that is the first comprehensive reporting on the priority research needs of American cities and counties. It considers the demand and supply of research: what are the research needs of local governments, and how can research outputs improve to “supply” or provide answers to better serve that audience?
The priority research needs for U.S. local governments are the following:
- Housing
- Economic Development
- Human Services
- Climate and Energy
- Transportation
- Community Engagement
- Service Delivery
Beyond any specific policy domain, local governments expressed the desire for support from the research community in three overarching areas: 1) evaluation; how can the research community measure and provide evidence that a policy intervention has achieved desired (or negative) impacts; 2) efficiency; how can the research community help local governments do more with less; and 3) data generation; how can the research community create and provide access to useful data that do not currently exist.
This report also focuses on the ecosystem itself; what are the current perceptions, barriers, and recommendations that can inform and improve how local governments and universities work together? Findings show that issues include:
- Perceptions: Incorrect perceptions are stalling trust between these institutions. University faculty and staff fear being seen as too intellectual and disconnected. Local government staff fear being seen as not intellectual enough.
- Barriers: The importance of relationship building and lack of networking: there are clear structural discovery systems on either side. Both are operating informally, and often, only by individual relationships. This report also provides evidence that the research demand/supply ecosystem is deeply relationship-dependent, creating high barriers for newcomers, inequity of access, and bottlenecks around well-connected individuals.
- Recommendations: The best way for research outcomes and publications to become actionable are:
- Make the research and findings as specific to a jurisdiction as possible
- Reduce paywalls
- Provide executive summaries with no technical jargon
- Provide peer-city or peer-county comparative analysis
- Go beyond observation and make recommendations
Finally, this report provides specific recommendations for local governments and universities to improve and grow the research-to-impact pipeline for one simple purpose: make research actionable, understandable, and accessible to communities across the country.
The singular recommendation that can strengthen the research-to-impact pipeline is this: research should have an audience that lives outside of the peer-to-peer review system.
CELS Playbook: Clean Electricity for Local and State Governments
State and Local Actions to Make Government Work for People, Reduce Utility Bills, and Deploy Clean Energy
Elected leaders across the country are staring down interlocking crises. Families and businesses are struggling to pay skyrocketing utility bills. Large new demands are straining the grid and overtaking the buildout of new power plants. And the public’s faith in government has hit new lows. We need a new playbook to solve these problems and make the government responsive to peoples’ needs.
What’s going wrong?
Utility bills are rising rapidly for households and businesses due to an administrative state ill-equipped to protect customers from costs and risks. The cost of power supply is increasing due to growing demand, long timelines to build new cheap clean energy, and volatile natural gas prices. Utilities are spending more money on the transmission and distribution grid for both maintenance and recovery from wildfires and other disasters. Today’s regulatory construct allows utilities to drive spending decisions and pass on all these costs to customers, and regulators are under-resourced and unwilling to find alternative solutions.
Meanwhile, we are not building clean energy nor upgrading the grid fast enough to meet demand growth and address climate change. And this problem will get worse as power-hungry data centers connect to the grid and electrification of buildings, vehicles, and factories adds additional electricity demand.
The old climate policy playbook is not equipped for this moment. While it has driven significant deployment of low-cost clean energy, it was not designed to address non-financial obstacles to building projects and upgrading the grid nor to fully mobilize the suite of finance tools needed for the energy transition, nor to demonstrate that the government can make peoples’ lives better, now and long term.
Where do we go from here?
Policymakers and advocates need an expanded playbook. One that addresses the full set of barriers impeding financing and construction of clean energy and grid modernization projects. One that targets the root causes of high energy costs. One that reworks the administrative state to make government work for the people.
FAS, with the help of partner organizations spanning ideology and function, launched the Center for Regulatory to Ingenuity to build a vision for a government that is agile and responsive and delivers affordable energy, abundant housing, and safe transportation for all Americans.
As part of this work, we have developed an updated set of policies and actions for state and local leaders to meet this moment. We started by identifying the barriers to deployment and the flaws in the old playbook, published in our report Barriers to Building. Now we are developing the “plays” in a new playbook—tangible actions that state and local leaders can take now to make near-term progress and pilot new solutions. These plays will live on this landing page, which we will continue to update with additional actions.
This playbook is not a laundry list of policies but rather a cohesive strategy to achieve two goals: (1) deploy the clean energy and grid upgrades necessary to make energy affordable and combat climate change and (2) create governments that tangibly improve peoples’ lives.1
Contents (click to jump to a section)
- Main Character Energy: Make Regulators Main Characters in Planning and Ratemaking
- Improve State Government Responsiveness to Clean Energy Projects
- Build Administrative Capacity to Plan for an Affordable & Reliable Grid
- Fix Broken Incentives to Expand Distributed Energy Resources
- Wield Creative Finance Tools to Drive Investment and Reduce Capital Costs
Main Character Energy: Make Regulators Main Characters in Planning and Ratemaking
Utilities and their regulators are responsible for major decisions about what infrastructure we build and how much people pay for energy. Utilities—which can be owned by investors, the public (e.g., municipal utilities), or members (i.e., electric cooperatives)—conduct detailed analysis and provide proposals on planning and ratemaking to their regulators. The set of solutions below focuses on investor-owned utilities, who are incentivized to prioritize projects that maximize the returns for their shareholders. As a result, they underutilize solutions that could save customers money but do not earn companies a profit, like rooftop solar or technology-or maintenance-based upgrades to existing transmission lines.
In a well-functioning system, regulators—whether Public Utility Commissions (PUCs) or locally elected officials—would rigorously interrogate utility analyses and direct the utilities to shape or revise their proposals to maximize benefits to the public at lowest public cost. This lens is needed to ensure that utilities are spending money wisely in the public interest and prevent unnecessary bill increases from overspending on the wrong solutions. Active regulators are also needed to incorporate long-term considerations in planning and ensure consideration of strategies that provide long-term benefits. However, regulators are often not well-equipped or politically willing to conduct detailed analysis and push back on utility proposals. Other intervenors, like consumer advocates and environmental organizations, are outspent by the utilities, who can recover the costs of their analysis and interventions through customer bills in most states.
The result is a reactive, short-term-focused administrative state that leaves the public frustrated. Regular people are frustrated with skyrocketing bills, clean energy companies are frustrated with slow processes and broken incentives, and both are frustrated with the government’s ability to solve big problems. The administrative state itself—the officials and staff who make up regulatory body and state and local governments—are also frustrated with their perpetually reactive role and with limited say in outcomes.
We should not accept the status-quo regulatory process as a given. As representatives of the public, regulators should have both the ability and motivation to actively drive toward an abundance of cheap clean energy, affordable bills, and a modernized reliable grid. Achieving this vision requires the right personnel, clear direction and support from governors, and adequate analytical capacity.
Solutions
I. Direct Regulators to Use All Tools to Lower Energy Bills and Deploy Clean Energy (Governors and Legislatures)
When regulators take a backseat and let utilities drive, they narrow the toolkit of resources that can help meet demand and as a result leave savings on the table. Regulators with a mandate to prioritize affordability and clean energy buildout can reduce bills by both better scrutinizing utility plans and taking a more active role in enabling clean energy deployment. This includes finding creative tools to get more out of the grid through distributed energy resources, alternative transmission technologies, and flexible sources of demand like electric vehicle charging and factories with electric appliances.
Governors can direct PUCs to audit utility investments to find opportunities for savings, re-evaluate utility business models and incentive structures, consider distributed energy resources and alternative transmission technologies in planning, and consider climate impacts in planning and ratemaking decisions.
Legislatures can set statutory requirements for PUCs to consider these opportunities and expand their mandates to include clean energy goals and highest net benefit criteria.
Legislators can require PUCs to find savings across the gas and electric systems, including by using beneficial electrification to reduce costs.
In 2021, the Maine legislature directed the PUC to consider emissions reduction targets and equity impacts in regulatory decisions.
Oregon Governor Brown directed its PUC to integrate the state’s climate pollution reduction goals and promote equity by prioritizing vulnerable populations and affected communities.
New Jersey Governor Sherrill directed the PUC to review utility business models and assess whether they are aligned with cost reductions for customers.
In 2024, Minnesota S.F.4942 mandated that the PUC establish standards for sharing utility costs for system upgrades, ensuring fair cost-sharing and advancing state renewable and carbonfree energy goals along with provisions for energy conservation programs for low-income households.
II. Even the Playing Field by Providing More Resources to the PUC and Consumer Intervenors and Increasing Data Transparency (Governors, Legislatures, and PUCs)
Electricity rates are determined by proceedings called rate cases, in which utilities submit proposals and justifications to regulators, other intervenors (such as consumer advocates, environmental organizations, and state and local elected officials) submit testimony, and the regulators hold hearings and make a decision. Most rate cases end in settlement agreements between the utilities and other intervenors, facilitated by the PUC. Utilities drive this process—they file initial proposals and have more information about their system than other participants. Well-resourced PUCs and public interest intervenors are important to interrogate utility proposals and ensure that settlement agreements are a good deal for regular people.
In order to take on more responsibility in grid planning and utility oversight, PUCs need additional staff and analytical capacity. For example, a legislative commission in Texas found that the PUC needs more staff and resources to independently analyze utility sector data and provide sufficient oversight to ensure reliability. Funding for staff and analysis has a great return on investment—state leaders can save customers money and get better outcomes for a relatively small price.
Moreover, consumer advocate intervenors are typically underfunded compared to utilities and so cannot compete with utility proposals. This disparity in funding places consumer advocates in a position of exclusively reacting to utility requests, rather than having the bandwidth to interrogate existing system inequities or to develop potential innovative solutions to address ratepayer needs. Utilities also determine the pacing of their rate case applications, which can put consumer advocates even more on their heels. For example, in a 2019 rate case in Colorado, Xcel Energy brought 21 witnesses, while only a few consumer advocate intervenors testified.
Utilities in most states can recover the full costs of analysis and intervention legal fees from customers, giving the utilities significant resources to drive the process. States can prohibit this practice, directly reducing bills for customers and reducing utility influence over the process.
In addition to resources for PUCs and intervenors, data transparency can help even the playing field. Utility data are often difficult to access, embedded in filings that often run thousands of pages, and not standardized. This lack of data transparency makes it difficult for PUCs and consumer advocates to track utility spending and effectively intervene in rate cases.
Legislatures can provide additional funding for PUCs to hire additional staff and conduct independent analysis.
Legislatures and PUCs can prohibit utilities from recovering the costs of political activities from customers and limit the amount of legal fees that are recoverable from customers.
Legislatures can establish mechanisms to ensure that low‑income, consumer, and environmental justice advocates can participate meaningfully in PUC proceedings. Several U.S. states have implemented intervenor compensation programs or similar initiatives that reimburse reasonable costs for nonprofit organizations and community groups engaged in utility regulatory processes.
PUCs can charge utilities to fund independent analysis of utility proposals on behalf of customers.
PUCs can assess their processes with an eye toward reducing participation barriers for non-traditional docket participants, such as groups representing low-income or environmental justice communities.
PUCs can standardize reporting on utility costs and increase data transparency both during and in between rate cases.
Governors can direct agencies to conduct analysis to inform PUC proceedings and hire technical talent to engage with the PUC. Legislators can authorize and fund state agencies to conduct independent, proactive analysis to inform PUC proceedings, with opportunities for public input on the analysis.
California passed AB 1167 in 2025 that put an end to the use of ratepayer funds for political lobbying and strengthening enforcement against investor-owned utilities (IOUs) that illegally use ratepayer funds.
A 2023 Colorado law prohibited utilities from charging customers for lobbying expenses, political spending, trade association dues, and other similar activities.
In Illinois, a 2021 law expanded the Consumer Intervenor Compensation Fund to compensate consumer interest intervenors in planning and rate cases.
The Oregon PUC provides both Intervenor Funding and a dedicated Justice Funding program, supporting groups representing environmental justice communities and low‑income customers, with clearly defined funding caps for eligible participants.
Improve State Government Responsiveness to Clean Energy Projects
To build a clean energy project, developers must navigate a complex mix of state environmental permits, local and/or state siting approvals, and utility and grid operator interconnection processes. While federal permitting reform receives the most attention, most clean energy projects do not require federal approval and often get stuck due to state processes and local restrictions. For example, 73 percent of wind and solar projects that faced opposition in the 2010s were contested only at the state and local level, not federally.
In many places, these federal, state and local processes are not working for anybody—the clean energy developers who seek to build projects, the local communities in need of jobs and economic development, and the families and businesses struggling with rising utility bills.
What is going wrong? State agencies are often stretched thin, and outdated processes make it difficult to respond quickly to new projects. Most states lack a single designated authority at the state level who can oversee and enforce timelines. Permitting approvals often involve regional offices who take different approaches, increasing complexity and uncertainty for developers applying for permits. Inconsistent decision timelines increase risk for projects, raising costs.
In most states, local governments have control over siting, and each municipality has different processes and requirements, adding complexity for developers. Most states also lack a state authority that can ensure local governments do not unreasonably block or delay projects. About 20 percent of U.S. counties now have formal restrictions on clean energy projects.
State and local leaders can play a critical role in addressing these challenges. Governors can target agency capacity where it is most needed and legislatures can complement these efforts by funding permitting offices, creating statutory timelines to standardize reviews, and giving them enough decision-making power to actually meet those deadlines. State governments can leverage their capacity to evaluate costs, risks, and benefits—across timescales and geographies—of projects to inform decisions. Local governments can similarly improve and standardize their processes and support state implementation.
This brief focuses on a subset of solutions that help states be more responsive to clean energy projects. These solutions expand and improve government capacity to build clean power faster, lower utility bills, and demonstrate that the government can be an active and effective partner in solving problems.
Solutions
I. Increase Permitting Certainty and Consistency (Governors and Legislatures)
Uncertainty and inconsistency in permitting processes increases costs and delays projects. Timelines to hear back from an agency might vary considerably from one project to another, which adds uncertainty to project timelines. Some review processes are run by regional offices which may take different approaches to project evaluation (e.g., using different assumptions for modeling the impact of a project) and mitigation requirements. As a result, the same developer might go through a bespoke process for two very similar projects in different parts of the same state.
In addition, the scopes and goals of state environmental laws are often outdated. For example, environmental review often does not consider system-wide effects or second-order emissions impacts. As a result, environmental review often gets stuck on project-by-project analysis and lacks an overarching vision for the system.
States can address these issues minimal to no cost and without sacrificing the quality of environmental review by increasing certainty and consistency of permitting processes, centralizing capacity to run processes across agencies, and adjusting the goals and scope of environmental statute to include system-wide impacts and overarching climate goals.
Governors can issue clear guidance and standard operating procedures for analysis of impacts for different clean energy project types required under different laws (e.g., state clean water or environmental review statutes) and set timelines that agencies must follow for review and decisions.
Governors can set clear permitting goals for agencies and empower the “machinery” expertise (e.g., staff engineers, lawyers, environmental specialists, etc.) to meet those goals. Governors can also develop programs to upskill existing staff to expand the technical capacity needed for permitting operations.
Governors can issue guidance on mitigation requirements that projects can take to shorten review.
Legislatures can also set statutory decision timelines and limitations on what impacts are considered to further increase certainty for projects. In some cases, legislatures may need to provide agencies with the authority and resources required to standardize mitigation requirements and speed up timelines.
Legislatures can re-align the goals and requirements of permitting statutes and environmental review to prioritize system-wide goals.
Legislatures and governors can exempt certain clean energy projects from state environmental permitting or create simplified permitting pathways for such projects.
Legislatures can mandate and support adoption of instant digital permitting for distributed energy resources located at homes and businesses, drastically reducing the cost to install rooftop solar and energy storage.
Governors and legislatures can consolidate agency authorities, reducing the number of process steps for developers and conflicts between agencies.
In Pennsylvania, solar developers must obtain a stormwater permit from the Department of Environmental Protection (DEP), a process that involves working with regional conservation districts on stormwater analysis and agreement on mitigation requirements. Developers have struggled with inconsistent approaches among conservation districts on modeling assumptions and mitigation requirements. In December 2025, DEP issued updated its Solar Panel Farms Frequently Asked Questions to clarify key analytical inputs for solar developers and will increase consistency in conservation district approaches.
The New York State Legislature passed the Accelerated Renewable Energy Growth and Community Benefit Act, mandating specific timelines for permit reviews, consolidating authority across agencies, and providing funding to support dedicated staff. This statutory framework has significantly reduced approval times for large-scale renewable projects.
Pennsylvania Governor Shapiro also directed state agencies to evaluate the timelines for each permitting process and set a specified maximum timeline for eligible projects by which applications will be processed.
In 2025, state legislatures in New Jersey, Texas, and Florida passed laws requiring local governments to adopt instant digital permitting processes for distributed resources.
Arizona House Bill 2003 allows utilities to replace conductors or structures on an existing transmission line without needing to apply for a Certificate of Environmental Compatibility (CEC) as long the line has a valid CEC (or has been grandfathered in) and all original conditions continue to be met.
In Minnesota, HF 4700 consolidated certain permitting responsibilities into a single law and shortened the timeline for state regulators to review and permit clean energy projects.
Texas passed SB 20 in 2005, establishing Competitive Renewable Energy Zones that smooth the process of developing and integrating renewable energy projects into the grid. This program has helped bring more than 18 GW of wind capacity online.
II. Increase Siting Certainty and Consistency (Governors, Legislatures, Local Leaders)
In most states, local governments handle siting of clean energy projects. Municipalities and counties within the same state may take wildly different approaches to siting, including different fee structures, setback requirements, public input requirements, approval processes and timelines, etc. This variability makes it harder to build clean energy projects without added benefit to communities. State leaders can improve certainty and consistency without sacrificing project quality or local benefits.
Legislatures can standardize processes across local governments, including by setting standard timelines for decisions, prohibiting excessive restrictions on clean energy projects and grid upgrades, and limiting the reasons for which a project can be denied approval.
Legislatures can move siting authority to the state level for large projects or projects that meet certain criteria (e.g., projects that create good jobs and economic benefits).
State lawmakers can create a streamlined, one-stop permitting process for distributed energy resources. This process would consolidate building, zoning, and environmental approvals. Such a framework reduces delays, lowers costs, and provides developers and homeowners with greater certainty.
Governors can develop model siting ordinances and encourage or incentivize local governments to adopt them.
Local governments can voluntarily work together to align processes.
Colorado developed a clean electricity code repository that included principles for smart local code design for clean electricity deployment.
Michigan HB 5120 financially incentivizes local governments to avoid overly restrictive ordinances and creates a state-led pathway for projects to by-pass overly restrictive local ordinances.
Arizona House Bill 2019 standardized timelines for municipalities to make permit decisions and took steps to ensure municipalities are responsive, including requirements on providing contact information and communicating with permit applicants.
A bill introduced in the Oregon legislature would exempt clean energy projects eligible for soon-to-expire federal tax credits from the state’s onerous state siting process.
Hawaii’s proposed SB588 creates a self-certification and standardized permitting system for behind-the-meter solar and storage projects.
Illinois HB 4412 set statewide renewable energy project siting standards that takes supremacy over unduly restrictive local decisions.
Indiana’s SB 411 set voluntary siting standards for wind and solar power projects. If communities meet those standards, they’re pre-cleared for project development.
New Mexico’s Renewable Energy Transmission Authority established an MOU with the Federal Permitting Improvement Steering Council to improve coordination and receive assistance on eligible permitting projects.
III. Create Dynamic Agencies that Can Respond to Project Needs (Governors and Legislatures)
States should build dynamic, flexible agencies that are responsive to evolving barriers to clean energy deployment. Government responsiveness is often thwarted by limited or outdated information, poor information sharing across agencies, and lack of centralized decision-making. Governors and state agency leads often do not know which specific processes are causing delays or uncertainty for projects. Where agencies have that information, it is often not shared among agencies or with the governor’s office without intervention. Agencies often do not have the mandate or the capacity to track and share this information.
A mandate and resources to collect detailed information on project barriers, and to share that information across agencies, allows states to be more responsive to clean energy industry needs. Using this information, states can ensure that limited agency capacity is targeted where it can have the greatest impact and helps agencies anticipate, rather than react to, common obstacles.
Governors can identify the specific government processes and other barriers that are holding up projects and increasing uncertainty through direct engagement with developers and formal processes like Requests for Informations. Governors can survey developers for specific information on the failure points in the process of getting projects built, across permitting, siting, interconnection, etc. Governors can then address the critical bottlenecks by adjusting the processes that are outdated or not serving the public or shifting capacity where processes must be maintained but are moving too slowly.
Governors can build centralized processes—for example by designating a senior official as state-wide lead for permitting coordination—that can quickly respond to bottlenecks and share information across agencies. Governors and legislatures can designate a single agency as the lead on permitting, siting, and project assistance and provide that agency with the authority to make decisions. For these steps to work, governors must provide the lead official or agency with the adequate decision-making authority and capacity to run the process.
Legislators can provide the resources for the above-referenced information collection and the authority for centralized permitting processes.
The Colorado Energy Office and Department of Natural Resources conducted an extensive survey of developers, local governments, community organizations, and other stakeholders to detail clean energy siting and permitting issues and develop a plan to address them.
Following passage of HR 1 in Congress, which rapidly phased out tax incentives for certain renewable energy technologies, Governors’ offices in states like North Carolina and Pennsylvania organized inter-agency dialogue with private sector stakeholders to identify and accelerate projects that could form cost-saving federal incentives before they expire.
The California Energy Commission runs a consolidated permitting process that centralizes staff capacity to manage permitting, with specified review timelines for clean energy projects that meet certain job quality and community benefits criteria.
Washington House Bill 1216 authorized the Department of Ecology to run a consolidated permitting process to collect all relevant information from developers and coordinate across agencies to speed up review. That bill also created an interagency Clean Energy Siting Council to improve how projects are sited in the state.
In New York, Governor Kathy Hochul directed state agencies to accelerate renewable project approvals by reviewing and reforming agency processes to reduce backlogs and coordinate efforts across departments, from environmental review to project eligibility screening, focusing agency attention on projects ready to leverage expiring federal tax incentives.
IV. Speed Up Interconnection Timelines (Governors, Legislatures, and Public Utility Commissions)
Long timelines for approvals to connect to the grid and to complete necessary transmission upgrades are one of the largest drivers of project cancellations and delays. States have little control over the interconnection process for projects connecting to the bulk transmission system, which is regulated by the Federal Energy Regulatory Commission. However, states can help projects use surplus interconnection processes, which enable faster approval for projects that are using excess interconnection capacity at existing generators. States also have control over the interconnection process to connect smaller projects to the distribution grid, which is run by utilities and regulated by the PUC.
Governors can direct state energy offices proactively identify and map surplus interconnection capacity and use state procurement authority, expedited permitting processes, and incentives to encourage clean energy development at those sites. This analysis should also include sites that may have excess interconnection capacity because projects (generators or new large load sources) fell through or shut down after triggering grid upgrades
Legislatures can require PUCs and utilities to consider using sites with underutilized transmission (e.g. sites of peaking gas-fired power plants that operate very infrequently) for clean energy development.
Governors and PUCs can identify zones where load is likely to increase (e.g., due to data center deployment, electrification of ports, and heavy-duty freight trucking) and fast-track distributed interconnection for projects in those areas.
PUCs can require that utilities develop rapid procurements for clean energy resources at sites with surplus interconnection.
In California, AB 1408 would have required state agencies, the state grid operator, and utilities to incorporate surplus interconnection capacity into long-term planning, but it was vetoed by Governor Newsom.
Virginia HB 1065, introduced in the 2026 legislative session, would task the State Corporation Commission with conducting a study of available surplus interconnection potential, and requires the state’s largest utility to procure new capacity via surplus interconnection.
In Indiana, HB 240, proposed in 2026, would require utilities in their integrated resources plan to analyze how much extra interconnection capacity they have and how they can use that existing capacity to bring more generation resources online. And, it would require such analysis to be done before permitting new power plants.
V. Provide Additional Funding for Siting and Permitting (Governors and Legislatures)
Clean energy permitting delays often stem from understaffed or under-resourced agencies struggling with outdated processes and technology. Legislatures can address this by increasing dedicated funding for permitting offices, enabling agencies to hire technical staff, invest in modern permitting platforms, and reduce backlogs. Allocating funding for targeted reforms can substantially shorten approval timelines. Potential examples of such include digital permit tracking systems, programmatic environmental reviews for common project types, and training specialized clean energy reviewers. States can start by digitizing permitting processes and making it easier for projects to submit the required information. States can then collect information from developers on the processes that are causing the most difficulty and provide targeted staffing and resources to address these bottlenecks.
Governors and legislatures can digitalize state permitting processes to improve transparency and interagency workflow management.
Governors and legislatures can institute state-level digital permitting platforms.
Governors and legislatures can create shared staff resources to hire the high-demand technical expertise necessary for permitting and enable those experts to move quickly between agencies in response to need.
State agencies and legislatures can initiate programmatic environmental reviews for clean energy projects subject to state environmental review laws.
Virginia’s Permitting Enhancement and Evaluation Platform (PEEP), now expanded as the Virginia Permit Transparency (VPT) system, was enacted by executive order. The VPT provides a public dashboard to track permits through state processes and workflow management tools for agency staff.
The Washington State Legislature created the Clean Energy Siting and Permitting (CESP) Grants Program, providing roughly $4.85 million to local governments, tribes, and state agencies to hire staff and modernize permitting workflows for solar, wind, and storage projects. By funding technical staff and process improvements, the legislature helped agencies reduce backlogs and provide more predictable permitting timelines for developers.
In 2025, Washington’s Department of Ecology, under direction from the legislature, completed programmatic environmental reviews for clean energy projects to speed permitting application and review of onshore wind, solar, and hydrogen projects in the state.
In Arizona, digitization of several environmental permitting processes helped reduce decision timeframes by 91 percent.
Build Administrative Capacity to Plan for an Affordable & Reliable Grid
Today, the U.S. bulk transmission system faces significant constraints that limit where new clean energy projects can be built and threaten overall grid reliability. Many regions with abundant clean energy resources simply do not have enough high-voltage transmission capacity to deliver that power to population centers. As a result, developers are increasingly unable to move generation projects forward even when siting, permitting, financing, and interconnection queue positions are in place. Without new transmission capacity, interconnection backlogs grow, power costs increase, and states are forced to rely on older fossil resources simply because they are already in place.
Transmission buildout is thwarted by barriers such as long planning timelines of 7 to 15 years, route identification, environmental review, litigation, supply chain constraints, and fragmented and inadequate planning processes.
While the permitting reforms described elsewhere in the playbook would help, we won’t build the transmission system that we need without improved planning. Building transmission lines requires utilities, developers, customers, and grid operators to work together to determine where a transmission line is needed and appropriately allocate costs across different stakeholders. Without a strong administrative state that can facilitate the process and collect and share all the required information (such as congestion on current lines, hotspots of demand growth, areas with high potential for cheap clean energy, etc.), this process often fails and very rarely results in optimal expansion of the transmission system. Today, states and grid operators lack administrative capacity to conduct this planning process, which is hamstringing our ability to expand the grid.
States can build the capacity to improve planning in order to spur development of transmission lines with the greatest benefit for the public.
Solutions
I. Include Advanced Transmission Technologies in Planning (Legislatures, Governors, and Public Utility Commissions)
Advanced Transmission Technologies (ATTs) can be used to increase grid capacity on current rights-of-way, alleviating congestion and allowing for more efficient energy transfer without building new infrastructure. Utilities being able to increase efficiency and cost effectiveness of their infrastructure is especially important as load growth continues to increase across the country and raise retail electricity bills. For example, installing high-performance conductors increases the amount of electricity that can be transferred over an existing transmission line. By one estimate, reconductoring with these technologies could double transmission capacity on the current grid. Dynamic line ratings allow lines to carry more electricity when weather conditions are good, rather than defaulting to conservative limits on line capacity. Each type of ATT has its own advantages and benefits.
PUCs can dictate standards, enforce rules, conduct studies, and establish new policies that require and incentivize utilities to evaluate and deploy ATTs.
Legislatures can require utilities to include evaluation of ATTs in planning processes, conduct studies on ATT potential and deployment opportunities, and analyze ATTs as potential enhancements to new transmission infrastructure.
Governors can petition ATT rulemakings to the PUC via an executive order, can integrate ATTs into funding criteria for grid or resilience projects and direct economic development agencies to study the economic impacts of ATTs, and convene ATT task forces to set direction and collaborate with educational institutions to develop workforce training programs focused on ATTs installation, operation, and maintenance.
Utah’s SB 191 requires utilities to conduct an alternatives analysis for ATTs in IRPs and also provides language that the Commission can approve cost-recovery for ATTs if it is determined the deployment is cost-effective.
Ohio’s HB 15 requires that utilities summarize ATT evaluation in power siting board certificates and furnish annual 5-year reports on ATT deployment opportunities, including congestion mitigation studies and that the PUC evaluate the potential of ATT deployment including consultation from stakeholders via two public workshops.
Governor Wes Moore’s December 2025 Executive Order Building an Affordable and Reliable Energy Future creates a Transmission Modernization Working Group that makes ATT policy recommendations to the Maryland Energy Administration, which in turn makes formal petitions to the PUC.
Montana House Bill 729, adopted in 2023, enables the state PUC to set cost-effectiveness criteria to allow utilities to deploy advanced transmission conductor technologies and recoup the cost via their ratepayers, similar to investments in new energy generation.
II. Create a New Transmission Planning Authority (Governors and Legislatures)
Lack of coordination between transmission and generation planning creates inefficiencies and prevents smart clean energy development. In deregulated markets—and in some vertically integrated states—transmission and generation planning processes occur largely in isolation without systematic processes to align long-term clean energy expansion with major grid upgrades. While the federal government has authority to set the rules for planning regional and interregional transmission lines, state leaders have tools at their disposal to expand transmission buildout and improve planning.
Legislatures can create transmission planning authorities explicitly authorized to identify transmission corridors that can expand low-cost clean energy generation, lead on the permitting and siting of transmission lines, secure project finance, negotiate and collaborate with other states on interstate transmission plans, provide advice on transmission priorities and planning needs for the state, and enter into public or private partnerships to help with project development. These authorities must be empowered and resourced to collect all the necessary information (e.g., congestion on the existing system, load forecasts, sites of cheap clean energy, etc.) and to attract top talent with expertise in utility planning, project development, and financing.
Governors can create a transmission advisory or coordinating committee and reorganize state agencies, boards, and commissions to serve the purpose of a transmission authority or to create one.
New Mexico passed the Renewable Energy Transmission Authority (RETA) Act in 2007, creating RETA and authorizing it to “plan, license, finance, develop and acquire high-voltage transmission lines and storage projects to help diversify New Mexico’s economy through the development of renewable energy resources.”
In Colorado, SB21-072 created the Colorado Electric Transmission Authority (CETA) to plan and develop transmission lines to increase reliability and deploy more clean energy. CETA has very similar powers to New Mexico’s authority.
III. Require Integrated Transmission and Generation Planning (Governors, Legislatures, and Public Utility Commissions)
Coordinated planning is essential to ensure that transmission is expanded in the right places and that new clean energy investments can flow to areas with sufficient transmission capacity. Around 35 states require their utilities to develop Integrated Resource Plans (IRPs), which act as a roadmap for how the utility will meet future forecasted electricity demand over a specific time period. Although transmission and generation are key inputs for energy supply, they are usually not included in these plans. The result is piecemeal grid planning, as transmission providers and developers focus on smaller lines which meet near-term needs and are profitable within their own footprint.This shortcoming is a product of both process—regulators and state agencies have not been mandated to link transmission and generation planning—and capacity, where the administrative state lacks the right staff and resources to conduct integrated planning.
Integrating these processes can ensure better coordination between load and generator interconnection, a more holistic understanding and roadmap of current and future grid reliability and supply chain needs, help avoid duplicative investments and ensure costs for upgrades remain reasonable, and can lower the likelihood of stranded or undersized assets. This integrated planning is especially important in places with projected load growth, whether from data center buildout or electrification of buildings, heavy-duty transportation, or factories.
Governors can direct relevant agencies to work with grid operators, PUCs, and utilities to encourage integrated planning.
Legislatures in vertically integrated states can require utilities to conduct IRPs where they don’t already do so and further require generation and transmission planning to be integrated.
PUCs can require utilities to link transmission and generation planning.
Enacted in 2021, Nevada S.B.448 requires an electric utility to amend its most recently filed resource plan to include a plan for certain high-voltage transmission infrastructure construction projects that will be placed into service before 2029.
In 2022, a Memorandum of Understanding (MOU) between the California Independent System Operator (California ISO), the California Public Utilities Commission (CPUC), and the California Energy Commission (CEC) ensured that the planning and implementation of new transmission and other resources were linked, synchronized, and transparent.
IV. Ensure Effective Implementation of FERC Order 1920 (Governors, Legislators, and Public Utility Commissions)
Recent federal actions, such as FERC Order 1920, have the potential to be a useful tool for states if implemented correctly and efficiently. FERC Order 1920 requires long-term, forward-looking, multi-value regional planning. It was designed to improve transparency in local transmission planning, including by conducting local stakeholder meetings. Under this filing, transmission providers must produce long-term, at least 20-year, regional transmission plans at least every five years, which must utilize seven specific categories of forward looking factors, select projects based on different economic and reliability benefits, and consider the use of grid-enhancing technologies.
Governors can take a more active role with PUCs to guide their involvement in regional transmission planning processes established under FERC Order No. 1920.
State legislators can hold hearings with PUCs on how utilities, regional transmission planners, and state officials plan to participate and support regional planning and put the order into action.
In the mid-Atlantic, 69 legislators from 10 states called on PJM to implement FERC Order 1920 without delay due to the benefits of reliable, affordable and clean electricity it will bring to their constituents.
Fix Broken Incentives to Expand Distributed Energy Resources
Misaligned incentives for utilities have limited the types of technologies and solutions in play to meet demand growth and maintain reliability.
Distributed energy and grid flexibility solutions—such as rooftop solar, flexible demand, smart charging for electric vehicles, and distributed storage, as well as alternative transmission technologies—can help meet demand growth faster and cheaper than solely relying on large power plants and bulk transmission upgrades. Distributed energy resources are particularly useful in an era when interconnection delays and economic pressures on rate payers have slowed down efforts to build large-scale transmission projects and generation facilities.
Investor-owned utilities make a profit based on a fixed rate of return on certain expenditures. Utilities typically do not earn a return on distributed energy resources and grid flexibility programs so many utilities have underinvested in these solutions and underutilized them for grid planning.
Largely because of these broken incentives, there’s a lack of capacity to value, procure, and orchestrate these distributed resources, thus limiting their ability to scale as a utility resource. In most states, utilities and regulators do not consider distributed resources in planning processes. Where distributed resources are considered, it is often only through voluntary offerings and limited pilot programs.
Leveraging distributed resource solutions requires state leaders to fix broken incentives and build new government capacity tools to facilitate uptake and utilization of these technologies.
Solutions
I. Incentivize adoption and use of distributed energy resources (Legislatures and Public Utility Commissions)
States can correct misaligned incentives by creating mechanisms to value the benefits of distributed energy resources and incentivize optimal utilization of distributed resources to reduce spending on new generation and grid upgrades.
Legislatures can establish Virtual Power Plant programs—run by utilities, public developers, or partnerships between public developers and utilities—to aggregate distributed energy resources to best serve the grid and compensate these resources accordingly.
Legislatures can incentivize adoption of appliances and customer energy systems, like smart chargers and thermostats, capable of supporting the electricity grid.
Legislatures and PUCs can require utilities to create mechanisms to value the benefits of distributed resources and adjust profit motives through performance-based ratemaking to align utility incentives with expanding distributed resources.
In Colorado, SB24-218 required Xcel Energy to create a new mechanism to compensate customers for distributed energy resources that can benefit the grid.
In New Jersey, Governor Sherill issued an Executive Order directing the PUC to establish a Virtual Power Plant program to aggregate distributed energy resources and use them to benefit the grid.
II. Require procurement of distributed resources (Governors, Legislatures, and Public Utility Commissions)
Utilities generally do not procure distributed resources, focusing instead on procurement and buildout of large-scale generation. Formal procurement of distributed resources can maximize use of the existing grid and get new generation onto the grid fast, which can avoid spending on the distribution grid and save customers money.
Legislatures can set statutory requirements for minimum procurement amounts.
PUCs can require utilities to conduct procurement of distributed clean energy.
Governors can direct PUCs to initiate solicitations for distributed clean energy and set timeframes and minimum procurement amounts that PUCs and utilities must meet.
In Minnesota, Xcel Energy has proposed a Distributed Capacity Procurement of 50-200 MW of distributed energy storage. Under the procurement, the utility would contract with a developer to build small-scale batteries and pay households, businesses, and non-profit organizations to host them.
In Illinois, the Clean and Reliable Grid Affordability Act required the procurement of 3 GW of storage and created a mechanism to compensate households and businesses for customer-owned storage. The state has a government agency—the Illinois Power Agency—that develops procurement plans and runs competitive procurement processes.
In Maine, the PUC issued a Request for Proposals for solar projects on PFAS-contaminated farm land.
In New Jersey, Governor Sherill directed the PUC to procure an additional 3,000 MW of community solar through the NJ Community Solar Energy Program.
III. Improve interconnection processes for distributed resources (Governors and Public Utility Commissions)
The interconnection process is a major hurdle for new clean energy projects. Utilities run the process to connect projects to the distribution grid, and PUCs regulate this process. Because utilities typically do not profit from distributed generation, they have generally deprioritized improvements to the interconnection process for distributed generation. PUCs have not had the mandate to push utilities on improving the interconnection process for distributed resources, which means capacity has not been channeled toward this goal. State leaders can force utilities to speed up the interconnection queue to get more clean energy on the grid.
Governors can direct PUCs to improve the interconnection process with specific goals (e.g., reduce the application approval time by a specified amount) and penalties for non-compliance.
PUCs can work with utilities to make improvements to the interconnection process to shorten timelines and provide more certainty to developers. PUCs can dedicate staff resources to this issue to ensure rapid improvements to the process.
Colorado Governor Polis in an August 2025 directive tasked state agencies to pursue flexible interconnection and voluntary curtailment for distributed energy and community solar projects; work to facilitate the pre-purchase of project equipment and/or affiliated electric transmission and distribution infrastructure; and to pursue updates to interconnection standards for customer-sited projects, including the use of meter collar adapters and other measures to minimize the cost and time. The Colorado PUC is now working to allow projects to use flexible interconnection agreements—alowing projects to connect to the grid in constrained places if they agree to use technologies that can operate flexibly and avoid straining the grid—which can speed up timelines and reduce interconnection costs.
In New Jersey, Governor Sherrill directed the PUC to take several steps to standardize and speed up interconnection of clean energy projects to the distribution grid.
IV. Improve data and analysis and increase transparency to optimize deployment (Governors, Legislatures, and Public Utility Commissions)
Expanding distributed energy resources in the right places is key to maximizing benefits to the grid and reducing costs for customers. States can help inform strategic adoption through data collection and analysis on the optimal places to deploy customer-owned resources and increase transparency of those data for the public.
Policymakers can analyze opportunities that will make the most impact through avoided costs of updating the grid or building bulk generation. Distributed energy projects are most useful when they minimize constraints on the distribution system and reduce peak demand. Information sharing can be facilitated through either Integration Capacity Analysis or regular Integrated Distribution Planning.
State leaders can also identify potential projects to provide tailored support by engaging developers and site hosts directly through Requests for Information. States can then use this information to direct support to projects. For example, data collection can enable aggregation and standardization of information about siting readiness, interconnection considerations, terms and conditions, and resilience needs which can then be converted into a pre-qualified inventory for procurement or financing.
Governors can direct agencies to engage with developers and other stakeholders to identify barriers to getting distributed projects built and identify solutions to support projects (e.g., connecting host sites with developers and financiers).
Governors and PUCs can improve distribution system planning to better anticipate infrastructure needs and ensure that distribution system planning is synced with grid operator interconnection processes and transmission planning.
PUCs can require utilities to conduct and publish analysis on hosting capacity and locational value of distributed clean energy resources and establish a market for distributed energy resources.
In Massachusetts in 2024, the PUC directed utilities to incorporate “non-wires alternatives”, including distributed energy resources, into system planning. Now, one of the state’s largest utilities has launched an innovative mechanism to quickly procure distributed energy resources that provide the greatest costs reductions and grid benefits.
The New York Build‑Ready program identifies potential host sites and shifts early development duties (i.e., negotiate the initial lease and start the interconnection process) from private developers to the state, which accelerates the move of renewable energy projects from the idea stage to construction.
In New Jersey, Governor Sherill directed the PUC to work with the utility to improve hosting capacity maps for distributed energy resources.
V. Create a Municipality or Sustainable Energy Utility in lieu of an Investor Owned Utility (Legislatures and Local Leaders)
Creating a municipal utility or sustainable energy utility (SEUs) can address bypass misaligned incentives and focus on the technologies and strategies that most benefit the public. These entities differ from traditional investor owned utilities because they are not-for-profit, owned by the communities they serve, and are run by local government. Benefits of this utility model include securing more affordable electric rates for consumers, achieving renewable and clean energy goals more quickly, increasing local control and governance over energy decisions and infrastructure, and contributing to local economic development.
Legislators and local leaders can spearhead the effort of municipalizing by developing a concept and authorizing or establishing a municipal utility. State and local leaders can explore different options, including developing a supplemental utility that procures generation but still uses the private utility’s poles and wires or fully acquiring the private utility and creating a public entity to run the full system.
In 2024, residents of Ann Arbor, Michigan voted to authorize Proposal A: Creation of a Sustainable Energy Utility. The Sustainable Energy Utility (SEU) will be an opt-in, supplemental, community-owned energy utility that provides 100% renewable energy from local solar and battery storage systems.
In 2005, residents of Winter Park, Florida voted to authorize the city’s use of bonds to buy the local distribution facilities from the incumbent utility.
The state of Nebraska receives all of its electricity from publicly owned sources.
Wield Creative Finance Tools to Drive Investment and Reduce Capital Costs
Rollbacks of federal financial support have threatened the viability of many clean energy projects. State and local leaders can help keep projects alive and build new ones with creative financing tools. In some cases, this means taking a more active role in coordinating across public and private sector actors, while in others that means building entirely new administrative capacities to perform more ambitious financial transactions or act as a public developer.
In addition, the grid is facing new challenges that require massive investments. For example, recovery from and preparation for wildfires is inflating energy bills in the west. Gulf states are facing similar costs from hurricanes. States need creative finance tools to ensure that these costs do not continue to raise bills for regular people and small businesses.
Beyond merely acting as a source of capital, governments of every shape and size actively participate at every stage in the project development and planning lifecycle to bring down the total cost of projects. These include lowering financing costs, securing stable or catalytic financing, and providing an avenue to complement other functions the state is undertaking. Local governments can engage in public development functions, including through creative finance tools and engagement with community choice aggregators, rural electric cooperatives, and energy service companies.
Solutions
I. Empower development entities with the legal authority and staffing to pursue high priority projects (Governors, Legislatures, Local Leaders)
State leaders can help ensure that infrastructure authorities, city and county development corporations, or energy departments of a given jurisdiction have the relevant borrowing authority, ownership and operation powers, and partnerships capabilities to support project development.
To be successful, state financing entities or public developers need clarity and certainty on how projects they support can participate in electricity market operations, including whether projects can participate in utility procurement processes or interact with grid operator interconnection processes. State financing also must be coordinated with other grid planning processes.
Given the overlapping interests state and local economic development agencies may hold, this process will demand adequate staffing resources and may require significant stakeholder engagement with private sector actors, government officials, and others.
Legislators can write or amend enabling authorities to explicitly provide state and local entities with the financing, bonding, ownership, and partnership authorities necessary to support, finance, own, and/or operate projects. These authorities should include co-financing and co-development options to blend public and private support. Legislators can also make sure that these authorities are flexible and broad so that state development can be competitive with private developers.
Legislators can allow use of public financing tools to support certain projects. In particular, legislators can expand the bonding authority available to state agencies for use on clean energy projects.
Legislators can establish state and/or utility procurement targets for clean energy, storage, and grid projects and provide direction and clarity for state financing entities to service these procurements.
Governors can use their authority over appointments and interagency coordination to align disparate entities around specific tangible objectives.
Governors can draw on recent public private partnerships in the offshore wind industry to structure offtake, procurement, and other commercial activities with utilities and developers across a range of clean energy projects. State entities can seed virtual power plants, solar, wind, energy storage and other clean power projects that mutually derisk projects for both public and private developers alike.
Governors and legislators can provide expedited permitting and siting processes for publicly sponsored projects.
City and county officials can form project-specific entities or special purpose authorities to make projects financeable.
In New Mexico, the Renewable Energy Transmission Authority (RETA) was established in 2007 and was granted statutory power to exercise eminent domain to acquire property or rights of way for eligible renewable energy projects. This authority has been critical in overcoming fragmented land acquisition barriers.
The Connecticut Green Bank’s Solar Marketplace Assistance Program (Solar MAP) serves as a public developer to finance and build solar projects for K-12 schools, allowing the state to own the assets and sell power back to districts at a discount. While the Green Bank has acted as a public public developer in some form since 2014, projects from Solar MAP are projected to deliver tens of millions of dollars in savings all without incurring any upfront costs for districts.
In Colorado, SB 21-072 in 2021 created the Colorado Electric Transmission Authority as a special-purpose development authority granted power to issue bonds and corridor acquisition tools.
II. Use pooled loan funds like state bond banks to lower borrowing costs and build project pipelines (Governors, Legislatures, and Local Leaders)
Pooled borrowing authorities offer transaction efficiency and credit strength for cities, counties or small utilities paying the fixed costs of a standalone bond issuance by aggregating relatively modest projects into standardized pools. This reduces the issuance and underwriting costs, and can often enhance credit resulting in lower borrowing costs.
Bond banks are valuable in practice because they are a repeatable financing infrastructure that can be improved and expanded over time. Governors offices, county executives, and mayors can direct agencies to build a steady pipeline of eligible projects (using Requests for Information or direct engagement) and then work with relevant financing authorities to standardize project intake, selection, and reporting and make the whole process more repeatable.
Once local governments experience lower borrowing costs and faster execution through a standardized conduit, the model becomes politically sticky and easier to scale, especially when paired with complementary tools like revolving funds or credit enhancement that can serve smaller borrowers and accelerate project turnover.
Legislators in states that lack a bond bank can establish one capable of pooling local loans, issuing bonds, and relending the proceeds. They can further work to standardize project solicitation, underwriting, and closing cycles to ensure the institution creates a regular cadence.
Legislators in states that have a bond bank can expand eligible project types (to include clean energy projects and resilience priorities like building retrofits, microgrids, etc.) and create standard project templates.
Governors can work with state agencies to centralize the origination of bonds for a public developer in their state’s bond bank or otherwise help public developers and other financing agencies exercise their bonding authority.
Governors can make regular use of bond banking authority a priority by directing agencies to run a standing intake process, and appoint or empower relevant state personnel to highlight pooled lending as an innovative solution.
City and county officials can create a rolling inventory of eligible projects, bundle them into multi-jurisdiction project aggregators and engage with existing bond banks on technical assistance for project scoping and diligence.
Vermont’s Bond Bank issues bonds backed by repayments of its loans to individual municipalities, school districts, etc. and maintains a dedicated Municipal Climate Recovery Fund. The bank has the ability to backstop non-payment by municipal or county entities that fail to pay based on state funds allocated to municipal or district borrowers in what is known as an “intercept mechanism.”
Virginia Resource Authority’s Resilient Virginia Revolving Fund was established in 2022. Jointly administered with the state’s Department of Conservation and Recreation, the pooled borrowing platform provides financial assistance for flood-mitigation projects across the state.
III. Require energy utilities to supplement portions of their debt or equity with public bonds (Governors, Legislatures, and Local Leaders)
A unique characteristic of public development is that strategic capital deployment has the potential to derisk private investment. Mandating that utilities replace a portion of their high-cost equity with state backed public debt or revenue bonds optimizes the project’s capital stack, thereby reducing the average cost of capital and reducing the total financing costs for capital-intensive grid infrastructure. Investor-owned utilities typically finance large infrastructure projects through a mix of debt and equity with regulators guaranteeing a return on equity (ROE) to attract private investors. Because this ROE is significantly higher than the interest rates on public debt, requiring public bonds to supplement the capital stack can dramatically reduce the long-term costs that are ultimately passed on to ratepayers.
This mechanism leverages the state’s superior credit rating and tax-exempt status to fund the most expensive portions of development while leaving the utility to focus on its core competencies of construction and grid operation. Establishing a public financing facility in this way allows the public sector to act as a sponsor investor for projects of high public interest, such as interregional transmission lines. By providing lower cost debt, states can ensure that critical energy targets are met without placing an undue financial burden on households. This approach creates a more stable investment environment and allocates risks more effectively across public and private stakeholders.
Legislators can mandate investor-owned utilities make use of state-backed revenue bonds or other forms of public debt to finance high-priority capital investments such as grid resilience or interregional transmission.
Legislators can authorize state infrastructure banks or other financing authorities to act as sponsor investors and displace high cost equity of a project’s capital.
Governors can establish dedicated clean energy project finance working groups to examine the full scope of infrastructure financing tools needed to derisk capital investment in transmission, generation, distribution and other electricity assets.
Regulators and state energy offices can lower the costs passed along to ratepayers by integrating public financing facilities directly into RFP processes, allowing bidders to access lower-cost capital.
City and county officials can pass local resolutions advocating for a specific local utility project to be financed via public bond rather than traditional utility equity to ensure the lowest possible rate impact for their residents. A similar strategy can be pursued via written submission or intervention within PUC docket proceedings.
City and county officials can collaborate with state energy offices to identify projects that are ideal candidates for public debt supplements.
California’s 2025 law SB 254 establishes a state public financing facility (the Transmission Infrastructure Accelerator) to replace high-cost utility equity with lower-cost public debt for new transmission projects, directly reducing the ratepayer impact of CAISO’s multi-decade development plans. The law requires utilities to finance billions of dollars of grid hardening investments using bonds instead of utility equity financing, reducing costs for customers and preventing the utility from excessively profiting off of this set of expenditures.
Maine’s Clean Energy Financing Study recommends operationalizing state revenue bond authority and establishing a working group on large clean energy project finance to optimize the capital stack for clean energy and transmission projects.
IV. Develop greater public understanding about the development levers available to public or quasi-public entities (Governors, Legislatures, and Local Leaders)
Some financial functions like loan issuance, co-financing, and non-dilutive debt financing may be well known to state energy offices, green banks, and certain infrastructure authorities. But in general, public financing is hampered by a lack of clarity, information, and standardization of different agencies’ authorities.
States can maximize the impact of public resources by establishing clear financing authorities and responsibilities, providing state authorities with broad powers to flexibly support projects, ensure that public finance is prioritizing the right investments, and providing clear direction on how publicly sponsored projects support utility procurement or grid operator processes. In addition, standardizing state and local financing entities drives down costs by making processes more repeatable and can pave the way for more effective federal support in the future. By surfacing all the capabilities public entities currently have and may wish to develop in the future, policymakers and advocates can align on objectives to strengthen the public developer toolkit and bring clean energy projects closer to fruition.
Governors can inventory borrowing, contracting, and financing authorities and provide clear guidance on roles and responsibilities between agencies.
Governors and legislatures can require reporting on key performance metrics like deal volume, borrower participation, and time-to-close to help encourage institutionalization.
Governors and legislators can publish analysis and information on areas to focus energy project development and create special zones for the installation, procurement, manufacturing, or operation of energy projects of various kinds. These industrial zones could provide access to a variety of benefits: expedited permitting, siting, interconnection, specific public finance facilities, funds for resiliency + operation, and various other coordination benefits from other interested state agencies.
Legislatures can provide agencies with clear financing authorities, direction on what types of projects to support, and a broad set of tools to flexibly support projects.
City and county officials can examine if there are relevant state laws that require additional ordinance/resolution to use. Some tools to activate and then specify rules to create repeatable administrative playbooks.
Houston, Texas had to pass an authorizing city ordinance to activate a state program known as Property Assessed Clean Energy (PACE). The program allows commercial and multifamily property owners to finance energy efficiency, renewable energy, and water conservation improvements and has invested over $540 million dollars statewide since its inception in 2016.
Montgomery County, Maryland created a green bank in 2016. In 2022, the county passed a statute to direct 10% of the county’s fuel tax revenue to the Montgomery County Green Bank each year. The green bank completed a new bus depot for EV buses in 2022 co-located with a 6.5 MW microgrid that can run independent of the local utility.
Colorado has an EPC program that lends against a project’s anticipated cost savings to finance building retrofits.
Solving the Clean Energy Infrastructure Finance Rubik’s Cube
Building Blocks to Make Solutions Stick
Capital is not the constraint, alignment is: Catalyzing large-scale climate and energy infrastructure requires government to act as a systems integrator—synchronizing policy, de-risking commercialization, modernizing valuation, and coordinating markets so private capital can move with speed and confidence.
Implications for democratic governance
- Visible coordination builds credibility: Investors, communities, and companies need to see how policy pieces fit together. Fragmented, asynchronous policymaking erodes trust and slows deployment.
- Risk tolerance must be publicly legitimized: If democratic institutions punish every failed deal but ignore portfolio-level gains, agencies will default to paralysis. A mature democracy must distinguish responsible risk-taking from mismanagement.
- Transparency is market infrastructure: Open data, common modeling tools, and clearer capital pathways empower regulators, communities, and innovators to interrogate and improve investment decisions.
Capacity needs
- Systems-level policy choreography: Agencies capable of synchronizing rules, guidance, financing programs, and permitting reforms on coordinated timelines rather than rolling them out in isolation.
- Transaction-speed infrastructure: Staffing models, underwriting playbooks, and surge capacity that match private-sector deal timelines while maintaining integrity.
- Interstate coordination platforms. Formal mechanisms for states to harmonize standards, pool procurement, share data, and replicate successful pilots without restarting from scratch.
- Accessible technical/economic infrastructure: Publicly credible data sets, modeling tools, and valuation methodologies that lower barriers to entry and allow meaningful third-party scrutiny.
Deal templates and archetypes: Clear, standardized financing pathways that signal how government capital will engage at different risk tiers and technology stages.
Jump to…
- Executive summary
- “Come together”: Defragmenting markets through regional coordination
- “That’s what friends are for”: Overcoming commercialization barriers through partnerships
- “Highway to the deployment zone”: Faster, risk-weighted transaction execution
- “Okay, now let’s get in formation”: programmatic policy synchronization for fast market formation
- “C.R.E.A.M.”: More holistic valuation tools and methodologies
- “Take it to the bridge”: Rethinking the ‘missing middle’ problem
- “The Next Episode”
Executive Summary
Historic commitments. Huge demand. Massive cost reductions. Ready technologies. Yet, infrastructure deployment levels are underperforming their potential. What gives? The U.S. clean energy sector has achieved remarkable milestones: solar and wind have tripled since 2015, costs have fallen 90%, and annual clean energy investment now exceeds $280 billion. Yet deployment has arguably fallen short of what both markets and the climate moment demand. The culprit isn’t a single bottleneck: not permitting, not subsidies, not technology readiness alone. The real constraint is misalignment across the multiple interdependent factors that investors need to see in place before committing capital at scale.
Think of it like a Rubik’s Cube: solving one face means nothing if the other five stay scrambled. This paper identifies six strategic levers that, when pulled in concert, can unlock the conditions for large-scale capital deployment:
- Market defragmentation: breaking down the patchwork of ~3,000 utilities and state-by-state rules that trap promising solutions in regional silos
- Commercialization partnerships: deploying innovative public-private joint ventures to de-risk capital-intensive, first-of-a-kind energy infrastructure that traditional markets won’t move on alone
- Transaction execution speed: closing the yawning gap between private-sector deal timelines and the multi-year gauntlet of public financing processes
- Policy synchronization: coordinating the release of rules, funding programs, and guidance so investors see a complete picture, not a puzzle missing half its pieces
- Holistic valuation: building common information infrastructure and market structures that capture the full economic value of energy solutions, moving beyond narrow cost metrics that systematically undervalue transformative infrastructure
- Proactive investor engagement: later-stage investors jumping in sooner and addressing the hard questions early to help bridge the “missing middle” between venture and infrastructure finance
The good news: the capital exists, the technologies are ready, and infrastructure is a solvable problem. With over 1,000 GW of clean energy in development and electricity demand projected to grow up to 50% over the next decade, the infrastructure build-out represents one of the largest capital deployment opportunities in American history. And global demand for U.S. clean energy technology has never been higher. The barriers identified in this paper are structural and systemic, not fundamental; most of the solutions proposed are actionable in the near term, without waiting for perfect legislation or perfect markets.
The window is open, but not indefinitely. For policymakers and investors alike, the question is not whether to act, but whether to act with the clarity, coordination, and urgency the moment demands. The frameworks, partnerships, and policy tools outlined here offer a practical roadmap for unlocking decades of economic growth, cost-of-living relief, and energy security for communities across every region of the country and beyond. The energy transition is not a cost to be managed; properly coordinated, it is a generational economic opportunity.
America has experienced extraordinary momentum in the growth and transformation of the energy sector. Solar and wind generation has more than tripled since 2015. In 2024, 50 GW of solar power was added to the U.S. grid, which is not only a record, but is the most new capacity that any energy technology has added in a year. Technology costs have plummeted dramatically: utility-scale solar and battery energy storage have each fallen 90% since 2010. Those declines have made them among the lowest cost forms of electricity in many places. Domestic manufacturing capacity has also surged with hundreds of clean energy manufacturing facilities, many of which have already come online. These technologies and projects have been reinvigorating communities and creating jobs across the nation, and we should see the benefits from these advances continue as capital is flowing into this sector at an unprecedented rate. Clean energy investment in the United States more than tripled since 2018 to $280 billion annually, with multiples more in commitments, and private markets alone have raised nearly $3 trillion over the past decade. For more established technologies like utility-scale solar and onshore wind, financing has become standardized. This includes established project finance structures, robust secondary markets, accurate energy production forecasts, and predictable returns that align with the needs of institutional capital. These asset classes now exhibit many of the hallmarks of market maturity: transparent pricing, deep liquidity, sophisticated risk assessment frameworks, and predictable transaction execution. This progress has been galvanized by unprecedented governmental support, including the Bipartisan Infrastructure Law of 2021 and the Inflation Reduction Act of 2022 (IRA), alongside bold state policies, aggressive corporate clean energy procurement, sustained advocacy, and relentless technological innovation.
Yet despite these achievements and the trillions of dollars in committed capital, the pace of deployment has arguably fallen short of what the market opportunity demands and what the climate crisis requires. Hundreds of IRA-supported energy and manufacturing projects have faced delays or cancellations: much of that is due to increased economic and logistical uncertainty (e.g. in the cost and availability of equipment, permitting timelines, import and export regulations); much of that too is due to sharp reversals in federal funding priorities (e.g. tax incentive changes from the One Big Beautiful Bill Act (OBBBA), direct project cancellations). Moreover, emerging solutions are still taking time to achieve true commercial liftoff. Despite billions in federal funding allocations, only a few carbon capture projects have meaningfully progressed, with others indefinitely delayed or cancelled. Growth of some demand-side energy solutions, like behind-the-meter solar and virtual power plants, has remained relatively regional, despite favorable economics. Advanced nuclear energy, though it remains a policy priority, has been challenged with long delivery timelines, and many project investors remain wary of the risk of cost overruns. Sustainable aviation fuel production capacity increased by ten times in 2024, but it is still a very small fraction of jet fuel demand. Furthermore, transmission capacity has not grown nearly as quickly as needed and remains a key constraint to progress.
The situation – deployment deficiencies despite historic support – can be entirely solved with just one thing… and that is… to stop acting as if there is just one thing. The relative underperformance described earlier is not attributable to any singular constraint: contrary to what some have argued, for instance, it’s not solely about removing permitting roadblocks nor creating more subsidies. Rather than seeking silver bullets, real progress can be made by recognizing that there are multiple elements involved and misalignment between those elements have curbed the rate of progress.
Accelerating new energy project investment is somewhat like solving a Rubik’s Cube. The key to solving the puzzle lies in its interdependence: every twist of one face ripples across the others. You can solve one face perfectly, but if the other five remain scrambled, you haven’t solved anything in the grand scheme. Progress requires coordinated progress across multiple dimensions simultaneously in the right sequence. The cube rewards systems-thinking and algorithms over siloed, non-coordinated actions. All six of its faces must align to win.
The same is true for new energy finance. When most project investors look at a sector, they approach it as a puzzle and look for as much alignment of the full picture as possible before being sufficiently comfortable to deploy capital. That’s the indicator for the risk-reward balance being in the right place to justify investment. Rather than defining the theory of progress by singular issues, policy and industry stakeholders need to create sufficient alignment of multiple puzzle pieces at the same time.
This paper offers a few perspectives on how to achieve the conditions for larger-scale capital deployment, drawing on both lessons learned and promising concepts from across the industry. Like the six faces of the Rubik’s cube, six priority strategy areas are articulated: market defragmentation, commercialization partnerships, transaction execution speed, policy synchronization, holistic valuation methodologies, and proactive investor engagement. Note that while some of the underlying strategies may take time and require deep structural realignment, most of the concepts discussed herein are actionable in the near term. A range of stakeholders, from policymakers to infrastructure investors to community and industry advocates, need to move in concert to solve the puzzle and unlock greater investment.
The opportunity before us is immense. With over a thousand gigawatts of clean energy in development and electricity demand projected to grow up to 50% over the next decade, the infrastructure build-out required represents one of the largest capital deployment opportunities in American history. Similarly, global demand for U.S. clean energy technologies to be a bigger part of the mix has soared over the past few years, as many countries are seeking to diversify away from China or access some of the more unique technologies that the U.S. is developing. And solutions can’t come quickly enough, in this era of fast-growing energy demand, spiking electricity bills, aging physical infrastructure, and burgeoning new industries, not to mention a plethora of old and new technology solutions and operational strategies poised to meet the moment. The question is not whether the capital exists (it does!), whether energy solutions are available (they are!), nor whether there is a silver bullet salve (there isn’t!). It’s whether we can align the six faces of our energy finance cube quickly enough and strategically to channel the right types of capital where it’s needed most, when it’s needed most. The energy transition presents a real opportunity to drive economic transformation that, when properly coordinated, can unlock decades of strong economic growth, cost-of-living reductions, innovation, and prosperity across every region of the country and the planet.
Chapter 1. “Come together”: Defragmenting markets through regional coordination
For many companies across multiple sectors, the U.S. market can seem like the golden goose. Its big population, high income, diversified economy, and strong purchasing power typically mean large total addressable markets (TAM). While those drivers can be true, the reality for many energy and climate solutions, especially early on, is that the large TAM is challenging to realize, as the market can be highly fragmented. In those sectors, the U.S. is less of a “market” per se and more of a loose mosaic. There are about 3000 utilities, ranging from large publicly traded corporations to rural cooperatives, in regulated and regulated markets. States and territories not only have different market drivers, but they also have their own regulations and regulators, business processes, permitting requirements, and market rules. This also complicates the go-to-market as you typically need large and locally-focused commercial organizations to tap the markets, which can be expensive and time-consuming to build, especially for newer companies. It also often means a less efficient path to scalability, as each set of local customers and regulators need to be brought up to speed and convinced about the fit of a solution (compared to having a few entities that speak for the entire country). In addition to the commercial elements, this dynamic introduces technical barriers to scalability, especially where deep integration and redesign are required to meet local requirements. Over the years, this has often flummoxed both U.S. startups and experienced foreign investors, who have approached the U.S. market with high expectations, only to be confounded by these complexities.
Harmonize local requirements to avoid the piloting death spiral
To the extent possible, to promote more rapid and widespread investment and deployment of solutions that could benefit their communities, local (and national) governments need to work more closely together to harmonize market designs and project requirements. Oftentimes, a solution provider may implement a solution in one state, but when they go to another state, that utility might make them start from the beginning and prove themselves all over again – many innovators have likened these continuously repeated pilots to death by a thousand cuts. If a good solution is successfully deployed in one place, the barriers associated with deploying the same solution in another market need to be lower. This concept applies to permitting and design as well. The more that permitting processes and tariff structures, or modular elements within, can be templatized, time and uncertainty are reduced. Additionally, uniformity lowers development costs because the solution doesn’t have to be fully reengineered for each locale. This could also correspond to standardizing equipment and project technical requirements between them, to minimize costly product redesigns and reengineering. Furthermore, state stakeholders seeking to deploy similar solutions should consider entering into reciprocal partnerships or MOUs, supporting collaboration that’s both technical (e.g., between their utilities and independent engineering organizations) and policy-focused (e.g., between their policymakers and regulators). As such, when a solution under that type of agreement is evaluated and approved in one state, when that solution is brought forward in a partnering state, the solution can be given an expedited evaluation and approval process.
Not just physically, but digitally
The dynamic described previously is not limited to hardware. It is present for many software solutions as well, particularly those that have to integrate with local operators’ control systems. For example, a locality may be interested in deploying a virtual power plant (VPP), which is a relatively low-cost, software-based approach to aggregate distributed and controllable energy resources in order to provide large-scale energy services. A VPP deployment would have to connect to a utility’s and/or grid operator’s distributed energy management system (DERMS) to talk to devices, energy management systems (EMS), energy dispatch and trading system for wholesale market participation, customer information systems to track billing and energy usage, and also be cybercompliant – note too that several utilities have yet to fully roll out these foundational modern digital systems. Not only that, each utility and grid operator might have their own implementations (vendors, versions, configurations, rules) of these systems. Even outside of controls-oriented functions, the variety of data structures, naming conventions, and IT systems can make it difficult to access available market data (e.g., energy pricing), electricity tariff rate structures, and other highly important information. This is a reason why you tend to see that many energy software solutions have their operations concentrated in just a few markets, as the costs and time associated with integrating with another market’s cadre of systems can be hard to justify and thwart efforts to scale.
This is an area where states can work together (along with their respective utilities, grid operators, technology providers, and regulators) to agree upon more uniform ways to structure data, access market information, and securely interface with market and control systems. This could include partnering with groups that are developing common standards and protocols (such as RMI VP3, LF Energy), building an implementation roadmap across those states and utilities (accelerating implementation of FERC Order 2022), and taking corresponding legislative actions to ensure investments are made to build out the enabling foundational digital systems.
Aggregated demand and collaborative procurement
In a similar vein, collaboration between state and national governments can level the playing field and expand markets. When it comes to infrastructure, states and countries may often endeavor to ensure that local manufacturing capacity and supply chains are set up within their territories – this can create long term economic growth opportunities, reduce equipment delivery risks, and improve the public’s return on their investment.
However, issues can arise when multiple states are trying to duplicate efforts in the same sector. Take offshore wind in the early 2020s. Multiple eastern U.S. states were not only supporting a new wave of projects, many funding programs had requirements that the projects needed to source materials and equipment from suppliers located in the state.. The effect of this for a small, burgeoning industry was dilutive and slowed down factory investments as the scaling factors were harder to justify. After all, there are only so many blade, monopile, vessel, and cabling factories that can be supported at a given time, especially early on in the industry’s development. In response, thirteen states and the federal government signed a memorandum of understanding, where they agreed to take more regionalized, collaborative approaches to procurement and supply chain development.
Relatedly, an area where significant improvements could be made is around the procurement of critical common equipment. To accommodate the load growth from new factories, data centers, and building and vehicle electrification efforts, there are many pieces of equipment that will be needed irrespective of what types of energy are associated: things like transformers, circuit breakers, switchgear, and so on. There are considerable production capacity shortages and long lead times on these, which raise costs and create execution risks for projects. Despite the robust market demand, manufacturers have been somewhat hesitant to invest in expanding production as they worry that the demand will not materialize, which would leave them with underutilized or even stranded assets.
State and local governments can respond to these challenges in multiple ways. For instance, they could pool together their demand and drive standardization of the equipment so that the equipment is more fungible and interchangeable, as has been previously highlighted by the U.S. National Infrastructure Advisory Committee’s report on protecting critical infrastructure. Also, states can create well-defined demand guarantees where they can provide assurances to manufacturers and consumers that necessary equipment will be there, as needed.
For example, in 2013, the Illinois’ Department of Transportation led a seven-state procurement initiative to jointly acquire a standardized set of efficient locomotives and railcars, with additional funding provided by the Federal Railway Authority to support domestic manufacturing. This effort pulled forward new, more efficient railway vehicles into the market and lowered lifecycle costs. These concepts can additionally apply to secondary markets as well, for example, providing residual value guarantees on heavy-duty electric trucking procurement, to help mitigate risks on the initial purchase (e.g., traditional resale markets not emerging or asset residual values not being realized as projected).
Chapter 2. “That’s what friends are for”: Overcoming commercialization barriers through partnerships
The next facet of the cube pertains to early market formation and investment into technologies that are not yet fully commercialized, especially first-of-a-kind (FOAK) and early-of-kind (EOAK) infrastructure, and why capital formation has been easier for some types versus others. Differences in the ability to demonstrate, commercialize, and scale new infrastructure do not purely depend on the ultimate value of the solution; they are often driven by how the characteristics of that infrastructure affect the pathway to value realization, particularly the inherent capital-intensity and modularity of the solution.
For highly modularized solutions with lower capital requirements, the pathway can be much more straightforward. Take solar photovoltaics. Though module R&D and fabrication are far from trivial, technical demonstration and deployment are relatively simple. One can usually install and field-test new solar quickly and inexpensively. The advantage extends when scaling the solution to bigger projects: once you reach megawatt scale, given the modularity of solar cells and their balance of plant (inverters, cables, trackers, etc), you can obtain a reasonably clear picture of how even gigawatt-scale projects should fundamentally work. Other highly-modular solutions like batteries and EV chargers have enjoyed similar advantages. Tesla, for instance, in order to address potential range anxiety issues for its customers, leveraged its own balance sheet and government funding to build out a network of standardized superchargers, taking advantage of charger modularity to build in waves. This characteristic has made it easier for rapid demonstration and scaling of those solutions, as the financial community can enter the market, investigate, learn, and expand with relatively low risk.
However, the commercialization process becomes significantly more challenging with more capital- intensive and complex technologies, such as carbon capture, nuclear, e-fuels, etc. For some of these types of solutions, the early projects can require billions of dollars to construct and demonstrate. For these types of infrastructure, smaller-scale systems might not provide comparatively representative technical proof points of how the larger system needs to operate. Furthermore, larger sums of capital are often needed for early deployments. What’s more, the financial risk can be compounded as the long-term payoff is not guaranteed, as FOAK & EOAK projects typically have more uncertainty, and the learning rates of subsequent projects may also be less obvious. Furthermore, instead of a typical first-mover advantage that you often see with new technologies, early project investors here might actually suffer from a first-mover disadvantage, where they have the risk and cost of participating in the earlier projects, but don’t accrue the benefits and learnings that are seen in projects executed down the line.
To help address these types of challenges often seen with capital-intensive, less modular FOAK and EOAK infrastructure projects, adopting a new suite of partnership structures can greatly help to accelerate market formation and improve investability.
Multi-project joint ventures
Catalyzing capital for this class of infrastructure may mean going significantly farther than providing a few incentives and having strong advocacy efforts. More complex and elaborate agreements, private and/or public, are often necessary to drive deployment. Particularly in the form of deployment coalitions, consortiums, and joint ventures that support multiple projects. At the highest level, this can exist in several forms and can be originated by the private or public sector, as appropriate. For illustration, a few examples of commercial approaches to scale new nuclear energy projects, roughly in increasing order of relative deployment impact:
- Demand-side vehicles/buyers clubs. This is where consumers get together and make commitments to purchase a solution. This can be used as a market signaling mechanism to stimulate demand and attract solutions. There are example programs run by governments, such as the First Movers Coalition, and others led by the private sector, such as the Advanced Clean Electricity Initiative (Google, Nucor, Microsoft). Large corporations and governments both have a long track record of entering into energy procurement contracts for solutions with attributes that they have deemed highly desirable: like low-carbon, rapidly deliverable, and/or resilient. Some of these entities have started to strike above-market offtake contracts and guarantees in order to bring those solutions to market faster: as have some hyperscalers with nuclear, Amazon with electric delivery vans, NYSERDA with offshore wind, or the US Air Force with enhanced geothermal.
- Supply-side vehicles. This is where a development consortium, public or private, might be formed to deliver the new solutions across multiple projects and geographies, immediately delivering sufficient scale to suppliers while also amortizing some of the deployment risk. As an example, the UK government formed Great British [Energy -] Nuclear as a publicly-funded, but arm’s length, national development company of small modular reactor projects. It has been charged with performing activities like finding sites around the country, raising capital, selecting partners, and facilitating construction. As part of partner selection, it ran a fairly competitive solicitation inviting technology vendors from around the world to showcase their offerings and enter into contention to deploy across a range of sites. It ultimately selected a vendor, Rolls-Royce, as its preferred partner and is moving forward with project development. It is leveraging the power of its sovereign backing to derisk the projects, navigate regulatory hurdles, attract capital, and internalize the societal benefits of the investment (job creation, manufacturing, etc).
- Joint supply-demand venture. Here, customers and developers tightly collaborate to deploy new solutions. A version of this was created with Amazon & X-energy’s partnership to deploy small modular reactors. Here, building off X-energy’s DOE Advanced Reactor Demonstration Program-supported deployment grant and partnership with Dow Chemical (serving as both industrial offtaker and infrastructure delivery expert partner), Amazon not only committed to purchase the energy from 5 GW of SMRs across multiple sites to meet their needs, but also invested in the company itself, meaning they are more aligned and get to share in X-energy’s future success. In turn, this could be particularly meaningful as it represents the first close representation of a project ‘orderbook’ for nuclear in the US in recent years, and thereby provides an opportunity to standardize their product (particularly given the same customer and same vendor), reduce future execution risk, and accelerate cost reductions through improved learning rates. Separately, but with some parallels, the U.S. DOE’s Gen III+ SMR program, created in 2024 and awarded in 2025, also prioritized applications that formed deployment consortia.
All of them offer significant advantages over pursuing projects on an individual basis. They provide demand signals to supply chains to create manufacturing capacity and to labor groups to create a workforce. Generally, both of these stakeholders may need to see firm demand signals before they will undertake significant investments, which are in turn typically needed to reach a solution’s cost and performance entitlement (otherwise creating a chicken-and-egg problem). They also create more concrete opportunities to drive project standardization; this not only allows a technology to achieve faster learning curves, but it also helps to derisk and justify the investment by providing a more tangible line of sight to the large market. The point for manufacturers and workforce development groups is equally applicable for financiers, who often want to see a pipeline of repeatable opportunities before spinning up their underwriting teams.
Risk and reward sharing
Having an orderbook of the first several projects, per se, may not be sufficient to create sufficient activation energy at the project level. Though it sends a good signal for supply chains and others, it does not necessarily address the first-mover disadvantage issue that may exist.
A differentiator in partnership approaches, including the ones described previously, is how to think about alignment and value creation. Traditionally, the way in which governmental entities approach financial partnerships is through mechanisms like subsidies, loan guarantees, offtake guarantees, backstops, and fast-tracked processes. These help to reduce financial exposure to stakeholders, but alone, they miss a key part of the story: the long-term upside that can be created by successfully deploying and opening up a market for these solutions.
Usually, for the product owners and corporate investors, this is more naturally accounted for and balanced against the downside risk. For instance, companies, from software providers to aircraft manufacturers, might sell their first products as loss-leaders, providing lower pricing to early adopters to reduce the risk to the buyer, which they justify knowing that they should be able to rapidly recoup the costs of early expenses (and even failures) over the broader market pool, if they are successful. For large infrastructure projects, this is more challenging to achieve. When projects are highly capital-intensive, the financial exposures may be too great for the product company and/or equity investor to bear – that company or fund might be entirely wiped out by an individual project failure (for example, Westinghouse had to declare bankruptcy in 2017 when their two U.S. nuclear projects faced challenges). Project stakeholders and financiers might be asymmetrically exposed to the downside risk and, therefore, be inclined to avoid investing in early projects. For a promising technology, you may often find several customers (e.g., electric utilities) lining up for the ‘third’ or ‘nth-of-a-kind’ project, which would likely be derisked and less expensive, while at the same time taking a passive wait-and-see approach with the first project – which produce a stalemate if the first project is hard to get off the ground.
This is where deployment partnerships with structures that more fully align economic incentives and share in both the downside risk and upside of value creation can be a powerful catalyst for action. Amazon’s structure, for instance, creates this too by being involved in multiple projects where they would ultimately benefit from improvements over time and also through their equity investment into X-energy itself, so they should (depending on terms) continue to benefit down the line if the company is successful.
In addition to encouraging the formation of joint ventures and consortia as described earlier, states and/or national governments can work together to strategically invest in key solutions, run competitive tenders to prospective providers, and strike profits-sharing agreements and/or warrants (as opposed to pure equity) in situations where government investment played an outsized role in value creation.
A potential example of this is the recent $80 billion framework agreement between Brookfield and the US Government to deploy new large nuclear reactors. Notably, beyond packaging existing products and authorities (e.g., low-interest government loans for projects), this proposed deal additionally stipulates a proposed profits sharing mechanism where the US Government would receive a share of future profits from reactor sales. Noting that this partnership is early-stage and important details have yet to be disclosed, this mechanism could be appropriate here as you have a hard-to-commercialize sector, with strategic national and geopolitical value, with effectively no domestic competitive products.
State entrepreneurship
Public sector funders can also play a significant role in creating and incentivizing these kinds of deployment partnerships. Though more commonplace in countries with state-run industries, countries with free market economies have often found it more delicate to navigate. There may be legitimate concerns about how governments’ “picking winners” may create adverse incentives and undermine competitive markets, in some cases. Plus, it may confuse governments’ role of trying to maximize public benefit, versus showing favoritism or extracting economic rents from corporations.
All that said, there are models for state entrepreneurship that can be very powerful here, balancing the need to pull forward solutions and capital, while protecting the public and maintaining market competition – particularly in markets that are pre-commercial, have few players, have outsized national strategic benefits, and otherwise would not develop on their own without heavy external intervention. These are cases where, though the market benefits are considerable, the activation energy may be too high to stimulate deployment without deep governmental intervention.
Furthermore, consider where you have first-mover disadvantage challenges but a strong set of prospective fast-followers. To avoid the risk of a Spiderman-meme-like situation where stakeholders (e.g. local utilities or individual states) are pointing at each other to make the first move, downstream project investors could, for instance, co-invest (debt, equity, or backstops) for the first project, even on a minority basis, which would both mitigate risk on the first project and also enable them to access a cost-effective pathway to the technologies they desire to build down the line. You do not traditionally see state and local entities investing in infrastructure projects in other jurisdictions, but doing so could be net beneficial as a faster and lower-cost way to derisk and execute their own projects.
In addition, where appropriate, public financing entities investing domestically (e.g., states, green banks, federal agencies) could, where appropriate, consider extending their authorities to borrow some concepts from the US’s international playbook. Organizations like the Development Finance Corporation (DFC) can make equity investments in strategic, high-value projects, particularly where the normal capital markets would otherwise struggle to enter until the investment thesis is more clearly actionable. Such a process would have very clear scopes, firm guardrails, clear commercial competition plans, and be compatible with legal and market structures to create the intended benefits without confusing or distorting markets.
In any scenario, there should be a corresponding plan for how the public profits would be used. For example, it could be used to raise capital for other governmental activities or directly returned to the public in some way. Or it could be efficient to recycle the funding towards related activities and balancing of the governmental ‘venture capital’ portfolio, as would a strategic sovereign wealth fund.
Chapter 3. “Highway to the deployment zone”: Faster, risk-weighted transaction execution
There’s a common cliché from finance that time kills or wounds all deals. Increasing the speed of policy formation and deal execution is essential to unlocking growth and investment, especially for newer sectors. We will particularly focus on the public capital side of the equation. Here, there is a great mismatch between public and private investment decision time scales. In the private sector, deals are expected to be completed on the order of months or even weeks. Often in the public sector, this can add many months to years, with a high degree of uncertainty, depending on the program. There can be many idiosyncrasies associated with public funding – e.g. infrastructure projects with federal dollars may have additional compliance requirements (e.g. for environmental regulation or for domestic manufacturing). Though there are many deep policy questions here, for this discussion, this piece will focus on ways to accelerate the process.
Staffing for success
While it’s easy to say that the government should move faster, the reality for many is that the individual government program officers are typically working at a rapid pace. This is especially true at the political level where the motivation to make progress in a short amount of time tends to be very high. Particularly, when it comes to developing new programs, they do a massive amount of work, mostly unseen by the public, and with very few resources and are often overstretching to meet deliverables.
The other side of this is that when new programs and initiatives are rolled out, there often isn’t a similar level of flexibility in staffing levels and allocations. In fact, staffers at the federal, state, and city level can get overwhelmed by the volume of direct work and information requests, following the relevant laws and statutes. A new capital program may get introduced, but the number of people to implement that program might not change rapidly. For instance, the Inflation Reduction Act of 2022 introduced and/or influenced dozens of tax credit programs, and accordingly had to issue almost a hundred pieces of new guidance, so the market could act on them. A relatively small group of people led by the Treasury Department’s Office of Tax Policy were charged with generating that official guidance (as required, to ensure consistency and fairness). In addition, a number of the programs therein had complex elements which required deep technical expertise (e.g. tax law, energy markets, carbon accounting, energy technology) to complete their work well – skills that are in relatively short supply and in high demand, both inside and outside the government. The rate of progress was also slowed by some ambiguity in the law itself, where key technical questions (e.g. accounting methodologies and criteria) accordingly needed additional time to be addressed during implementation instead of beforehand. The associated teams ran at breakneck pace to complete all those issuances in just two years. Yet, many engaged market actors who were excited to proceed with shovel-ready projects experienced challenges, as they waited for guidance, thereby slowing initial progress.
To meet the rapid needs of an eager market and particularly in times when the governments are trying to push comprehensive reforms, agency leaders and legislators need to consider ways to make sure that the implementing organizations are sufficiently well-staffed and resourced. This should not only consider program staff (both existing and new), but also functional teams (e.g. legal, communications, stakeholder engagement) where bottlenecks can often form as they support multiple programs. This can include having surge capacity resources (either short and/or long-term, internal or external) bringing on technical and subject matter experts to help enable fast and fair processing. Also, an implementation staffing needs assessment should ideally be conducted as part of the policy-formation & legislative process so that the appropriate resources can be allocated early and efficiently. To further ensure efficiency of resource allocation and implementation speed, legislators should consider expedient ways to drive greater clarity and specificity at the point of legislation, where applicable.
Iterative capital deployment programs
It is tempting and important to get things totally right on the first try, particularly when it comes to government funding, which is highly scrutinized. That said, an approach which has delivered success is to release capital in phases. Here, instead of issuing all funding at one time, the program office (especially for a large competitive program) might split the deployment into phases over time. The first phase is executed quickly and the subsequent phases are introduced later. While this may introduce some short-term friction, it not only gets capital and projects moving faster, perhaps as importantly, it gives both the funders and market chances to build momentum, learn from one round, and improve towards the next ones. A good example of this was the DOE’s Grid Resilience and Innovation Partnerships (“GRIP”) program. This was a $10 billion program from the Bipartisan Infrastructure Law to enhance grid flexibility and improve the resilience of the power system against extreme weather. That $10 billion allocation was split into three phases to be issued over a few years. Over those three phases, the quality and ambition of the applications and programs funded increased significantly as all stakeholders were able to learn and adapt with each round.
Progress over perfection
For public programs, this is a major challenge driven by misalignments of risk tolerance. Many ambitious government funding programs are faced with a strange pickle. On one end, they have the duty, mandate, and power to drive innovation, do deals ahead of the commercial markets, and derisk promising solutions to a point where they can scale on their own and deliver the broad public benefits associated with that solution. On the other hand, the funds being used are raised from people’s hard-earned money or a state or country’s valuable natural resources, and neither should be handled frivolously. The fear of the political fallout from the latter may drown out the benefits of delivering the projects; that is, in the eyes of an underwriter or program officer, the downside risk may often outweigh the upside creation; doing no deal may feel safer than doing a ‘bad’ deal.
Take the case of two companies that received funding from the DOE’s Loan Programs Office (LPO) in the early 2010s. One is the solar cell manufacturer Solyndra, whose idea was to decrease the cost of solar energy by using cylindrical, thin-film solar cells that could capture the sun’s energy from multiple angles, compared to their conventional flat-panel counterparts, and thereby bring down their levelized costs. Solyndra received $535 million in federal loan guarantees, which it defaulted on and went bankrupt when market conditions changed (they, and other promising new solar companies, were undercut by plummeting prices of silicon solar cells from China). The default filled the news cycles for months, sparked several congressional hearings and investigations, and also left a profound imprint in the minds of many program funders. No government underwriter wants to be dragged to the Hill or see their name in the papers for this reason. On the other end of the spectrum, you have a little-known car company called Tesla, which received a $465 million loan to expand electric vehicle production. Tesla, as we know, went on to become one of the most transformational and successful automobile companies in recent history. And yet, comparatively little fanfare has been made about the government’s role in making that a success. Two loans, about the same size, issued around the same time, by the same organization. Not only did their actual outcomes differ, but the financial outcomes to the upside and the political fallout to the downside were diametrically-opposed.
This contrast becomes even more stark when looking at the broader picture. Again, taking the example of the LPO (now Office of Energy Dominance Financing (EDF)). It has historically had a loss rate of less than 3%, which is on par with most commercial and investment banks – entities which often invest in markets with more proven solutions and less uncertainty. Moreover, other governmental programs, like DARPA, NIH, and ARPA-E, also have strong investment track records. All that said, the perception of risk has led to an overly deep risk conservatism and a sense of fear of doing deals that might go sideways. This creates huge process drag for the entire organization and curbs the rate of progress that can be made, as underwriting processes can become elongated and difficult to navigate. Furthermore, for many loan applicants, it can take several years to get through the loan process; anecdotally, some applicants have complained it was slower and harder to access than what they could wrestle from the commercial market.
Overall, this is a situation where the tolerance for and understanding of losses for public financing needs to be reconceptualized and appropriately balanced, given the mission. Not all losses are bad. Individual losses are not necessarily detrimental if outweighed by net gains. There are significant opportunity costs of not taking risks appropriately. More can be accomplished without jeopardizing public interest.
Given the governments’ having historically demonstrated their ability to be good stewards of capital across long periods of time, the inherent risk that accompanies the pre-commercial asset classes they support, and the urgent need to make progress and unlock markets, this is a situation where more streamlined, faster underwriting processes that increase the speed of execution are critical and warranted. Furthermore, governmental funding organizations need to be given more ‘air-cover’ so that individual misses do not get over-politicized, but are understood to be reasonable elements of a process for progress. Note that accomplishing this process and cultural shift requires major work internally with staff, policymakers, and the broader public. This is an aspect where integrating concepts like state entrepreneurism and more balanced portfolio-based risk and reward approaches, described previously, can also unlock new investment and risk management strategies, greater societal benefits, and increased comfort to staff and leaders to usher in that kind of transformation.
Creating longer and more durable windows of action
A more obvious reason to move quickly on policymaking is to deliver benefits faster to project and community stakeholders – which is, of course, the main objective of the policy in the first place. But beyond that, investors understand that the political time windows covering favorable conditions could be short. This is particularly acute for assets with long development cycles and/or high upfront costs: e.g., building new manufacturing facilities or developing interregional transmission lines can take years. Indeed, it was estimated that 60% of committed IRA-funded clean energy manufacturing projects were originally not slated to come online until between 2025 and 2028.
Moreover, the IRA timeline created some interesting time crunches. Though the law was thoughtfully conceived with some longer time horizons for tax credits, in practice, the actionable investment window for that version ended up being incredibly short. The law was passed in August 2022. It then took time for programs to be formed and guidance to be released, as described earlier. In parallel, the investment community had to learn, come up the learning curve on the new opportunities, and build ecosystem collaborators (which themselves are also reacting and forming). Next thing you knew, as the election window started to ramp and policy uncertainty increased, many investors started to park their capital and take a wait-and-see approach in early 2024, as evidenced by strong increases in fund ‘dry powder’ (raised but uncommitted capital) but a sharp dropoff in actual capital deployment and assets under management at that same time.
Moving quickly is critical to give investors, communities, and other associated stakeholders as much time as possible to understand the landscape, develop deployment pathways, build new solutions, and ideally iterate, given the chance to take more shots.
That was a shorter-term perspective. In the spirit of leveraging speed to open the front end of the window, longer term, some thought should be made to how one extends the investability window. Investors typically do not decide to invest in a project purely due to the project’s merits. Particularly when entering a new sector, it’s also driven by the commercial prospect of potential follow-on deals. Short political windows and the associated ‘stroke-of-pen’ risks often raise major flags for the risk committees at financial institutions. As was mentioned earlier, many IRA programs arguably had much less than two years of impact. Deeper policy stability is critical to ensure continued, long-term investment. That type of stability has, at least historically, been a hallmark of the US regulatory & commercial system and a positive differentiator in the race to attract capital and talent from across the globe. For sectors with high strategic value, high early capital requirements, and long investment cycles, policymakers should consider more mechanisms to provide longer-term policy guarantees to give investors assurance that they have long enough windows to justify their business cases.
Chapter 4. “Okay, now let’s get in formation”: programmatic policy synchronization for fast market formation
Catalyzing the deployment of new infrastructure is usually enabled by a bevy of policy actions. This can be important as transforming a sector may require several changes in economics, behaviors, and processes. Especially when resulting from expansive new legislation and/or executive actions, the government may be required to deliver a host of new policy programs, including new rules (e.g., permitting reforms, categorical exemptions), funding allocations and programs, implementation guidance (e.g., for tax rules), informational reports (e.g., National Lab technical studies, commercialization reports, etc). These types of activities are highly valuable as they tackle different aspects of the deployment challenge and take huge amounts of effort to be effective. However, they often get rolled out and implemented on separate and independent processes. This can actually stall and frustrate deployment efforts as most investors will want to see the major policy puzzle pieces locked in place before getting comfortable enough to deploy capital – for most risk organizations, ‘stroke-of-pen’ risks are often viewed as red flags. Conversely, this hesitation can cause consternation for policymakers and advocates who may feel that they have done the heavy lifting in passing new legislation, but don’t see a corresponding flood of serious commitments immediately after.
Policy deliverable schedule alignment
One way to address this is to implement a visible and synchronized schedule, showing all the related policy efforts and programs for an initiative. The interdependencies between those activities would be easier to identify and allow relevant stakeholders to see when all the major puzzle pieces would be in place, and in turn, also align their investment and advocacy efforts accordingly.
An example of this comes from carbon capture: the federal tax credit for carbon dioxide sequestration (45Q) was first issued in 2008; however, the first set of tax guidance was not issued until 2020, as the IRS, Treasury, EPA, and other agencies had to build a suite of complex regulations around reporting, verification, stakeholder comments, and more. A consequence is that, though the tax credit was in place (and though there were strong complementary financing capabilities from renewables tax equity and thermal power plant development already in place), little to no investment went into this sector, effectively ‘wasting’ many years of eligibility and frustrating many interested stakeholders.
By contrast, take advanced transmission technologies: despite being rapidly-deployable and cost-effective solutions to increase transmission and distribution system capacity and performance, they have been historically underutilized. To increase awareness and deployment, the federal government developed a suite of products including the formation of the Federal-State Modern Grid Deployment Initiative, grant funding via the DOE’s Grid Resilience Innovation Partnerships program, loan funding from the LPO’s Energy Infrastructure Reinvestment program, categorical exemptions in federal environmental permitting on upgrading existing transmission lines, a Pathways to Commercial Liftoff report on grid modernization, new national deployment goals, technical reports and new assistance programs by the National Labs, and more. These were all released within a couple of months of each other in 2024, so the market had a fuller picture to which it could react and begin to make greater progress. Since then, dozens of states have passed new laws, and the number of projects being pursued and funded has also been on the rise.
Capital source navigators
Relatedly, new legislation may result in several new governmental funding programs or changes in missions for existing funding programs. Many of these efforts and changes might go unnoticed or disproportionately utilized. Take the energy- and climate-tech startups, who may be seeking capital to grow or transform their businesses. Government capital tends to be attractive as it is often willing to embrace early technology risk (unlike most commercial capital), is often non-dilutive to the company’s capital stack, and can give the company extra visibility. Most people in the energy sector will know of programs like DOE’s ARPA-E (Advanced Research Program Agency for Energy) or the Loan Programs Office. Far fewer may know that there may be funding available through ‘non-energy’ agencies like the US Department of Agriculture, Small Business Administration (SBA), the Governmental Services Administration (GSA), or the Department of Defense (DOD). These have increased the pool of capital available and provided a wider array of financing products that increase the chances that the right kinds of capital are available to serve the spectrum of company needs.
Initiatives like the Climate Capital Guidebook, published in 2024, can be helpful to make these types of programs less opaque and easier to access, especially for startups and small businesses. At the state level, databases like DSIRE USA have been providing a beneficial service aggregating information on state incentive programs for years.
Making information on federal, state, and/or municipal funding programs highly accessible and searchable from a centralized, common location is key. Or else they may get lost, buried in webpages that few know where to access. This process can be further enhanced using cross-cutting discovery technological tools. For example, AI-based agents could be used to continuously and automatically map these programs and keep information organized. Also, large language models can be deployed to allow stakeholders to more readily identify and compare programs of best fit (matching things like user capital needs versus program ‘ticket sizes’, usage restrictions, and eligibility requirements).
Zooming out from individual needs, this would also augment new solution developers’ and investors’ ability to more comprehensively understand the relationship between governmental capital programs and the role they play in energy solution commercialization and deployment. For instance, related to technology readiness, it would make it easier to chart what programs are available to technologies at different stages of maturity. From the National Science Foundation (NSF) for fundamental research, or the Advanced Research Program Agency-Energy (ARPA-E) for more applied technology development and early manufacturability demonstration, to planning grants from various agencies, and federal tax credits for infrastructure projects.
Similarly, funding programs could be mapped against project development phases. In areas like international project finance, this exercise would be valuable to demystify which programs and institutions are suitable for various phases of project development. In some cases, an international energy project developer using American technology might need to navigate a gauntlet of different funding institutions: from the US Trade and Development Association (USTDA) provide grants for front-end engineering development (FEED) studies, the Export-Import (EXIM) bank for domestic manufacturing loans, to the Development Finance Corporation (DFC) for equity co-investment and political risk insurance. Not to mention multilateral development banks like the World Bank and International Finance Corporation (IFC), which themselves have an array of funding programs and instruments. Here, providing clearer, more cohesive representations of how a patchwork of funding sources can work in tandem and be packaged together can have outsized strategic competitiveness for American companies. This would thereby help level the playing field for companies competing against companies backed by governments that can provide fully wrapped financing solutions.
Chapter 5. “C.R.E.A.M.”: More holistic valuation tools and methodologies
This facet addresses the challenge, which is still too common in solution valuation. Not of valuing the companies themselves, but of proving to customers and investors that the proposed energy solution is worth adopting. Especially because regulation alone is usually not a salve for driving energy transition activities in free market economies. While regulation may steer what should happen, costs and economics are often bigger drivers of how quickly that transition occurs. Borrowing a chemistry analogy, economics determines the activation energy and kinetics of transition policy. Solutions need to demonstrate their fit and attractiveness in often economically competitive and constrained environments. In addition, stakeholders with shared interests (e.g., not just federal, but also at the industry and state & city levels) should invest to build a common valuation infrastructure (e.g., resource characterization data, system models, and more) that helps lower the barriers to deployment and investment. Doing so will also make it easier to appropriately size any associated financial programs (like subsidies and grants), to ensure that there is sufficient catalysis to get multiple stakeholder groups moving and investing.
Understanding end-use unit economics
This means that solution providers, policymakers, and advocates need to develop very deep understandings of the commercial drivers and realities of the markets they are looking to serve. They need to put themselves in the shoes of their customers and related stakeholders. This is particularly important when trying to sell solutions into new and competing markets or applications that may not otherwise be required to change (e.g. regulations on fuel use or emissions). This should seem obvious and has always needed to have been a primary focus, yet it’s a step that some innovators, policymakers, and advocates have not always adequately prioritized.
This is a recipe for failure, particularly in the infrastructure space. Saying it’s good for the world is not sufficient to get traction; a need does not mean there’s a market. Getting a strong, detailed, and accurate understanding of customer unit economics is foundational to the success of any infrastructure. By their nature, this should encompass more rigorous estimations of how much a solution costs to produce and deliver (which often tends to be underestimated in early stages, and leave stakeholders to be surprised later on by cost overruns upon implementation). And it should likewise reflect an understanding of the customer’s cost and value drivers, as this affects project revenue and adoption readiness. This is a question that sometimes gets missed in the early stages, but in later stages, particularly when seeking significant capital to fund projects, it becomes highly pertinent as investors tend to take a much more critical viewpoint of the economic potential of the project, both to the upside and downside. As part of the process of developing detailed assumptions, they should develop reasonable sensitivities and scenarios that illustrate how the financial performance of the project may vary due to changing internal and macroeconomic conditions.
Diagnosing this early not only helps the solution providers to be better positioned for commercial success with their customers but also enables them to catch potential flaws early enough, make different design choices, and ensure the product’s value proposition is more robust and resilient. In turn, this helps reduce project risk and gives comfort to financial investors along the way.
Governments can help by driving easier price discovery and transparency, collaborating with project stakeholders (especially developers and customers) to compile and share relevant cost and value data in more public forums. Reports generated by government agencies (e.g., the series of Pathways to Commercial Liftoff reports by the US Department of Energy’s Office of Technology Transition/Commercialization), national laboratories, and private third-party analysts (e.g., BNEF, S&P, Lazard) have made strong contributions in attempting to fill those information gaps. Continuing to support and drive efforts like those would be valuable. Having state and regional actors (e.g., groups of state economic development organizations) can also help to make them even more granular, and perhaps more local, which would drive even more actionability. They could also compel information disclosure through legislation (akin to efforts in healthcare and drug pricing transparency over the past few years) or require a greater level of disclosure as a part of some government-funded programs, especially with new industries.
Development of accessible, trustworthy technoeconomic evaluation analysis tools
Intending to perform the types of deep technoeconomic analyses described in the previous paragraph is one thing. But having the ability to do that is another. In many situations, you either suffer from data unavailability or asymmetric access. Taking the electricity system, for example, there are very few stakeholders (usually utilities and grid operators) with deep access to information about how the system and its underlying assets are performing. Sometimes, this is intentional due to potential concerns around security and market manipulation. Sometimes, like often in the case of customers requesting their own historical hourly usage data, the process might be archaic and difficult.
That said, a major downside of this situation is that third parties are often not in a position to interrogate resource plans, challenge priorities, or test and validate new ideas. There are third-party analytics tools, but they sometimes don’t have the requisite data, fidelity, or trust to ensure that their results cannot be readily dismissed. That also makes it too easy for grid operators to dismiss new ideas without adequately considering them. Not having accessible data and models can result in excluding beneficial solutions from being part of the menu of options or from making it to market.
This has been a pretty common battle with an array of more ‘disruptive’ energy technologies, like distributed energy systems and advanced transmission technologies. But it also occurs with generation. One example is the prospective transition of the Brandon Shores coal plant in Maryland. The plant’s owner, Talen Energy, filed to retire the facility because it was no longer economically viable to operate (following a trend with many coal plants around the country). However, the grid operator, PJM, sought to force an extension of its operation (via striking a reliability must run (RMR) contract) for four years until new transmission capacity can be brought in, citing potential reliability concerns. State and congressional officials strongly opposed this plan as extending the operation via an RMR contract would increase costs to ratepayers and increase local pollution. A group of advocates and energy experts, led by the Sierra Club and GridLab, proposed replacing the plant with a mix of energy storage, reconductoring, and voltage supports, which they claimed would be not just cleaner, but more cost-effective than what was proposed – even more so in the likely event that the transmission project is delayed. Note too that a similar concept was deployed in New York City at the Ravenswood power plant. However, that suggestion was dismissed by PJM, without real allowance for iteration, arguably for not using the right modeling methodology, and for not being a project sponsor. Aside, the decision is also reflective of PJM’s energy storage market structure’s inability to effectively value energy storage’s benefits as both energy and transmission assets. Though an agreement was ultimately reached, many stakeholders view this settlement as suboptimal, not just because of its outcome (even more so as it does not help wider issues like high-capacity market prices), but because the advocates did not have the modeling tools in place to meaningfully evaluate the options and force a more substantive dialogue with the grid operator.
Rebalancing this situation is essential in order to allow additional key stakeholders like policymakers, regulators, project developers, solution providers, and other experts to interrogate the opportunity space and propose actionable new ideas. This should include creating common, accessible infrastructure for data, models, and evaluation methodologies that an interested stakeholder would need to know to assess potential project options – which are otherwise difficult or prohibitively expensive to access. Note too that government endorsement of these tools greatly assists the credibility of the prospective solution.
The Australia Energy Market Operator (AEMO), for example, did exactly this by implementing the world’s first connection simulation tool. This tool is a digital twin of the country’s electric grid that project developers are using to rapidly evaluate their prospective solutions in an accurate, safe, and trustworthy environment.
Also consider two approaches to accelerate geothermal project development, where insufficient quantification of the resource can add significant development costs and project risks. One example is Project InnerSpace: this is a collaborative effort funded by philanthropy, the US federal government, and Google to provide a common, open set of surface and subsurface characterization data. Another example is the Geothermal Development Company: this is a special purpose vehicle fully-owned by the Kenyan government, which performs the resource characterization and steam development themselves, and shares the information with prospective geothermal power producers, which in turn has significantly lowered the barriers to entry and made Kenya a global leader in geothermal power production. Both approaches are being used to help project investors to be more targeted, capital-efficient, and prolific.
Similarly, Virginia recently passed a grid utilization law requiring their local utilities to measure the utilization of their transmission and distribution systems, including establishing metrics as well as plans to improve those metrics. If implemented well and the associated data is made available, having that data can drive more targeted investments, more cost-effectively manage customer energy bills, and allow new solutions like virtual power plants, distributed energy, and grid-enhancing technologies to be appropriately valued and play bigger roles in the energy solution mix.
Quantify, aggregate, and internalize external benefits and costs
Many solutions labeled for “climate” often have a wide array of other benefits – lowering costs, boosting reliability, creating jobs, improving health, to name a few. In some cases, reducing emissions might be a secondary or even tertiary benefit. This often results in the cost-benefit of a potential solution being understated and capital being underallocated. Alternatively, it can lead to greenwashing, where the benefits are overstated relative to the impact and capital being misallocated.
In some cases, like in electricity markets and industrials, decisions are made on a narrower set of financial criteria, ignoring the broader value proposition – e.g., which solution has the lowest upfront cost? What is the least costly way to meet power demand on hourly basis or does the solution pay itself back in three years? There have been some directional approaches that at least help with the first issue of benefit underquantification.
An example is with FERC Order 1920, a rulemaking that covers new approaches to transmission planning and cost allocation. It called on state decision criteria to be expanded from just cost and reliability to a consideration of seven benefits: avoided or deferred infrastructure costs, reduced power outages, reduced production costs, reduced energy losses, reduced congestion, mitigated extreme weather impact, and reduced peak capacity costs. As it gets implemented across the country, that should provide a considerably more fair and holistic basis to assess the potential benefits of transmission projects and will likely increase the viability of game-changing concepts (e.g., reconductoring, interstate/interregional transmission).
In other cases, like in some larger governmental grantmaking or policy efforts, a suite of benefits may be quantified but are estimated and presented in siloes, where the benefits appear to be orthogonal and nonadditive. So, the key to addressing that is developing valuation frameworks that do the difficult task of weighing the benefits together in a clear manner that’s directly relatable to the investment thesis. Applying ways to translate those benefits commensurately into the project’s financial terms is critical to ensure they get prioritized and realized. The IRA had elements of this at least conceptually, for instance, by applying bonus credits to low-carbon energy projects that paid fair wages, were put in economically disadvantaged areas, or used domestically-manufactured equipment. On a local level, there are state laws like Montana’s transmission law that established a new, elevated cost-recovery mechanism for transmission projects that used more efficient, high-performance conductors.
Going further along the point of internalization, there are more structural issues where markets may not be designed to solve for the outcomes that stakeholders are seeking. In electricity, for example, power markets generally solve for meeting demand for the least cost in a short time period (following a narrowly-defined reliability scheme). This may not only ignore solutions that may save more money over longer periods of time, but it also does not explicitly solve for attributes like resilience, sustainability, or flexibility. That often means external out-of-market solutions are needed to create desired outcomes (e.g., reliability-must-run contracts, tax credits, renewable energy credits). While those have had a great impact on their specific goals, they are imperfect and may have unintended consequences, like distorting market behavior or disincentivizing cost-cutting innovations. Solving that on a greater scale and more fundamental basis in some segments may require greater reforms, like redesigning electricity market structures, revisiting the Energy Policy Act and Federal Power Act, and more.
Chapter 6. “Take it to the bridge”: Rethinking the ‘missing middle’ problem
For the last face of the cube, it is incumbent to describe the role that a whole cadre of investors needs to play. Not just venture and early stage, but particularly later stage capital providers such as project financiers (equity and debt), institutional investors, pension funds, insurance funds, and even utility balance sheets. They have a massively important role and arguably need to be more proactively involved with ensuring the maturation of promising earlier-stage solutions. The ‘missing middle’ problem in energy, where a lack of transition and demonstration capital thwarts promising venture-backed solutions from progressing to mainstream infrastructure, is well-known. While continued innovation is needed to form new capital solutions to fill that gap, there is a lot that investors can do to shrink the gap and make that chasm easier to traverse.
Engaging earlier to pull companies to maturity
Though they control the greatest share of the majority of assets under management and can support bigger ticket sizes, by the nature of their investment mandates, late-stage investors’ risk tolerances tend to skew more conservative. They tend to concentrate their efforts on solutions with established track records and large addressable markets that provide greater certainty of execution and relatively consistent returns. And they typically have more than enough deal volume to justify their focus in those markets. Consequently, though these investors typically at least follow major new trends, they tend to be hesitant to enter newer markets; in fact, many are content to just ‘wait’ for the market to come to them before they engage. This introduces several challenges.
First, at the most basic level, it means that many lower-cost sources of capital may be hard to access for newer solutions, whether climate and otherwise, which makes it more challenging for them to compete on a level playing field early on. Next, as importantly, it means that companies may miss critical opportunities to get sharper earlier. The attributes that are valued by investors change dramatically over the life cycle.
For instance, many early-stage products are rated by things like their uniqueness, differentiation, disruptiveness, and total addressable market. Those attributes that tend to attract venture capital and garner the most market visibility in the media. By contrast, at later stages, especially for project uniqueness and differentiation, might actually be seen as sources of risk: risks that get compounded if the solution is supplied by a new market entrant. Moreover, fungibility and supplier optionality may hold even greater weight. To mitigate the risk of a situation where things could go wrong with a project’s vendor, project investors are often comforted knowing there are substitutes that can be brought in as part of a contingency plan. In addition, though addressable market matters to both groups of investors, early-stage investment tends to focus on alignment with macroeconomic trends. While for later stages, micro arguably supersedes macros, as diligence tends to be more deeply and narrowly-focused on project-specific questions, like contracts, pricing, and execution. Furthermore, late-stage underwriters may need to conduct deeper diligence and acquire more data to get comfortable with the new technical attributes, features, and vendors, which can be a drag on their process as they have to spend more time and money to go through that process. Whereas for earlier stage investors, that deep focus on the new features already tends to be an integral part of the diligence and value creation process, and they get rewarded for that accordingly.
Overall, not understanding these differences can create shocks for new companies that have had great success with attracting capital early, but get stopped in their tracks when graduating to the next level of maturity. This has also often meant that prospective solutions providers miss the opportunity to sharpen their pencils, address more detailed questions, and have their key assumptions stress-tested. Even if they are not prepared to transact, later-stage investors, especially in infrastructure project finance, and their partners (such as independent engineering firms and insurers) should devote additional time to engage with promising technologies early on and bring them along. This is also in the investors’ interest as it allows them to get up the learning curve faster, be better positioned to take advantage of those opportunities when the markets come around, and ensure that the solutions that do make it to later stages are of higher quality and more likely to yield successful transactions.
Formulate deal templates and archetypes
The previous steer for early engagement comes with a conundrum. For many investors, it is hard to meaningfully engage until there is a complete deal on the table. By complete, I mean a fully-fleshed out representation of all core theses, puzzle pieces, assumptions, and more, mapped to a specific, actionable situation. This is the scaffolding onto which the financing packages are built and the basis by which most risk managers are trained to evaluate the financeability of a solution. This can be true for governmental and commercial financing programs; “bring us deals to look at” is a common refrain.
The conundrum comes because many of those details may not be fully known early on, so there may not be a fully-formed deal to bring per se, and in addition, there might not be a clear set of underwriting criteria that the company can aim for. Even for offices like the DOE LPO, which was proactive and engaged, struggled to get significant market traction for a few years, both to their and the market’s frustration. Instead of both sides staring at each other like the Spider-man meme, the impasse can be broken by creating deal templates and archetypes, which can take a more hypothetical representation of assumptions, including reasonable scenarios, and frame what the financial structure and execution pathway would look like for that.
The LPO, for instance, did just that, creating several deal archetypes based on customer type and technology, with terms and execution timing aligned based on the associated risk. For instance, deals involving more established energy solutions (e.g., solar, storage, transmission) with investment-grade utilities providing corporate guarantees were allowed to have faster execution processes than some other deals, commensurate with the comparatively low level of credit risk involved. Next, deals associated with a narrow focus, like the Advanced Technology Vehicle Manufacturing program, tended to have more well-defined execution processes. For others that may have more default risk (e.g., newer technologies, non-investment grade counterparties), those deals might take more time to diligence and underwrite. This benefitted both the Office and the applicants, as this created clear, agreeable expectations for each. Providing this clarity greatly increased deal volume and traction, as more clients brought more loan applications and had greater clarity and confidence in the transaction process they were entering into.
More broadly, this is an area where the companies, industry associations, and other advocates can play an active role, independently and in collaboration with governments. Formulating early pictures and archetypes for financial stakeholders and investors can significantly enhance feedback and capital formation.
Collaborate, Celebrate, and Replicate
Finally, energy investors, especially in less mature sectors, need to find ways to be more open, as appropriate, about their investments and investment strategy. Though for product companies, this usually comes a bit more naturally and is necessary to market their products, later-stage investor communications for individual deals often tend to be more guarded and high-level. This is usually not because the information is unavailable. Sometimes this can be hard as they may be attempting to protect sensitive information, or the move may be to protect potential market share and not lead other players on to the same strategy (particularly if they worked hard to open a new market). The richer information transfer usually happens privately during deal execution (e.g., as part of due diligence) or during project- or fund-level capital raises.
In newer spaces, however, progress itself is often catalytic (rising tides floating all boats). The easiest way to persuade a risk committee to invest is to show precedents and comparables. An underwriter can stand up with greater confidence when they can show that someone else has done it before. It can be even more validating when that ‘someone else’ is a competitor. Project investors should endeavor to share more about how they got comfortable with the deals, market, and/or technologies involved, as appropriate. This is not out of altruism. Especially in emerging sectors, rapidly expanding the market and creating a foundational flywheel can be commercially more beneficial for the firm versus purely protecting market share. Getting more investors comfortable makes the pie bigger and encourages other investors to pursue their own projects. This then sends actionable signals to ecosystem stakeholders (e.g. supply chains) to invest and create production and delivery efficiencies. The efficiencies improve unit economics, reduce risk, and increase returns both down the line and potentially even on the early projects (e.g. operating expense and replacement parts scarcity reduction). These scale efficiencies and flywheel creation should help investors generate more deal flow and revenue, building on the expertise they have built and leadership position they have established. The advantages can be extended where significant public funds were allocated to the project, perhaps in exchange for preferable financing terms from the public funding institution.
Similar ideas apply to public sector funding programs. Making announcements about projects and ribbon cuttings, though important, are shorter-term and quick-hitting communications strategies that tend to be more formulaic. Instead, policymakers should take a page from commercial product marketing. Policymakers should view their deployment policy efforts as their products. As such, they should create consistent, thematic narratives to which individual initiatives, projects, legislation, and rules, can all be framed. Even though they may be exhausted after delivering the policy itself, government officials and program officers should not undervalue the uplift phase. It is crucial to plan to spend ample time and resources to explain and repeat the micro- and macro significance of each product to investors and community stakeholders, especially in today’s competitive information environment. Building greater public buy-in both nationally and with communities, is crucial, especially to longer-term, transformational projects. Government funders should work hard to bring along additional local governments, nonprofits, and investors along to collaborate, celebrate, and replicate the successes.
Akin to what was mentioned in the valuation section about common information infrastructure, working closely with investors, industry, and other stakeholders to collate and amplify key investment theses, lessons learned, and others will be key to building investor confidence and creating more of a flywheel effect for follow-on investments.
Conclusion. “The Next Episode”
Taking a step back, we have laid out many ideas and concepts in this paper: harmonization, collaboration, acceleration, synchronization, valuation, and amplification, to name a few. It may seem daunting to look at policymaking across so many vectors, particularly in a manner where many of those puzzle pieces have to align and move in sync in order to unlock significant and consistent investment. That said, the power of a strong and well-intentioned administrative state, at both national and local levels, lies in its very ability to wrap its arms around big challenges, partner with private industry, and leverage its resources to create high-value solutions with outsized benefits. This has repeatedly been proven in the US and globally across time and multiple sectors: whether it’s going to the moon or inventing life-saving medical treatments; building massive infrastructure, or delivering nanoscale electronics. States and towns should roll up their sleeves, find creative ways to collaborate, develop foundational information tools, and remove unnecessary market barriers. Investors should take an even more active role, making their needs known to early-stage companies and policymakers, building consortia to pull new opportunities forward, and creating an actionable set of commercial opportunities that they would find attractive. What’s more, acting now to design and implement new, actionable administrative structures, especially at the state and local level, will not only create more high-value pathways for progress now, but if it is well-coordinated, it can also lay a foundation for federal actions that can be taken, nearer and longer term. Though the challenges and journey are complex, the opportunity before us is massive, the imperatives are clear, the transformations are tractable, and success is achievable. This can be done, so let’s get busy!
Why Credit Access Makes or Breaks Clean Tech Adoption and What Policy Makers Can Do About It
Building Blocks to Make Solutions Stick
For clean energy to reach everyone, government can’t just regulate behavior. It has to actively shape credit markets in partnership with the private sector.
Implications for democratic governance
- Financing programs need governance that is visibly fair, transparent, and accountable to enable trust–without that, low trust drags down their efficacy.
- Build broad constituencies to set and drive the agenda.
- Treat local lenders and communities as active implementers, not passive beneficiaries.
Capacity needs
- Talent, playbooks, and governance structures to run policy-enabled finance (credit, guarantees, revolving funds) with speed and integrity.
- Faster contracting, simpler reporting, and fewer transaction frictions.
- Clear guidance on identifying and resolving tradeoffs, instead of allowing decisions to bog down in case-by-case analysis paralysis.
- Staff who can translate between agencies, investors, and communities.
- Connective tissue to and between states to replicate smart practices and share toolkits for financing mechanisms that move beyond one-time infusions of cash.
- Quasi-public structures that give government agility without sacrificing public interest and accountability.
Access to affordable credit is a necessary condition for an equitable energy transition and an inclusive economy. Markets naturally concentrate capital where risk is low and returns are predictable, leaving low-income communities, rural areas, and smaller projects behind. Well-designed federal policy can change that dynamic by shaping markets—reducing risk, creating incentives, and unlocking private capital so clean technologies reach everyone, everywhere. This paper explores how policy-enabled finance must be part of the toolkit if we are going to drive widespread adoption of clean technologies, and can be summarized as follows:
- Problem: Clean technologies require upfront capital; tax incentives alone are insufficient for small, distributed projects and underresourced borrowers. Without targeted credit solutions, the energy transition will deepen existing economic and environmental inequities.
- Opportunity: Policy‑enabled financial services—direct investments, tax incentives, and loan guarantees—have a proven track record of expanding access to credit and driving inclusive economic growth. The climate policy playbook should be expanded to incorporate lessons from other sectors and programs that have incorporated these interventions.
- Case study: The Greenhouse Gas Reduction Fund (GGRF) was designed to augment grants and tax incentives contained in the Inflation Reduction Act by seeding revolving capital, leveraging national financing hubs, and mobilizing local lenders to scale clean investments. This program was stopped in its tracks early in the Trump administration, but lessons from its design and early implementation should be leveraged by local, state, and future federal programs.
The critical role of policy-enabled finance to drive widespread economic opportunity
Access to affordable credit is not just a financial tool—it is a cornerstone of economic opportunity. It enables families to buy homes, entrepreneurs to launch businesses, and communities to invest in technologies that reduce costs and improve quality of life. Yet, across the United States, access to credit remains deeply uneven. Nearly one in five Americans and entire regions – particularly rural and Tribal communities – are excluded from the financial mainstream, limiting their ability to thrive.
Private-sector financial institutions—banks, private equity firms, and other lenders—are designed to maximize profit. They concentrate on markets where risk is predictable, transaction costs are low, and deals are easy to close. This business model leaves behind borrowers and communities that fall outside these parameters. Without intervention, capital flows toward the familiar and away from the places that need it most.
Public policy can change this dynamic. By creating incentives or mitigating risk, policy can make lending to or investing in underserved markets viable and attractive. These interventions are not distortions — they are strategic investments that unlock economic potential where the market alone cannot, generating economic value and vitality for the direct recipients while yielding positive externalities and public benefit for local communities. And, importantly, these policy interventions act as a critical complement to regulation. Increasing access to credit is often the carrot that can be paired with, or precede, a regulatory stick so that people are not only led to a particular economic intervention, but they are also incentivized and enabled.
For decades, policy-enabled finance has delivered measurable impact through multiple programs and agencies designed to support local financial institutions – regulated and unregulated, depository and non-depository – that are built to drive economic mobility and local growth. These policies and programs have taken multiple forms, but can generally be put in three categories:
- Direct Investments: Programs like the CDFI Fund Financial Assistance awards that provide enterprise grants to Community Development Financial Institutions (CDFIs) to support balance sheet strength and increased lending and the Emergency Capital Investment Program (ECIP) that made equity investments into community development credit unions and banks.
- Tax Credits and Incentives: The Low-Income Housing Tax Credit (LIHTC), New Markets Tax Credit (NMTC), Opportunity Zones, and renewable energy credits like the Investment Tax Credit and Production Tax Credit have spurred billions in private investment for housing, community development, and clean energy.
- Loan Guarantees: Small Business Administration, U.S. Department of Agriculture, and Department of Energy guarantee programs, among others, reduce risk for the lender, enabling small businesses, rural communities, and earlier stage companies to access credit otherwise unavailable at transparent and affordable rates from participating financial institutions.
These tools enjoy broad recognition and bipartisan support because they work. They increase access, availability, and affordability of credit—fueling job creation, housing stability, and economic resilience. Policy-enabled finance is not charity; it is a proven strategy for broad and inclusive economic growth and a key tool for the policy-maker toolkit to support capital investment, project development, and adoption of beneficial technologies in a market-driven context that can increase the effectiveness of a regulatory agenda.
Most importantly, policy-enabled finance has led to major improvements in wealth-building and quality of life for millions of Americans. The 30-year mortgage was created by the Federal Housing Administration in the 1930s as a response to the Great Depression. Before this intervention, only the very wealthy could afford to buy a home given the high downpayment requirements and short-term loans. Since this policy change, thousands of financial institutions have offered long-term mortgages to millions of Americans who have bought homes that provide safety and security for their families, strong communities, and an opportunity to build wealth through appreciating assets. Broad home ownership is a public good, but until the government created the right policy and regulatory framework for the markets, it was out of reach for the majority of Americans.
Similarly, the Small Business Administration’s loan guarantee programs started in the 1950s supported financial institutions, including banks and non-bank lenders, in extending credit to small businesses that would otherwise be difficult to serve with affordable credit. These programs have collectively helped millions of small businesses access the credit they need to grow their businesses, create wealth for themselves and their families, provide critical goods and services in their communities, and create a diverse and vibrant local tax base.
The financial markets, without these types of interventions, are not structured to prioritize access and affordability. Well-designed policy and complementary regulatory interventions have been proven to drive different behaviors in the capital markets that yield real benefits for American families and businesses.
The role of access to credit in driving an equitable energy transition
The public and private sectors have spent decades and billions of dollars investing in the development of clean technologies that reduce greenhouse gas emissions, create economic benefits, and deliver a better customer experience. Now that these technologies exist, the challenge is to deploy them for everyone, everywhere.
The barrier to widespread deployment is that most clean technologies require an upfront investment to yield long-term benefits and savings (i.e., an initial capital expense to reduce ongoing operational expenses) – technologies like solar and battery storage, electric vehicles, electric HVAC and appliances, etc. – which means that people and companies with cash or access to credit are adopting these better technologies while those without access to cash or credit are being left behind. This is yielding an even greater divide – creating economic savings, health benefits, and better technologies for those who can afford them, while leaving dirty, volatile, and increasingly expensive energy sources for the lowest-income communities.
Many of the federal policy interventions to support deployment of these new technologies to date have been through tax credits. These policies have been very popular, but are not often widely adopted, particularly in rural and lower-income communities, because, (a) they are complex, (b) they often require working with individuals or businesses with large tax liabilities, and (c) they typically come with high transaction costs, making smaller, more distributed projects harder to make work. The energy transition is a huge wave of change, but it is made up of many small component parts – individual buildings, machines, vehicles, grids – so if our policies fail to enable small projects to get done, we will fail to transition quickly and equitably.
To deploy everywhere, households and businesses need credit to offset capital expenses. To expand access to credit, we need supportive clean energy policies that work within and alongside local financial services ecosystems – just like we’ve seen with housing and small businesses.
Regulation is insufficient to drive widespread adoption
Pursuing a carbon-free economy is a massive undertaking and, understandably, much of the state and federal government’s toolkit has focused on regulation of people and businesses to drive behavior change – policies like fuel economy standards, pollution restrictions, renewable energy standards, and electrification mandates. This is an important piece of the puzzle – but insufficient to drive broad (and willing) adoption.
Take, for example, the goal of electrifying heavy-duty trucks in and around port communities. States like California have attempted to set a date at which all new trucks on the registry must be zero-emissions vehicles. Predictably, this mandate was met with a lot of pushback from truck drivers, small operators, and industry associations who struggled to see a path to complying with this regulation without a major increase in cost.
It wasn’t until the regulation was paired with direct incentives for truck purchases and an attractive and feasible financing package for vehicle acquisition and charging infrastructure that the industry actors started to come around. This has helped change behavior of both buyers and incumbent sellers in the market.
Policy-enabled finance creates tools – often used in conjunction with other policy mechanisms – that can more effectively meet people where they are with affordable, appropriate, and tailored solutions and can help demonstrate a feasible path to adoption that can help buyers and sellers in these markets adapt accordingly.
The Greenhouse Gas Reduction Fund as an innovative policy-enabled finance program
The Greenhouse Gas Reduction Fund (GGRF) is more than an emissions initiative—it is a strategic investment in economic equity and market innovation that took lessons in program design from many sectors and programs of the past. Designed with three core objectives, the program aims to:
- Reduce greenhouse gas emissions at scale
- Deliver direct benefits to communities, particularly those that have been historically underserved by the financial markets
- Transform financial markets to accelerate clean energy adoption and resilience
GGRF programs, including the National Clean Investment Fund, the Clean Communities Investment Accelerator, and Solar for All, were built to complement other Inflation Reduction Act (IRA) programs by occupying a critical middle ground between grant programs and tax credits. Grant programs provide direct, one-time support for projects and programs that are not financeable (i.e., not generating revenue). Tax credits are put into the market to incentivize private investment for anyone interested in taking advantage but are not typically targeted to any specific project or population.
GGRF bridges these approaches. It channels capital into markets where funding does not naturally flow in the form of loans and investments, ensuring that clean energy and climate solutions reach every community—but does so in a way that often extends the benefits of the tax credits and incentive programs so that they reach a broader set of projects and communities where the incentive is insufficient to drive adoption. GGRF focuses on increasing access to credit and investment in places that traditional finance overlooks by reducing risk and creating scalable financing structures, empowering local lenders, community organizations, and national financing hubs to deploy resources where they are needed most. Also, because the program makes loans and investments, it recycles capital continuously – akin to a revolving loan fund – so that the work filling gaps in market adoption can continue for decades.
GGRF’s design was built on a strong foundation of successful direct investment programs for local lenders, such as CDFI Fund awards and USDA programs. What makes it unique is its scale—tens of billions of dollars—and its centralized approach, leveraging national financing hubs to drive systemic change with and through new and existing local financial capillaries (i.e., credit unions, community banks, green banks, and loan funds). This program was not built to drive incremental progress; it is a market-shaping intervention designed to accelerate the clean energy transition while promoting widespread economic growth.
Unfortunately, the program was stopped in its tracks when the Trump administration illegally froze funds already disbursed to awardees, leading to multiple lawsuits to restore funding. Without this disruption, awardees and their partners across the country would be driving direct economic benefits for families and communities across all 50 states. In the first six months of the program, awardees had pipelines of projects and investments that were projected to create over 49,000 jobs, drive $866 million in local economic benefits, save families and businesses $2.7 billion in energy costs, and leverage nearly $17 billion in private capital. The intention and mechanics of the program were working – and working fast – to deliver direct economic, health and environmental benefits for millions of Americans.
Moving at the speed of trust: Bringing the public and private sectors together for effective implementation
For a program like the Greenhouse Gas Reduction Fund to succeed, both the private and public sectors need clarity, confidence and accountability. But most importantly, they need a baseline of trust between the parties to support ongoing creative problem solving to implement a new, scaled program with exciting promise and a limited blueprint.
For the private sector, certainty is paramount. Investors and lenders (and importantly, their lawyers) require clear definitions, consistent requirements, and transparency about the availability of funds, requirements of use, and the ability to forward commit capital to projects and businesses. They need mechanisms to leverage public dollars with private capital and assurances that counterparties will be shielded from political, compliance, and policy risk. Flexibility is equally critical, allowing actors to adapt to rapid market shifts and technological innovations without being constrained by rigid program structures. Understanding these requirements – and the needs of the financial market actors involved – is outside the comfort zone of most government agencies and employees and requires significant experience and capacity building to strengthen this muscle. Nimble thinking is not often associated with government agencies, but in policy-driven financial services, it is paramount.
At the same time, the public sector has its own requirements which require patience and understanding from the private sector. Policymakers and the EPA, the implementing agency of the GGRF, must ensure that funds are used properly and that Congressional and public oversight is robust. This means designing programs that comply with all laws and regulations while advancing policy priorities. It requires mechanisms for accountability—certifications, reporting, and transparency in how funds flow – along with safeguards against undue influence from purely profit-motivated private actors. Balancing these needs is not optional when managing taxpayer funds; it is the foundation for building trust and ensuring that the program delivers on its promise of reducing emissions, benefiting communities, and transforming markets.
Implementation requires striking the balance between the needs of the private and public actors; this was difficult and time consuming for both the federal employees and for us as private recipients. There was pressure to deploy quickly to demonstrate impact and the value of the program, but it took a long time to get contracts signed and funds in the market because of the many requirements of the public and private parties involved. We speak different languages, are solving for different constraints, and work in drastically different environments – all which led to complexity and delays.
Internal EPA requirements and federal crosscutters (i.e., federal requirements from other related laws that applied to this program) increased time to market and transaction costs. Many of these requirements came with high-level policy objectives without the ability to get to a level of detail required for capital deployment.
For example, two of the major policy crosscutters were the Davis Bacon and Related Acts (DBRA) requirements around labor and workforce, and the Build America Buy America (BABA) requirements for equipment manufacturing and component parts. While the agency and private awardees were aligned at a high level on policy intention – good-paying jobs and domestically-manufactured goods – down streaming these requirements to borrowers and projects required significantly more detail and nuance than was available to the agency, adding weeks and months onto implementation and frustration among private counterparties.
Clear expectations up front on how to manage the trade-offs – policy priorities versus capital deployment – could have helped create a high-level framework for implementation, which was a one-by-one review of use cases to determine feasibility and applicability. This added complexity and friction to the process without driving outsized results.
More requirements and complexity led to slower, more costly deployment, which meant fewer communities would benefit from the program’s goals of cutting emissions, creating jobs, and cutting household and business costs.
Another key feature of the program for the National Clean Investment Fund and Clean Communities Investment Accelerator was the ability for the federal government to leverage a Financial Agent to administer the funds. This arrangement was developed between the EPA and Treasury, leveraging a long-standing practice of the Treasury Department of contracting with external banks to provide financial services that were hard for the government to provide directly. This was particularly important for the National Clean Investment Fund program because the disbursement of funds into awardee accounts enabled the awardees to meet a core statutory requirement to leverage funds with private capital. Without this function, the cash would not be available on the balance sheet of the awardees and would be difficult to leverage with private investment.
Lastly, the reporting requirements for the program were complex, making it hard to provide clarity on what data collection was required for early transactions. Again, both parties recognized the importance of transparent data collection and dissemination but implementing that intent in practice was time consuming. A simple, standardized framework to get started that could evolve over time would have helped reduce uncertainty and supported faster deployment.
Altogether, the cross-sector translation – finding common ground between two disparate worlds – added many months onto the process of getting the program to the market which, in the current political climate, was time not spent doing the important work to educate a broad set of stakeholders on the program’s promise, potential, and purpose. A lot of this complexity could have been reduced by developing a baseline of trust between the parties through the application and award process, complemented by a common goal to improve program implementation over time.
Strange bedfellows create weak alliances
In addition to the programmatic elements of translation, the actors involved in implementing direct investment strategies tend to be unknown entities to government agencies and Congress. Even though many of the implementing organizations – the “awardees” – have been around for decades doing similar work, there were weak ties with Congress, federal agencies, and other related stakeholders. Similarly, there was a lack of understanding of the role that nonprofit and community-based financial organizations play in addressing market gaps. This mutual lack of understanding and engagement leaves room for misunderstanding, distrust or generalizations that can hinder the ability to make collective progress.
Within the agency, this was a new program type for the EPA, so requirements and design process took many months before anything was shared publicly. The Notice of Funding Opportunity was released nearly a year after the legislation was signed.
The unique form and function of the program and limited direct engagement with lawmakers and other stakeholders about the program left a vacuum of information, which led to skepticism and confusion. Because the funds were provided to awardees as grants, many interpreted this as just another grant program – a large federal spending package that would lead to “handouts” – instead of what it was, the federal government seeding a sustainable fund with “equity” that would be lent out, returned, and reinvested in perpetuity. For example, here is the Wall Street Journal editorial page,and later, the EPA press release conflating investments with “handouts”:
“Imagine if Republicans gave the Trump Administration tens of billions of dollars to dole out to right-wing groups to sprinkle around to favored businesses. That’s what Democrats did in the Inflation Reduction Act (IRA). The Trump team’s effort to break up this spending racket has led to a court brawl, which could be educational.”
The fact that this policy structure and the private sector entities charged with implementing it were relative strangers led to confusion and delay during a period that could have been spent on outreach, engagement, and education. Without that broad base of support, the program unnecessarily became a political punching bag.
To mitigate this risk going forward, there needs to be greater investment in relationship building, education, stakeholder engagement and capacity building within and among the implementing partners across all relevant government actors and their private sector counterparts, especially after award selections are made. This connective tissue would go a long way in creating a baseline of common understanding of the policy objectives, program design, and implementation partners involved so all parties are aligned on strategic intent and path forward.
Making policy-enabled finance programs work in the future
If we agree that policy-enabled finance is essential to drive the energy transition and deliver broad benefits, the next step is asking the right questions about how to design these interventions for success, drawing lessons from the GGRF and other related programs.
First, what mechanisms should we use, and what are the trade-offs for each? Federally supported direct investment programs, such as managed funds, can deploy capital quickly and target underserved markets, but they require strong governance, thoughtful program design, and radical transparency, otherwise they are susceptible to the “slush fund” narrative or similar risks (i.e. conflicts of interest and political favors).
Tax credits and incentives have proven effective in attracting private investment, yet they often favor actors with existing tax liability and can leave smaller players behind. Guarantees reduce risk for lenders and unlock private capital, but they demand careful structuring to avoid moral hazard and can struggle to reach communities that are truly under-resourced.
Despite the many pitfalls of direct investment programs, they address a challenge that has plagued many of the more distributed policies: centralization and market making. Often in an attempt to let a thousand flowers bloom, policymakers underestimate the need for centralized or regional infrastructure to help with asset aggregation, data collection, product standardization, and scaled capital access. This yields local infrastructure that is sub-scale, inefficient, and unable to access the capital markets for private leverage – too small to truly shape markets.
While the GGRF’s future is uncertain given pending litigation, its purpose and role as a set of centralized financial institutions within the broader community-based financial ecosystem is critical – and needs to be more broadly understood as policymakers set future priorities.
Second, should government manage funds and programs internally or partner with external experts? Internal management within an agency offers control and accountability but can strain agency capacity and impede the ability to be an active market participant. It is also difficult to attract the right talent within the government’s pay scale, leading to an inability to recruit and high turnover. This model has been attempted through programs like the Department of Energy’s Loan Programs Office (LPO), but even that market-based program has been slower to execute, delaying critical infrastructure and technology investments by months, if not years.
On the other hand, external management brings specialized expertise and market agility, yet it raises questions about oversight and influence. No matter who the private party is, there is skepticism around the use of funds, their personal or professional gain, and their intentions with taxpayer money. In our deeply politicized world, this puts a target on the leaders of these organizations that may limit who is willing to play this role.
Quasi-public Structures
Despite the challenges, on balance it seems that internal agency management or a quasi-public structure is the most feasible path. Internal management pushes the boundaries of public agency function but goes a long way to build trust and accountability. Quasi-public structures seem to be a good compromise when feasible. Other countries have figured out how to manage these programs within a government or quasi-government agency (see the Clean Energy Finance Corporation and Reconstruction Finance Corporation, both in Australia). We can too.
At the federal level, credit programs should be managed by agencies with the skills and capacities to hold an investment function, like the Department of Energy or the Treasury Department, and leverage lessons learned from programs like DOE’s LPO and EPA’s GGRF to structure new entities. Or – like many of the state and local green banks have done – create quasi-public entities that have public sector governance and appropriations but otherwise operate independently as financial institutions with their own balance sheets, bonding authority, and staffing structure.
Lastly, if public-private partnerships are preferred, who should the government work with to implement policies meant to expand access to capital and credit? Nonprofit financial institutions often prioritize mission, community impact and are willing to arrange complex financings that require a higher touch approach but often lack scale and institutional capital access. For-profit firms bring scale and expertise but often find it hard to manage a government program with a mindset or culture that differs from their typical profit-maximization frameworks.
Depository institutions such as banks offer stability and regulatory oversight, whereas non-depositories can innovate more freely to reach the hardest to serve communities. Regulated entities provide robust and trusted infrastructure and controls, but unregulated actors may move faster and can be more creative in supporting traditionally under-resourced opportunities. Specialty firms bring deep sector or asset-class knowledge, while generalists offer broad reach and experience in managing across asset classes.
To identify the optimal path, it is helpful to look to existing programs for lessons. The U.S. Treasury’s Emergency Capital Investment Program (ECIP) demonstrates how direct investment into regulated depository institutions can mobilize significant capital for underserved communities through an existing financial ecosystem. The Loan Programs Office shows what internal management can achieve for large-scale projects. Tax credit programs like the New Markets Tax Credit (NMTC) and Investment Tax Credit (ITC)/Production Tax Credit (PTC) illustrate how incentives can transform markets, while guarantee programs such as the U.S. Department of the Treasury’s Community Development Financial Institutions Fund (CDFI) Fund Bond Guarantee and SBA 7(a) and 504 guarantees highlight the power of risk mitigation in activating and standardizing products to support secondary market access. These precedents offer valuable insights as we design future policies to accelerate a broadly beneficial energy transition.
Educating policymakers to build trust in the community finance ecosystem
Regardless of path forward, one thing remains critical – building better relationships between policymakers and the community finance industry, including community banks, credit unions, loan funds, and green banks. These are the boots-on-the-ground organizations that share a mission with many policymakers to expand economic opportunity and broaden access to capital and credit. And they are often the organizations navigating multiple public products and programs to bring affordable, quality financial services to communities.
The challenge is that most advocacy and educational work for these organizations has been siloed – there are groups representing credit unions big and small, those representing housing lenders, loan funds, green banks, and community banks. The disaggregation of these efforts has diluted the potential for policymakers to look at this ecosystem as a whole to determine how best to leverage it for public good. This is not to say that each of these individual groups does not have a role to play for their members – they all have different needs and requirements and deserve representation. But the broader industry would benefit from collaboration across these organizations to create a mechanism for these institutions to help with outreach, advocacy and education around policy-enabled finance overall. This would bring a strong and powerful group of actors together for a higher collective purpose and, ideally, create a large and diverse constituency with common goals.
State and local governments stepping up
In the near-term, the absence of federal support for clean technology deployment through policy-enabled finance creates an enormous opportunity for state and local governments to step up and push forward. Hundreds of local financial institutions were doing work to prepare for the delivery of GGRF funds to and through local projects and businesses to drive broader adoption of clean technologies. These organizations continue to have the skillsets, capacity, and pipeline to finance these projects – but need access to flexible and affordable capital to do so.
State funding efforts could mirror the program and product design of the GGRF to get deals done locally, working with one or more of the constellation of financial institutions preparing to deploy federal funds. Just because the GGRF’s programs were cut short, it doesn’t mean that the infrastructure and learnings generated should go to waste – if there are public institutions willing to commit capital, there should be many financial institutions across the country ready to put it to good use.
Conclusion
If our shared goal is an equitable, rapid energy transition, policy must do more than regulate — it must enable finance and focus on deployment, or getting great projects done. The Greenhouse Gas Reduction Fund showed both the promise and the pitfalls of large-scale, policy-enabled finance: when designed and governed well, these tools can unlock private capital, deliver measurable local benefits, and sustain long-term market transformation. When implementation gaps and weak relationships persist, even well-intentioned programs become politically vulnerable and ripe for attack. To make these programs successful within our current political context, future efforts should prioritize clear governance, cross-sector capacity, and sustained stakeholder engagement so public dollars can catalyze private investment that reaches every community.
Rebuilding Environmental Governance: Understanding the Foundations
Today we are facing persistent, complex, and accelerating environmental challenges that require adding new approaches to existing environmental governance frameworks. The scale of some of them, such as climate change, require rethinking our regulatory tools, while diffuse sources of pollutants present additional difficulties. At the same time, effective governance systems must accommodate the addition of new infrastructure, housing, and energy delivery to support communities. Our legal framework must be sufficiently stable to enable regulation, investment, and innovation to proceed without the discontinuities and gridlock of the past few decades.
In an increasingly divided atmosphere, it will take candid, multiperspective dialogue to identify paths toward such a framework. This discussion paper explores the baseline that we’re building on and some key dynamics to consider as we think about the durable systems, approaches, and capacity needed to achieve today’s multiple societal goals.
Building Blocks to Make Solutions Stick
Our environmental system was built for 1970s-era pollution control, but today it needs stable, integrated, multi-level governance that can make tradeoffs, share and use evidence, and deliver infrastructure while demonstrating that improved trust and participation are essential to future progress.
Implications for democratic governance
- Invest in strategic communications to build durable public understanding of environmental measures.
- Redesign public participation and engagement to transparently and rapidly weigh the inevitable trade-offs to a decision through open and informed consideration guided by clearly articulated principles.
Capacity needs
Modernize today’s system of cooperative federalism to address the lack of clear and intentional interconnections, adaptive feedback loops, and aligned objective, by:
- Rebuilding and sharing the data backbone at the state/local level, including working to preserve and sustain current environmental data; over time, expand interoperable data collection and make it genuinely usable to support consistent, evidence based state and local action.
- Using state/local powers for climate progress, integrating zoning, land use, infrastructure, and public health authorities into a whole-of-government discussion to address climate.
- Strengthening public sector networks to share data, best practices, and innovations across jurisdictions.
- Investing in tech-enabled capacity for states and local governments, (such as AI for decision support), satellite imagery, remote sensing, and digital modeling to lower the cost of monitoring, risk assessment, permitting, and compliance.
The early 20th Century saw the emergence of our first national laws regulating public resources— the Federal Power Act in the 1930s, the precursor to the Clean Water Act in the 1940s, and the first version of the Clean Air Act in the 1950s. Then, in a concentrated decade of new laws and massive amendments to existing ones, the 1970s saw a focus on assessing, controlling, and reducing pollution, while setting ambitious goals for human and ecosystem health. These statutes generally were constructed around specific resources—airsheds, watersheds, public lands, and wildlife habitat—and articulated specific roles for federal agencies and other levels of government. State efforts were incorporated into a nationwide system of cooperative federalism, while many states undertook their own initiatives to address environmental problems.
For half a century these laws—enacted with overwhelming, bipartisan congressional support— produced a great deal of success, with conventional pollution decreasing across many resources and regions and some species and habitats recovering. But we have plateaued in terms of broad improvements, and meanwhile novel pollutants and more diffuse, global threats have emerged. Political shifts, legacy economic interests, and a changing information landscape have played an important role, as amply recounted elsewhere.
The bipartisan legislation of the 1970s arose from both idealism and necessity, during an Earth Day moment that embraced ecological thinking in response to tangible harms to humans and the environment. The laws enjoyed massive public support and got many things right. Some were aspirational and holistic, such as the Clean Water Act’s “zero-discharge” target or NEPA’s vision “to create and maintain conditions under which man and nature can exist in productive harmony, and fulfill the social, economic, and other requirements of present and future generations of Americans.” The latter Act established the Council on Environmental Quality to coordinate this policy across the entire federal government.
Other advances came piecemeal, focused on specific resources. The U.S. Environmental Protection Agency (EPA) was cobbled together by an executive plan to reorganize several existing agencies and offices, then granted authority in a series of media-specific statutes that began with the Clean Air Act, Clean Water Act, and Safe Drinking Water Act, and later the Toxic Substances Control Act and Federal Insecticide, Fungicide, and Rodenticide Act. The Resource Conservation and Recovery Act, Superfund, and Oil Pollution Act addressed hazardous substances affecting the nation’s health and ecosystems. Implementation of all these laws required the Agency to develop in-house scientific expertise and detailed regulations that fleshed out statutory standards and applied them to specific sectors—an approach upheld for decades by the Supreme Court.
These laws made unquestionable progress on conventional pollution and waste, the visible, toxic byproducts of industrial production and consumer culture that had spurred the environmental movement and drawn a generation of lawyers to the new profession. But with specialization came fragmentation of environmental law into a plethora of subtopics, and a managerial, permit-centric legal culture that risked losing sight of ecological goals. Nor were the benefits distributed equally by race or class, as demonstrated by pioneering studies in the field of environmental justice.
As the field matured, it slowed, with congressional interventions becoming less frequent and more technical. Some of the last major amendments to a bedrock environmental statute were the Clean Air Act Amendments of 1990, enacted by a bipartisan Congress and signed by President George H.W. Bush. (The other prominent example is the Frank R. Lautenberg Chemical Safety for the 21st Century Act (Lautenberg Chemical Safety Act), a major amendment to TSCA in 2016.) Absent updated legislation, EPA regulations became paramount, but these had to run a gauntlet of shifting policy priorities, complex rulemaking procedures, litigation, and a transformed and often skeptical Supreme Court.
Critiques of this system date back almost as far as the statutes themselves. One ELI study listed 34 major “rethinking” efforts emanating from academia, blue-ribbon commissions, and NGOs between 1985 and 2014, across the political spectrum and ranging from incremental reforms to radical reinvention. One highly touted initiative, led by sitting Vice President Al Gore, resulted in some modest administrative streamlining. Most remained paper exercises, appealing to good-government advocates but lacking political support.
The stakes grew higher with increasing awareness of climate change. In June 1988, NASA and book-length treatments followed, sparking broad discussion of what was then a fully bipartisan issue. Vice President Bush campaigned on addressing it, and as President in 1992, he traveled to Rio de Janeiro to sign the U.N. Framework Convention on Climate Change. With successes like the 1987 Montreal Protocol on the ozone layer or EPA’s 1990 Acid Rain Program doubtless in mind, the Senate ratified the Framework Convention 92-0.
But climate change implicates much larger portions of the U.S. economy—energy, transportation, agriculture—at individual as well as industrial scales. While NEPA embodied the 1960s slogan that “everything is connected,” the lesson of climate change is that many things emit greenhouse gases, and all things will be affected by global warming. The need for systemic change proved to be an uneasy fit with existing site-specific, media-specific environmental laws.
Growing awareness of climate change and the scale of action needed to address it also generated a backlash from entrenched economic interests. By the mid-2000s, the Bush/Cheney administration had reversed course on federal climate commitments. It contested and lost Massachusetts v. EPA, a landmark ruling in which a narrowly divided Supreme Court held that the Clean Air Act applies to greenhouse gas emissions that affect the climate.
The Administration’s argument was captured by Justice Antonin Scalia’s flippant remark in dissent that “everything airborne, from Frisbees to flatulence, [would] qualif[y] as an ‘air pollutant.’” In Scalia’s opinion, real pollution must be visible, earthbound, toxic, inhaled, not a matter of colorless molecules interacting in the stratosphere. Even in dissent, this view set the stage for subsequent legal battles, right up to the present effort to revoke EPA’s 2009 “endangerment finding” that is now the underpinning of federal greenhouse gas regulation.
Climate change likewise laid bare the long-standing divide between environmental law, which historically regulated the power sector in terms of its fuel inputs and combustion byproducts, and energy and utility law, which focused more on transmission and distribution of the resulting power. (Both fields are further divided among federal, state, and local authorities, as discussed below.) Vehicle emissions similarly are regulated via both EPA tailpipe standards and National Highway Transportation and Safety Administration mileage standards, with California authorized to propose more stringent ones. When coordinated, this multi-headed structure produces steady advances, but in deregulatory moments it has become fertile ground for opportunism, retrenchment, and delay.
At the federal level, these questions have been exacerbated by massive shifts in administrative law, long the building block of environmental law and climate action, and in federal court rulings on the separation of powers, implicating the authority of federal agencies to issue and enforce rules. Successive administrations have run afoul of the current Supreme Court majority, whose “major questions doctrine” casts a shadow both on attempts to fit new problems into once-expansive environmental statutes, and on “whole of government” approaches that attempt to address climate change’s sources and impacts across the entire economy.
Tentative attempts by presidents to leverage executive power and emergency authority have been curtailed when invoked for regulatory purposes, but are running strong in deregulatory efforts and executive actions in the service of “energy dominance.” Whether the Supreme Court will articulate some principled limits, and whether those will be even-handedly applied to future administrations, remains to be seen. Meanwhile, the past year has seen a large-scale push to reduce environmental regulation, in parallel with abrupt reorganizations and steep reductions in the federal workforce and agency budgets. These actions were joined by sharp declines in environmental enforcement and U.S. withdrawal from environmental and climate-related international instruments and bodies.
In this uncertain atmosphere, attention has turned to new technologies and building the necessary infrastructure to effect growth in low- and zero-carbon energy. As clean energy alternatives have matured and become economically competitive, the climate imperative is pushing against long-standing environmental review and permitting procedures. That may well include NEPA, which is now attracting attention from all three branches of government and a robust debate about whether, or how much, its procedures might be slowing energy deployment.
Environmental issues were federalized for a reason: to counter pollution that crosses state borders and to prevent a race to the bottom. But decades of implementation have seen the blunting of some tools, expansion of others, and identification of gaps. Moving forward requires reaffirming that the environment is inseparable from societal health and well-being, economic stability, and energy systems. Any serious response must orient governance toward decarbonization, while embedding accountability, equity, and justice from the outset rather than inconsistently and often inadequately after the fact. Doing all this without sacrificing hard-won environmental gains will not be easy.
To meet the challenge of the worldwide crises of biodiversity loss, pollution overload, and climate change, creation of any new structure must be rooted in understanding the existing baseline for environmental governance.
- Cross-Cutting Objectives: Effective governance paths must overcome the persistent false dichotomy between the environment and the economy, making clear that energy production, economic prosperity, ecosystem health, and societal well-being are inextricably linked. Improved trust and participation are essential to sustaining and accelerating progress across these interconnected goals.
- Democracy, Expertise, and Regulatory Certainty: Our legacy environmental laws have seen many successes, but their media- and site-specific tendencies are in tension with the scale of action needed to decarbonize our economy, conserve biodiversity, and control pollution. Eroded trust, accreted layers of process, and increasingly extreme political actions and reactions hamstring progress. At the same time, rapidly advancing scientific knowledge and technology have greatly expanded our ability to anticipate environmental challenges and understand and react to the impacts of our actions. Harnessing these tools effectively can help us improve and accelerate our decisionmaking processes.
- Building a Structure Fit for Purpose: Environmental law necessarily operates at multiple scales: global, national, tribal, regional, state, and local. Our system of cooperative federalism centers authority around the federal and state governments, backstopped by treaty obligations, interstate compacts, and traditional state and local authority over land use and public safety and welfare. A strong cooperative federalism framework can foster collaboration across subnational jurisdictions, including by leaning into data collection, analysis, and dissemination to support decisionmaking. In addition, understanding the effects and drivers of private sector environmental actions can help to identify ways to leverage those actions to augment and fill gaps in public governance.
Cross-Cutting Objectives
Inseparable: Environment, Energy, Economy, and Society
The past half-century has demonstrated the impossibility of severing the environment from the economy, energy production, and social well-being. We must ensure the false dichotomy between environmental protection and economic development, characterized by an oversimplified idea that the two are in a zero-sum competition, also fades. The decades-old concept of sustainability (or triple bottom line) has not yet made its way into many of our foundational laws and governance structures.
Ignoring the complex relationships among environment, energy, the economy, and society favors short-term decisions that externalize impacts. This underlies the longstanding debate over the accuracy and efficacy of cost-benefit analyses, throughout their 40-plus year federal history, including questions about scope and how they handle uncertainty. For any project or program, system designers that consider an integrated suite of factors that move beyond basic environmental parameters or economic indicators (from public health to workforce development, from the supply chain to community well-being) have a greater chance of cross-sector success.
These governance challenges are also inseparable from shifts in how finance flows. Public and private financial tools—from subsidies and tax credits to loans, grants, and community-based financing—are increasingly shaping market behavior and determining whether policy objectives translate into real-world outcomes. Who controls these tools, how they are deployed, and when capital is made available all play a central role in driving or constraining environmental progress.
Bridging these gaps is, of course, easier said than done. But widening the aperture of considerations can connect decisionmaking to holistic industrial policies that account for a wider range of economic, social, and environmental factors. Accounting for this wider range isn’t just a nice-to-have, but essential to shared prosperity.
Foundational: Trust and Participation
A process, project, or program will move at the speed of trust—no faster and no slower. This refers to trust in institutions, in science, and in process.
Trust is earned through consistent transparency, clear accountability, and demonstrated responsiveness. For governance systems to function at the scale and pace required today, these principles must be embedded in decisionmaking in ways that are coherent and durable, rather than fragmented across a series of disparate steps and entities. Our traditional frameworks contain mechanisms to solicit and incorporate public input. But those mechanisms have limitations for all involved, both those trying to make their voice heard and those proposing the action and receiving input. (These range from when and how often participation occurs in the decisionmaking process to how the input is incorporated and decisions communicated.) Participation is foundational to our regulatory democracy and must occur early enough and in meaningful ways to improve decisions.
Effective participation also depends on clarity. People must be able to understand how decisions are made, what tradeoffs are being weighed, and where and how engagement can influence outcomes. But our frameworks still reflect reliance on elite and professional representation rather than widespread engagement. Trust—and the durability of outcomes—will increase when our processes have clearly articulated principles, transparently and rapidly weigh tradeoffs, and come to decisions through open and informed consideration.
The Concurrent Risk and Promise of Technology
Mechanization and industrialization created both unprecedented wealth and the pollutants that were the target of the 1970s wave of environmental laws. Emerging technologies likewise offer great promise, but also place familiar stresses—greenhouse gas emissions, water consumption, land use, waste—on the ecosystem and on human health and well-being. Our existing laws will need to respond and adapt to these problems as data centers and other novel demands reach greater scale, even as we evolve new ways of balancing those technologies’ potential against their up-front impacts and opportunity costs.
Technology also offers a potential path through the climate crisis, as solar and wind energy have become scalable and cost-competitive with traditional fossil fuels. Other clean technologies on the horizon, such as geothermal or fusion energy, retain bipartisan support and will require legal and regulatory guardrails if they mature and are integrated into the system. Battery storage and energy efficiency advances will help manage and reduce energy demand, and carbon removal and sequestration technologies may also play a role in curbing emissions. And at the outer limits of our knowledge, various geoengineering concepts are raising difficult questions about feasibility, decisionmaking procedures, unintended consequences, and accountability.
New technologies are also helping shape the implementation of environmental law in important ways. Existing tools such as satellite imaging, GPS location and geographic information systems, remote monitoring and sensing, and drones have fundamentally altered the way we view and record data from the physical world, in close to real time. Computer modeling and simulations have been a mainstay of climate science and policy, and other software innovations may improve environmental governance, including addressing long-standing issues of government transparency and public participation.
Effective messaging is essential to enhancing public understanding of interconnected issues and support for responses. It should be tailored to specific jurisdictions and informed by advances in research (e.g., behavioral science), learn from those thriving in today’s information ecosystem, and embrace strategies for reducing polarization.
How can we identify and address barriers to the development and equitable deployment of technologies that advance environmental protection while limiting their negative impacts.
Democracy, Expertise, and Regulatory Certainty
In a healthy democracy, public policy is guided by evidence, and truth is the shared foundation for collective decisionmaking, whatever the chosen outcome. When facts and scientific expertise are dismissed or minimized in favor of ideology, however, it becomes harder for citizens to deliberate, solve problems, and hold leaders accountable. The diminution and marginalization of science contribute to the erosion of democracy itself.
In the United States, our ability to build necessary infrastructure and take action has been slowed by the long timelines and sometimes overlapping requirements of our regulatory processes. This is exacerbated by the increasingly extreme policy swings we have been experiencing between administrations. The result is the twin challenge of how to increase the pace of our processes without lessening their protections, while also making our decisions more stable and durable.
Aligning Regulatory Certainty and Timelines
Regulatory certainty is not the same thing as rigidity. When done correctly, it can be the backdrop against which communities are able to plan for the future and companies can make informed decisions about where and how to invest. Regulation that is sufficiently clear on stable objectives does not have as much space in which to swing.
Long horizons with clear milestones matter: think of a national clean electricity standard, or the emissions-based equivalent, set on a 15- to 20-year glidepath. Confidence in long-term decisions, however, stems from effective inclusion, holistic analysis, and transparent decisions. The perspectives of subject-matter experts (in-house and external), and of those who manage and care about the resources or land in question, should be considered essential and actively pursued by policymakers.
Program-level thinking can help inform decisions at the project level. The energy transition will be remembered for feats of engineering—the thousands of miles of transmission lines, the buildout of battery storage—but its success will be determined by whether our framework listens, incorporates needed expertise, and produces rules that last long enough for people to plan their lives.
Evidence-Based Decisionmaking
For decades, the principle that good decisions require a good evidence base has been axiomatic. Dating back to 1945, the federal government has invested in science as a discipline and an idea, with government supporting the research to be conducted by public institutions and delivered as socially useful goods by the private sector.
Incorporating meaningful, often complex, evidence—including scientific data, traditional knowledge, and the needs, concerns, and priorities of potentially affected individuals—into decisionmaking is increasingly fraught. Climate change illustrates these challenges: despite decades of understanding by government officials and private sector decisionmakers about its causes and the need to act, economic and social interests have prevented effective policy and legislative response. Decisions are as good as the information they are based on. Emissions reductions ultimately depend not just on technical knowledge, but on institutions and governments capable of acting on that knowledge independently, transparently, and free from corruption and clientelism.
In a study assessing the effectiveness of the federal government’s efforts to improve evidence-based decisionmaking, the U.S. Government Accountability Office found mixed progress in: (1) developing relevant and high-quality evidence; (2) employing it in decisionmaking; and (3) ensuring adequate capacity to undertake those activities. These are foundational problems.
Compounding our challenges in making legislative and policy decisions based on accurate and pertinent evidence is the siren song of AI. Artificial intelligence promises many tools, ranging in complexity and autonomy from providing clerical tasks to generating substantive recommendations. (AI Clerical Assistive Systems automate certain administrative and procedural tasks, such as document classification and automatic transcription, and AI Recommendation Systems can contribute to judicial decision-making, for example, by analyzing legal codes and case precedents. Paul Grimm et al.)
AI is already being used across jurisdictions and agencies for environmental regulation, including planning, reviewing proposals, drafting environmental reviews, public participation and engagement, monitoring compliance, and enforcement. Recent federal policy has fueled the AI flame, with a 2025 AI action plan and multiple Executive Orders that offer the power to expedite permitting processes.
Enormous governance questions around AI have yet to be resolved. Technologies built by people reflect the values and assumptions of those who built them, and their use shifts power in decisionmaking processes. If a judge were called upon to review a decision made by such a tool, how could she determine the finding was reasonable under existing standards of administrative law? Can machine-generated analysis satisfy NEPA’s “hard look” review? These types of governance concerns dog AI tools wherever they are deployed but become particularly critical when they have the potential to become the decisionmaker in our legal and regulatory system.
The importance of having rigorous systems for identifying and considering trusted information to ground collective and democratic decisionmaking cannot be overstated. Until recently, dozens of scientific advisory committees routinely advised federal agencies to help bridge information gaps. Staggering recent losses of federal research funding and government programs and scrubbing of essential data sets means any path forward will likely require significant investments of both financial and human capital. When we rebuild, priority should be placed on ensuring all participants in decisionmaking have access to the same evidence, supported by the same systems.
Frontloading Regulatory Decisionmaking
Even as we work to improve how evidence informs decisionmaking, we face growing risks, uncertainties, and tradeoffs. The challenge is not simply to generate more information, but to make better use of what we already know through regulatory systems that reflect the integrated nature of the problems we face—without mistaking uncertainty for an absence of evidence.
Many conflicts arise because decisions are fragmented across regulatory silos and institutions. Consider a proposed electrical transmission line crossing a wetland. Decisionmakers must balance the imperatives of the energy transition, the conservation of biodiversity, the protection of water resources, and local economic opportunities. Yet these factors may be evaluated at different times, at different scales, and by different agencies. As a result, environmental permitting decisions can be made in isolation, long after foundational choices about the project’s purpose and design have already been locked in.
By the time site-specific questions arise, such as whether a particular wetland falls within the narrowed jurisdiction of the Clean Water Act, many broader tradeoffs have already been foreclosed.
A holistic approach would entail identifying the priority of certain projects and a system for weighing their impacts. For example, infrastructure decisions could happen at a systemic scale such as nationwide grid needs, providing context for decisions about individual projects and resources. Our decisionmaking processes need systems for weighing tradeoffs, and making them transparent, to enable systems-level planning and prioritization and effective engagement.
Hard decisions will have to be made regarding prioritized (and thus deprioritized) objectives. But frontloading data gathering, assessment, and decisionmaking on a national scale—through meaningful scenario planning, for example—could reduce the number of decisions made much further down the line in a project lifecycle and temper the uncertainty that can stem from permitting officials’ discretion.
We will be facing these types of tradeoffs with increasing frequency as needs mount to build infrastructure and housing, retreat from our coasts, manage and conserve species and ecosystems, and respond to and prepare for increasingly frequent and severe emergencies. In addition to an integrated approach for assessing impacts and making tradeoffs transparent, the system will need certain decisions to be made earlier in the decisionmaking processes and with a broader scope.
Acting (and Adapting) Amidst Uncertainty
Core tenets of administrative law structure decisionmaking with up front analysis and assume that we have full—or at least sufficient—information about circumstances and potential impacts to support a decision. But this is not always the case. When there are substantial uncertainties about conditions or the possible impacts of an action or rulemaking, adaptive management can improve outcomes by taking an iterative, systematic approach.
The uncertainties brought on by changing conditions due to climate impacts and unknowns about the consequences of proposed actions may call for an adaptive approach. And there are other situations where establishing sufficient evidence before taking irreversible action is appropriate. For example, we currently have limited understanding of the potential local and global impacts of geoengineering proposals to release aerosols into the atmosphere to block the sun’s rays, nor are there governing mechanisms in place to address them.
There are also situations where it is important to ensure that we do not indefinitely postpone action due to a desire to have all the answers before acting, such as infrastructure for transitioning away from fossil fuel combustion. When appropriate, effective adaptive management plans include procedural and substantive safeguards such as clear goals to set an agenda and provide transparency, an accurate assessment of baseline conditions to compare future monitoring data against, an outline of the thresholds at which management actions should be taken to promote certainty and assist with judicial enforcement, and is linked to response action.
Learning as we go and making appropriate adjustments may be justified in some contexts, and even essential when we do not have the luxury of time and must move ahead without critical information. Adaptive management can increase an agency’s ability to make decisions and allow managers to experiment, learn, and adjust based on data. But adaptive management’s flexibility comes at the cost of more resources and less certainty, which may also invite controversy. The sweet spot for adaptive management may be when managing a dynamic system for which uncertainty and controllability are high and risk is low. While uncertainties are proliferating, situations that meet those conditions are not the norm.
It would be beneficial for our environmental governance systems to explicitly identify conditions under which adaptive management may and may not be used, and to provide clear accountability mechanisms. The approach must fit with the practical realities of the working environment. For example, even if uncertainty and controllability are high and risk is relatively low, tinkering with large-scale energy infrastructure is not practical. Adaptive management may not be suited to regulatory contexts (1) in which long-term stability of decisions is important; (2) where decisions simply can’t easily be adjusted once implemented; or (3) where it is essential that an agency retain firm authority to say “yes” or “no” and leave it at that. It is a valuable tool to be invoked when truly necessary.
The interconnectedness of today’s global environmental challenges is in tension with the accreted framework of media-specific, site-specific laws and siloed agencies. Adjustments that help to align objectives, processes, and structures could scale impact.
Our framework should reflect commitment to and investment in gathering and analyzing information, from intricate science to the concerns of impacted communities; and be designed to incorporate and respond to changing information, such as through judicial review or other checks.
In part because of impacts already set in motion, we must consider when we cannot wait for more information before taking action on environmental and climate challenges. By their nature, some of those actions can be adapted on an ongoing basis, while others cannot. Clear parameters for differentiating will help ensure clear timelines and appropriate, effective processes.
Building a Structure Fit for Purpose
The triple planetary crises, a term coined by the UN Environment Programme, refers to the challenges of biodiversity loss, pollution overload, and climate change. They require large-scale mobilization and societal level adjustments. This magnitude of action requires a multifaceted system that can support and move myriad levers in a coordinated and balanced manner. The year she received the Nobel Prize in Economics, Elinor Ostrom published a paper capturing the tension but also necessity of this layered system, calling for a “polycentric approach” to addressing climate change.
The following discussion focuses largely on federal and state government action. In addition, Tribal Nations are vital sovereign authorities, partners, and voices in governance, including natural resource management, and their needs and knowledge are critical to effective, sustainable, just results. And as Ostrom recognized, private entities will also be instrumental in addressing climate change and other complex challenges; this includes not only corporations, as discussed below, but philanthropic organizations and a variety of other nongovernmental actors.
The Scale Challenge
Environmental regulation occurs at multiple levels: local ordinances, state laws and policies, interstate agreements, tribal laws, federal regulations, and international laws and norms. It also works at different resource scales, from managing a subspecies to protecting regional drinking water to setting nationwide air standards.
Jurisdictional nesting can provide comparative benefits at various levels for specific resources or pollutants. For example, working at the local level may allow for tailoring to specific circumstances to maximize benefits and the building of trust, while working at the state level can allow for the cumulative benefits of collective local action while also allowing for the testing of different approaches to federal implementation. Meanwhile, working at the federal and larger scale allows, among other things, the balancing of voices, and the establishment of shared objectives, standards, or requirements.
However, tiered systems can also be subject to gaps in implementation, such as when there is no mechanism to trigger enforcement of an international mandate at a national level. This may inadvertently impede interoperability and shared learning, such as by using different data standards, tools, or systems, and slow action due to competing or otherwise unaligned priorities. In addition, rarely do jurisdictional boundaries align with resource definitions, whether it be a hydrogeographic basin, extent of an air pollutant, or natural hazard vulnerability zone. Further complexity is added by questions around preemption, with changes occurring in longstanding understandings of federal versus state authorities under key statutes and regulatory structures.
Federal, tribal, state, and local governments must navigate these challenging dynamics as they work to effectively implement existing environmental laws and creatively address new environmental problems.
Cooperative Federalism
Federalism—whereby the federal government and states share power and responsibilities—is a central tenet of the U.S. governance system. A particular form, cooperative federalism, is embodied in most of the major U.S. environmental laws, including the Clean Air Act and the Clean Water Act. These laws establish a legal framework in which minimum standards are established at the federal level and individual states implement the programs. Today, over 90 percent of the delegable federal environmental programs are run by states. As a general matter, states are responsible for ensuring that federal standards are met but have the flexibility to impose standards that are more stringent than the federal standards.
In practice, the Congressional Research Service observes that the “precise relationship and balance of power between federal and state authorities in cooperative federalism systems is the subject of debate.” This debate has manifested in a variety of ways over the decades, including differences over the appropriate scope of federal oversight and levels of federal funding for state-delegated programs.
Environmental protection has advanced in many respects over time with cooperative federalism as its foundation, but few would argue there is no room for improvement. For example, a 2018 memorandum by the Environmental Council of the States (ECOS) captured a consensus among states that the “current relationship between U.S. EPA and state environmental agencies doesn’t consistently and effectively engage nor fully leverage the capacity and expertise of the implementing state environmental agencies or the U.S. EPA.”
In addition to the leeway that cooperative federalism provides to the states in implementing federal environmental laws, states are free to regulate or otherwise address environmental problems that are not covered by federal laws. As a result, states are often referred to as (in Justice Brandeis’ phrase) “laboratories of democracy” for testing innovative policies. Historically, states have served as testing grounds for environmental policies later adopted by the federal government. Given the current federal governance landscape, discussed below, what happens in the states may stay in the states (at least for quite some time)—making state laboratories one of the few promising options for advancing environmental protection.
Barriers to Optimal Functioning of Cooperative Federalism
In addition to the inherent systemic challenges outlined above with respect to multi-tiered jurisdiction and resource scale, there are broad societal barriers to maximizing the efficacy of cooperative federalism. The numerous overarching problems contributing to democratic dysfunction (e.g., channelized communication, primaries that yield extreme candidates who foster dramatic pendulum swings, lack of public trust) will contribute to impeding the optimal functioning of cooperative federalism for the foreseeable future.
The multitude of environmental governance-specific challenges identified earlier also significantly affect the functioning of cooperative federalism. These include, for example, long-standing congressional gridlock; new and emerging environmental harms that cannot be easily addressed within the existing, siloed framework; a Supreme Court changing its review of regulation; and regulatory pendulum swings that make consistency and stability difficult and hinder continuous improvement.
In addition, several additional barriers arguably weaken the foundations of cooperative federalism. These include: ineffective federal oversight of state programs (possibly both too stringent and too lenient in some respects); insufficient collection and dissemination of data (e.g., on environmental conditions, performance, pollution impacts), as well as inconsistent tracking of key environmental indicators; lack of state-specific effective risk communication and messaging; limited state resources for filling federal regulatory gaps or experimenting with innovative ways of implementing federal and state regulations; and insufficient federal funding for state programs. Recent critiques also point to the need to build out state administrative law to improve the functioning of cooperative federalism.
Opportunities for Renewing Cooperative Federalism
Recent developments in federal programs are disrupting many aspects of the country’s environmental protection efforts. These developments include drastic regulatory rollbacks, multiplied industry influence, curtailed input from scientists and other experts, rollback of federal grant funds to states and local governments, and sweeping staffing cuts resulting in loss of critical expertise.
Cooperative federalism has been particularly undermined by federal funding cuts (e.g., withdrawal of federal grants, reductions in revolving loan funds) and cuts to the federal programs that collect and analyze environmental data. Moreover, federal interference with independent or “more stringent than” state initiatives is taking a toll (e.g., response to California’s electric vehicle requirements ).
Given the barriers outlined above that make major statutory change infeasible, building an entirely new structure to replace cooperative federalism will be a nonstarter for the foreseeable future. However, ample opportunities exist to strengthen the existing structure in a manner that yields more effective and innovative approaches to environmental protection.
Front and center is building state and local governmental capacity to fill the gaps created by federal inaction and rollbacks as well as to lead on regulatory innovation. In so doing, states and local governments can serve as more effective laboratories of democracy and foster innovative federal action. And because states and local governments are on the frontlines of managing environmental and climate impacts such as floods and wildfires, as well as aging water infrastructure and other environment-related challenges, they are motivated to address the cause and effects of these harms, despite the intensely politicized nature of environmental issues such as climate change.
To be sure, renewing the existing structure is complicated by an uneven political landscape. For example, the level of political and popular support for environmental protection measures in the 26 states led by Republican governors differs from the levels of support in the 24 states led by Democratic governors, and the relative dominance of a particular party (e.g., trifectas or triplexes) is also a factor. These dynamics likewise influence environmental action by local governments when, for example, the potential for state preemption of local authority is a factor.
Nevertheless, the practical reality of increased extreme weather events, aging water infrastructure, and other environment-related challenges provides a strong incentive for all states and local governments to act. State and local efforts, however, are hindered by limited capacity in the form of staffing, funding, expertise, data, and other factors. For example, virtually all states could benefit in their decisionmaking from more robust data on local environmental conditions, and many states lack adequate funding, staff, and other resources.
Private Sector Synergies and Opportunities
Private environmental governance (PEG)—which can take a range of forms including collective standard-setting, certification and labeling systems, corporate carbon commitments, investor and lender initiatives, and supply chain requirements—is already making its mark across industries as diverse as electronics, forestry, apparel, and AI. For example, roughly 20 percent of the fish caught for human consumption worldwide and 15 percent of all temperate forests are subject to private certification standards. In addition, 80 percent of the largest companies in key sectors impose environmental supply chain contract requirements on their suppliers. And investors are increasingly taking environmental, social, and governance (ESG) into account, including risks related to climate change. A 2022 study estimated, for example, that assets invested in U.S. ESG products could double from 2021 to 2026 and reach $10.5 trillion.
As professors Vandenbergh, Light, and Salzman explain in their book Private Environmental Governance: “If you want to understand the future of environmental policy in the 21st century, you need to understand the actors, strategies, and challenges central to private environmental governance.”
Given the scope of PEG activities, it is not surprising that a range of regulatory regimes are implicated, including corporate governance, contract, antitrust, and consumer protection laws. In some cases, these legal regimes place constraints on the forms and scope of PEG initiatives. Many contend, however, that these constraints are inadequate, as reflected in recent efforts to severely curtail ESG initiatives.
Further, some scholars and advocates have criticized PEG from an entirely different perspective, citing concerns that PEG measures constitute greenwashing—that is, that they do not actually change corporate behavior and environmental conditions. Among other concerns is that PEG may undermine support for public governance measures in certain contexts.
Yet federal legislative gridlock, a dramatically swinging environmental regulatory pendulum, unregulated new technologies, and other factors point to needing a better understanding of how PEG can be leveraged to advance environmental protection efforts—including the improved functioning of cooperative federalism.
How can we use innovative approaches for preserving existing data and collecting new data on environmental conditions, regulated entity performance, and pollution impacts to enhance interoperability of local, state, and federal systems, foster consistency among assessments of risk, and help align priorities and approaches?
Problems such as climate change require a whole of government approach to address and could benefit from leveraging adjacent state and local regulatory authorities in areas such as land use (e.g., zoning), infrastructure, and public health.
Bolstering state and local officials’ networks for sharing data, best practices, and regulatory innovations may help align priorities and produce further progress on cross-jurisdictional problems as well as new challenges such as permitting reforms.
For example, asking—what are the effects of PEG (e.g., emissions reductions); what are the drivers of PEG (e.g., brand reputation, shareholder actions, employees, and corporate customers); are there ways to reduce greenwashing and greenhushing; and how can we ensure that PEG complements public governance.
For example, AI and advanced monitoring technologies—if thoughtfully leveraged—could lessen the burden on state and local governments, particularly those that are under-resourced, in their efforts to assess climate risk, develop resilience plans, and monitor regulatory compliance.
Conclusion
The environmental gains of the last half-century demonstrate that governance choices matter. The United States built a system capable of addressing the urgent environmental crises of its time by combining scientific expertise, democratic accountability, and enforceable legal standards.
Today’s urgent challenges—climate change, biodiversity loss, and pervasive pollution—demand a similar alignment under far more complex conditions. The challenge is not merely to regulate more, faster, or differently, but to recommit to decisionmaking that is credible and durable: by restoring confidence that evidence matters, that participation is meaningful, that tradeoffs get confronted honestly, and that rules will persist long enough to justify investment and collective effort.
The path forward lies neither in abandoning the foundations of environmental law, nor in relying solely on technological or private solutions. It will be found by strengthening and adapting existing governance structures—integrating cross-cutting objectives across domains, clarifying roles across jurisdictions, and rebuilding the shared evidentiary base and institutional capacity needed to act amid uncertainty, rather than deferring action in pursuit of unattainable certainty. And it requires clear communication about today’s complex, dispersed challenges that enhances understanding and reduces polarization.
At its core, the triple planetary crisis is a democratic and governance challenge: how societies decide, together, to protect people and places while sharing costs and benefits fairly. Meeting that challenge will require systems capable of carrying both technical complexity and public trust, as well as a sustained commitment to invest in institutions that can decide, act, and endure.
Costs Come First in a Reset Climate Agenda
Building Blocks to Make Solutions Stick
Durable and legitimate climate action requires a government capable of clearly weighting, explaining, and managing cost tradeoffs to the widest away of audiences, which in turn requires strong technocratic competency.
Democratic governance needs
- Clear articulation of tradeoffs in policy design, including who pays, who benefits (and when), and why.
- Think bigger and wider in building durable coalitions for climate action, mobilizing dispersed beneficiaries and taking advantage of policy Overton windows that cut across partisan lines.
State Capacity needs
- Intergovernmental delivery muscle and partnering capacity to enable state and local actors.
- Invest in technocratic state capacity where the big wins live, like permitting and siting, interconnection and transmission, or power-market governance, and implementation capacity to limit bottleneck-driven policy failures.
- Institutionalize rigorous ex ante and regular cost benefit analysis to guide design and mid-course corrections.
Key Takeaways
- The costs of climate policy influence whether reforms benefit society, as well as their likelihood of passage and durability. Four ways to categorize climate policy costs are: negative-cost policies (pro-growth policies with climate co-benefits); low-cost policies (costs below domestic climate benefits); medium-cost policies (costs below global climate benefits); and high-cost policies (costs above global climate benefits). Cross-partisan alignment is most evident among pro-abundance progressives and pro-market conservatives.
- Negative- and low-cost policies align with domestic self-interest and comprise a growing share of the abatement curve. For example, market liberalization in permitting, siting, electricity regulation, and certain transportation applications lower energy costs and have profound emissions benefits. A prominent low-cost policy is emissions transparency. Negative- and low-cost policies hold the most potential for durable reforms and are often technocratic in nature.
- Chronic underconsideration of costs has induced an overselection of high-cost policies and underpursuit of low- and negative-cost policies. Legislative policies, such as subsidies and fuel mandates or bans, often receive no ex ante cost-benefit analysis before adoption. Interventions receiving cost-benefit analysis, especially regulation, tend to underestimate costs.
- Innovation policy – namely public support for research, development, and early-stage deployment – can align with domestic self-interest and address legitimate market deficiencies. By contrast, industrial policy for mature technology carries high costs, often erodes social welfare, and is not politically durable. Notably, public support for mature technologies in the Inflation Reduction Act was not durable, but support remained for nascent industry.
- We recommend that a reset climate agenda focus on abatement results over symbolic outcomes, prioritize state capacity for technocratic institutions, and emphasize cost considerations in policy formulation and maintenance. Negative cost policies warrant prioritization, with an emphasis on mobilizing beneficiaries like consumer, non-incumbent supplier, and taxpayer groups to overcome the lobbying clout of entrenched interests. Robust benefit-cost analysis should precede any cost-additive policies and be periodically reconducted to guide adjustments.
Introduction
Public policy involves tradeoffs. The primary tradeoff for climate change mitigation is economic cost. Secondary tradeoffs include commercial freedom, consumer choice, and the quality or reliability of goods and services. Political movements seeking to address a collective action problem, such as climate change, are prone to overlook the consequences of tradeoffs on other parties, like consumers and taxpayers. This paper posits that the cost tradeoffs of climate change mitigation have been underappreciated in the formation of public policy. This has resulted in an overselection of high cost policies that are not politically durable and may erode social welfare. It also results in overlooking low or negative-cost policies that are durable and hold deep abatement potential. These policies can have broad political appeal because they align with the self-interest of the United States, however they typically require dispersed beneficiaries to overcome the concentrated lobby of entrenched interests.
A core, normative objective of public policy is to improve social welfare, which “encourages broadminded attentiveness to all positive and negative effects of policy choices”. Environmental economics determines the welfare effects of climate change mitigation policy by the net of its abatement benefits less the costs. The conventional technique to determine abatement benefits is the social cost of carbon (SCC). The barometer for whether climate policy benefits society is to determine whether abatement benefits exceed costs. Accounting for full social welfare effects requires consideration of co-benefits as well, granted these tend to be conventional air emissions with existing mitigation mechanisms covered under the Clean Air Act. Nevertheless, accounting for costs is essential to ensure climate policy benefits society.
Abatement costs also have a discernable bearing on the likelihood and durability of policy reforms. Climate policies exhibit patterns of passage, mid-course adjustments, and political resilience across election cycles based on the constituency support levels linked to benefit-allocation and cost imposition. This paper develops four policy classifications as a function of their abatement benefit-cost profile, and uses this framework to examine the political economy, abatement effectiveness, and economic performance of select past and potential policy instruments.
Political Economy and Policy Taxonomy
The translation of climate policy concepts into legitimate policy options in the eyes of policymakers can be viewed through the Overton Window. That is, politicians tend to support policies when they do not unduly risk their electoral support. The Overton Window for climate policy is constantly shifting within and across political movements with the foremost factor being cost.
In a 2024 survey of voters, the most valued characteristics of energy consumption were 37% for energy cost, 36% for power availability, 19% for climate effect, 6% for U.S. energy security effect, and 1% for something else. Democrats slightly valued energy cost and power availability more than climate effects. Independents and Republicans heavily valued energy cost and power availability more than climate effect.
Progressives have long exhibited greater prioritization of climate change policy, but cost concerns are driving an overhaul of the progressive Overton Window on climate change. In California, which contains perhaps the most climate-concerned electorate in the U.S., progressives have begun a “climate retreat” to recalibrate policy as “[e]lected officials are warning that ambitious laws and mandates are driving up the state’s onerous cost of living”. Nationally, a new progressive thought leadership think tank is encouraging Democrats to downplay climate change for electoral benefit. Importantly, they find that 61% of battleground voters acknowledge that “climate change is at least a very serious problem,” but that “it is far less important than issues like affordability.”
Similarly, veteran progressive thought leaders, such as the Progressive Policy Institute, now stress that “energy costs come first” in a new approach to environmental justice. While emphasising the continued importance of GHG emissions reductions, those policy leaders are making energy affordability the top priority, amid a broader Democratic messaging pivot from climate to the “cheap energy” agenda. The rise of cost-conscious progressives is particularly notable because the progressive electorate has expressed a higher willingness to pay to mitigate climate change than moderate and conservative electoral segments.
Economic tradeoffs, namely costs and more government control, has long been the central concern on climate policy for the conservative movement. The conventional climate movement messaged on fear and the need for economic sacrifice, which is the antithesis of the conservative electoral mantra: economic opportunity. Yet the conservative climate Overton Window emerged with a series of state and federal policy reforms when climate change mitigation aligned with expanded economic opportunity. However, pro-climate conservative thought leaders remain opposed to high cost policies, such as calling to phase out Inflation Reduction Act (IRA) subsidies for mature technologies.
Many leading conservative thought leaders continue to challenge the climate agenda writ large because of its association with high cost policies. For example, President Trump’s 2025 Climate Working Group report was expressly motivated by concerns over “access to reliable, affordable energy” while acknowledging that climate change is a real challenge. Similarly, a 2025 American Enterprise Institute report finds that the public is most interested in energy cost and reliability and unwilling to sacrifice much financially to address climate change. Meanwhile, climate-conscious conservative thought leaders like the Conservative Coalition for Climate Solutions and the R Street Institute continue to emphasize a market-driven, innovation-focused policy agenda that prioritizes American economic interests and drives a cleaner, more prosperous future. Altogether, it indicates a conservative Overton Window on negative and low-cost climate change mitigation.
While cost is driving the Overton Window within each political movement, it also buoys the potential for alignment across political movements. Political movements are not monoliths, but rather exhibit major subsets within each movement. The progressive movement has seen gains in popularity among its populist left flank, often identified as the “democratic socialist” wing, which contributes to ongoing debate about Democrats’ ideological direction. Climate policy initiated by this wing, however, is associated with high economic tradeoffs (e.g., degrowth) and has prompted a backlash within the progressive movement. By contrast, a subset of the progressive movement, sometimes labelled “abundance progressives,” has emerged to support a more pro-market, pro-development posture. This movement is especially responsive to energy cost concerns, and is an emerging substitute for the anti-development traditions of the progressive environmental movement. Overall, variances in the progressive movement are fairly straightforward to categorize linearly on the economic policy spectrum.
The Republican electorate views capitalism far more favorably than Democrats, but with modest decline in recent years. Republicans have trended away from consistently conservative positions associated with limited government, which historically emphasized the rule of law and a strict cost-benefit justification for government intervention in the market economy. They have migrated towards right-wing populism associated with the Make America Great Again (MAGA) movement. Right-wing populism is hard to operationalize for economic policy because it is not a standalone ideology, but a movement vaguely attached to conservative ideology. Generally, the “America First” orientation of MAGA implies positions based on the self-interest of the U.S., with the Trump administration prioritizing cost reductions in energy policy.
MAGA is further to the right of conventional conservatives on environmental regulation and general government reform. For example, conservatives have noted the contrast between conservative “limited, effective government” and the Department of Government Efficiency’s “gutted, ineffective government” reform approach. On the other hand, MAGA will occasionally back leftist policy instruments, such as coal subsidies, wind restrictions, executive orders to override state policies, and emergency authorities for fossil power plants. These are often justified to counteract the leftist policies passed by progressives (e.g., renewables subsidies, fossil restrictions, emergency authorities for renewables), resulting in dueling versions of industrial policy. In other words, ostensible overlap between MAGA and progressives on policy instrument choice actually reflects the use of similar tools used for conflicting purposes (e.g., restrictive permitting or subsidies for opposing resources; i.e. picking different “winners and losers”). Nevertheless, the disciplinary agent for right-wing energy populism has been cost concerns, which have influenced the Trump administration to pursue more traditionally conservative energy policies like permitting reform and lowering electric transmission costs.
This political economy identifies the broadest cross-movement Overton Window between moderate or “abundance progressives” and traditional conservatives. Regardless, both broad movements exhibit cost sensitivity and growing prioritization of U.S. self–interest. Distinguishing the domestic SCC from global SCC is essential to determine what policies are consistent with the self-interest of the U.S. versus the world as a whole. Traditionally, the U.S. government only considers domestic effects in cost-benefit analysis, yet the vast majority of domestic climate change abatement benefits accrue globally.
The first SCC, developed under the Obama administration, relied solely on a global SCC. Leading conservative scholars, including the former regulatory leads for President George W. Bush, criticized the use of the global SCC only to set federal regulations. They argued for a “domestic duty” to refocus regulatory analysis on domestic costs and benefits. Similarly, the first Trump administration used a domestic SCC. Although the second Trump administration moved to discard the SCC outright, this appears to be part of a regulatory containment strategy, not a reflection of the conservative movement’s dismissal of the negative effects of climate change. In other words, even if the SCC is not the explicit basis for policymaking, it is a useful heuristic for policymakers.
The proper value of the SCC is the subject of intense scholarly and political debate. It has fluctuated between $42/ton under President Obama, $1-$8/ton under President Trump, and $190/ton under the Biden administration (all values for 2020). The main methodological disagreement has been over whether to use a domestic or global SCC, with the Trump administration position guided by “domestic self-interest.” This suggests the original domestic and global SCC values may approximate the Overton Window parameters the best. This underscores the following policy taxonomy that characterizes climate abatement policies by cost relative to domestic and global SCC levels:
- Class I policy: negative abatement costs. Such policies are widely viewed as “no regrets” by scholars and political actors across the spectrum because they constitute sound economic policy that happens to carry climate co-benefits. The Overton Window is most robust for Class I policy. It typically takes the form of fixing government failure, such as permitting reform.
- Class II policy: positive abatement costs below the domestic SCC. These low-cost policies often fall within the Overton Window, because they advance U.S. self-interest (i.e., positive domestic net benefits). Class II policies have a small abatement cost range (e.g., up to $8/ton). One estimate puts them at 4-14 times smaller than the global SCC.
- Class III policy: abatement costs between the domestic SCC and global SCC. These medium-cost policies improve global social welfare, but are not in the self-interest of the U.S., excluding co-benefits. Most cost-additive policies that pass a global SCC test fall in this range, underscoring why climate change is an especially challenging strategic problem; those incurring abatement costs do not accrue most abatement benefits. Class III policies face inconsistent domestic support and often require international reciprocation to be in the self-interest of the U.S.
- Class IV policy: abatement costs exceeding the global SCC. These high-cost policies fail a climate-only cost-benefit test. In other words, Class IV policies erode social welfare, excluding co-benefits. Class IV policies may be effective at reducing emissions, but often leave society worse off. Class IV policies are challenging to pass and are hardest to sustain.
Policy Applications
There are myriad policies across the abatement cost spectrum. This analysis applies to particularly popular domestic policies already pursued or readily considered. This includes policies targeting the environmental market failure via direct abatement (GHG regulation) and indirect abatement (public spending, clean technology mandates, and fuel bans). It also includes policies targeting non-climate market failure, yet hold deep climate co-benefits (innovation policy). The analysis also examines policies that correct government failure and have major climate co-benefits (permitting, siting, and electric regulation reform).
Fuel Mandates and Bans
For the last two decades, the most prevalent climate policy type in the U.S. has been state level fuel mandates and bans. Last decade, the environmental movement came to prefer policies that explicitly promote or remove fuels or technologies, not emissions. This is despite ample evidence in the economics literature that market-based policies are more effective and carry far lower abatement costs. Nevertheless, the most common domestic climate policy instrument this century has been state renewable portfolio standards (RPS). The literature notes several key findings from RPS:
- RPS has substantial but diminishing abatement efficacy. RPS compliance drove the bulk of initial renewables deployment, but declined to 35% of U.S. renewables capacity additions in 2023. This reflects the improved economics of renewable energy, which went from an infant industry in the 2000s to a mature technology and the preferred choice of voluntary markets by the 2020s. Renewables also exhibit declining marginal abatement as penetration levels grow. This underscores the environmental underperformance of policies promoting fuel, not emissions reductions.
- Binding RPS increases costs, with large state variances based on target stringency and carveouts. RPS compliance costs average 4% of retail electricity bills in RPS states and reach 11-12% of retail bills in states with solar carve-outs. Stringency is a key factor, as some RPS are not binding due to strong market forces, whereas binding RPS increases costs. Abatement cost estimates of RPS vary widely, with one prominent study placing compliance with RPS from 1990-2015 at $60-$200/ton. Within the Mid-Atlantic region alone, implied states’ RPS compliance costs in 2025 ranged from $11/tonne to $66/tonne, with solar carveout compliance clocking in at $70/tonne to $831/tonne. The future abatement cost of renewables integration is highly sensitive to RPS stringency and technology cost assumptions, with one estimate of implied abatement costs ranging from zero (nonbinding) to $63/tonne at 90% requirement in 2050. This evidence qualifies RPS as a class II to class IV policy, depending on its design.
- States with stringent RPS face challenging compliance targets, prompting calls for reforms to mitigate cost. Compliance with interim targets has generally been strong but stringent RPS states are beginning to fall behind on their targets. For example, renewable energy credit (REC) costs are nearing alternative compliance payment levels. To reduce costs, popular reform ideas have included delaying compliance timelines, adopting a clean energy standard to capture broader resource eligibility, or making RECs emissions weighted.
- Modest RPS exists in some conservative states but aggressive RPS policy has, generally, only proven popular in progressive states. As of late 2024, 15 states plus the District of Columbia had RPS targets of at least 50% retail sales, and four have 100% RPS. Sixteen (16) states have adopted a broader 100% clean electricity standard, though the broad definition of clean energy dilutes expected abatement performance in some states. Overall, renewable or clean portfolio standards do not appear to hold broad Overton Window alignment potential beyond modest applications.
Micro-mandates have also sprung up, primarily in progressive states. These have often targeted the promotion of nascent or symbolic energy sources that the market would not otherwise provide, with the costs obscured from public view (e.g., rolled into non-bypassable electric customer charges). A good example is offshore wind requirements in the Northeast, which carries a high abatement cost (over $100/ton).
Fuel bans have become increasingly popular climate policy in progressive states and municipalities. Beginning in 2016, a handful of progressive states began banning coal. However, this does not appear to have created much cost or abatement benefit, as evidenced by a lack of commercial interest in coal expansion in areas without such restrictions. In fact, neither federal nor state regulation was responsible for steep emissions declines from coal retirements. Coal retirements were mostly driven by market forces, especially breakthroughs in low-cost natural gas production and high efficiency power plants. Policy factors, like the Mercury and Air Toxics Rule, were secondary drivers of coal plant retirement.
Around 2020, California, New York, and most New England states began adopting partial natural gas bans or de facto bans on new gas infrastructure through highly restrictive permitting and siting practices. Unlike coal restrictions, these laws have markedly decreased commercial activity, namely gas pipeline and power plant development, and in some cases caused economically premature retirements. This has caused “pronounced economic costs and reliability risk.” Resulting pipeline constraints drive steep gas price premiums in these states, which translate into a core driver of elevated electricity prices.
Insufficient pipeline service in the Northeast is especially problematic, as demonstrated by a December 2022 winter storm event that nearly led to an unprecedented loss of the Con Edison gas system in New York City that would have taken weeks or months to restore. Further, preventing gas infrastructure development does not provide a clear abatement benefit, because more infrastructure is needed to meet peak conditions even if gas burn declines. A prominent study found a 130 gigawatt increase in gas generation capacity by 2050 was compatible with a 95% decarbonization scenario.
Progressive states and municipalities have also pursued natural gas consumption bans. This policy may carry exceptional cost, especially for existing buildings, with potentially well over $1 trillion in investment cost to replace gas with electric infrastructure. One estimate put the cost of natural gas bans at over $25,600 per New York City household. A Stanford study projected a 56% electric residential rate increase in California from a natural gas appliance ban. Generally, conservative thought leaders and elected officials have opposed natural gas bans for cost as well as non-pecuniary reasons, including security concerns and the erosion of consumer choice. This applies even for prominent members of the Conservative Climate Caucus. Altogether, gas bans are considered class IV policy with virtually no Overton Window alignment.
GHG Transparency
GHG regulation takes various forms. The least stringent is GHG transparency, which addresses an information deficiency and lowers transaction costs in voluntary markets. This begins with reporting and accounting requirements on emitters (Scope 1 emissions). Public policy can help resolve measurement and verification problems that have eroded confidence in voluntary carbon markets. GHG transparency policy can also standardize terminology and provide indirect emissions platforms. For example, making locational marginal emissions rates on power systems publicly available lets market participants identify the indirect power emissions of power consumption (Scope 2 emissions). Progressives have consistently favored GHG transparency policy, while conservatives have typically supported light-touch versions of it like the Growing Climate Solutions Act.
The second Trump administration recently pursued removal of basic GHG reporting requirements on ideological grounds, specifically repeal of the GHG Reporting Program (GHGRP). This appears to reflect an optical deregulatory agenda over an effective one. Conservative groups have warned of the downsides of GHGRP repeal. Pressure to course correct may prove fruitful, given that the industry the Trump administration aims to assist – oil and natural gas – maintain that the U.S. Environmental Protection Agency (EPA) should retain the GHGRP. A recent analysis found that if states replace the GHGRP, new programs will be more expensive (Figure 2).
Many regulated industry and conservative groups instead support a low compliance cost GHG reporting regime with durability across future administrations. This not only applies to direct emissions reporting but indirect emissions reporting, as in the absence of federal policy industry faces a patchwork of compliance requirements across states and foreign governments. The same economic self-interest rationale justifies a role for limited government in emissions accounting, with an emphasis on the capital market appeal of showcasing the “carbon advantage” of the U.S. in emissions-intensive industries. An example is liquified natural gas, whose export market is enhanced by showcasing its lifecycle emissions advantage over foreign gas and coal.
The abatement effectiveness of GHG transparency has grown appreciably in the 2020s, as voluntary industry initiatives have sharply increased. This policy set enables an efficient “greening of the invisible hand” with staying power, as corporate environmental sustainability efforts appear resilient regardless of political sentiment, unlike corporate social endeavors. In fact, the aggregate willingness to pay for voluntary abatement from producers, consumers, and investors suggests that well-informed domestic markets go a long way towards self-correcting the externality of GHGs (e.g., convergence of the private and social cost curves). Certain voluntary corporate behaviors may even exceed the global SCC, especially commitments to nuclear, carbon capture, and other higher cost abatement generation financed by the largest sources of power demand growth. Well-functioning voluntary carbon markets could yield roughly one billion metric tons of domestic carbon dioxide abatement by 2030. Providing locational marginal emissions data can slash abatement costs from $19-$47/ton down to $8-$9/ton while doubling abatement levels from some power generation sources.
Overall, efficient GHG transparency policy described above is a low-cost mitigation strategy consistent with class II designation. Basic, federal GHG transparency policy may even constitute class I policy, because it avoids the higher compliance cost alternative of a patchwork of state and international standards that would manifest in the absence of federal policy. However, stringent GHG transparency policy may constitute class III or IV policy. Prominent examples include a recent California climate disclosure law and a former Securities and Exchange Commission proposed rule to require emissions disclosure related to assets a firm does not own or control (Scope 3). Such efforts may obfuscate material information on climate-related risk and worsen private-sector led emission mitigation efforts.
Direct GHG Regulation
Classic environmental regulation takes the form of a command-and-control approach. These instruments include applying emissions performance standards or technology-forcing mechanisms, typically for power plants or mobile sources. These policies vary widely in stringency and cost. Overall, command-and-control is widely considered in the economics literature to be an unnecessarily costly approach to reducing GHGs relative to market-based alternatives. It can also result in freezing innovation, by discouraging adoption of new technologies.
Federal command-and-control GHG programs have not been particularly environmentally effective, cost-effective, or demonstrated legal or political durability. The first power plant program was the Clean Power Plan, which was struck down in court, and yet its emissions target was achieved a decade early from favorable market forces and subnational climate policy. The most recent federal command-and-control approaches for GHG regulation were 2024 EPA rules for vehicles and power plants. A 2025 review of these and other federal climate regulations over the last two decades of federal climate regulations found:
- EPA’s cost estimates to be “extraordinarily conservative” with suspect methodology that was prone to error and inconsistent with economic theory;
- Assessed costs of $696 billion compared to regulators’ estimate of $171 billion, or an increase in abatement cost from $122/tonne to $487/tonne; and
- EPA is too optimistic in its assumptions of benefits.
The 2025 review study implies that past federal command-and-control had very high cost – well into class IV range. It has also been a top priority of conservatives to undercut. However, it is possible for modest command-and-control policy with class II or III costs.
Some conservatives, noting EPA’s legal obligation to regulate GHGs and the cost of regulatory uncertainty from decades of EPA policy oscillations between administrations, suggested modest requirements as a better option to replace high cost rules in order to mitigate legal risk and provide industry a predictable, low-cost compliance pathway. For example, conservatives argued that replacing high cost requirements for power plants to adopt carbon capture and storage (CCS) with low cost requirements for heat rate improvements may lower compliance costs more than attempting to repeal the Biden era rule for CCS outright. Similarly, the oil and gas industry opposed stringent GHG regulations on power plants and mobile sources, but often validated alternative low cost compliance requirements.
The first Trump administration pursued modest replace-and-repeal GHG regulation. The second Trump administration has opted for repeal policies and to eliminate the endangerment finding via executive rulemaking. However, regulated industry and many conservative thought leaders believe this is a strategic blunder, given the low odds of legal success, resulting in the perpetuation of “regulatory ping-pong that has plagued Washington, D.C., for decades.” If the courts uphold Massachusetts v. EPA and the associated endangerment finding, this implies that modest command-and-control policy may have durable political alignment potential. Yet this does not hold much abatement potential. In the absence of a legal requirement to regulate GHGs, there is unlikely to be broad political alignment for even modest command-and-control policy. Conservatives tend to view this as a gateway to more costly policies that will probably not meaningfully affect global GHG trajectories.
The 2025 review study understates the full cost of U.S. climate regulations because they exclude state and local levels. Although no comprehensive study of state climate regulation is known, command-and-control state regulations often raise major cost concerns as well. The cost and environmental performance of such state programs varies immensely, often owing to differences in the accuracy of abatement technology costs that regulatory decisions are based upon (e.g., the failure of California’s zero-emission vehicle program compared to success with its low-emission vehicle program). A recent example is California’s rail locomotive mandate, which projected to impose tens of billions of dollars in costs before being withdrawn. State command-and-control regulation is commonplace in progressive states, but not beyond, implying meager Overton Window alignment.
A more economical version of GHG regulation is a system of marketable allowances, or cap-and-trade (C&T). Over three decades of experience with C&T programs reveals two things. First, C&T is environmentally effective and economically cost effective relative to command-and-control policy. Second, C&T performance depends on its design quality and interaction with other policies. Abatement costs depend on stringency and other design features, but C&T in a backstop role is generally close to the domestic SCC, rendering it class II policy. Robust C&T generally falls in the class III policy range. C&T is an example of abatement policy that can be cost-effective on a per unit basis, but given the breadth of its coverage its total costs can be substantial. Recent developments in Pennsylvania indicate a possible preference for policies with higher per-unit abatement costs than C&T, which may reflect a political preference for policies with less cost transparency and lower aggregate costs.
Some environmental C&T complaints are valid, such as emissions leakage, but C&T effectiveness concerns are generally readily fixable design flaws. C&T effectiveness complaints are often the result of interference from other government interventions like fuel mandates, relegating C&T to a backstop role and suppressing allowance prices. Such state interventions triggered anti-competitive concerns in wholesale power markets overseen by the Federal Energy Regulatory Commission (FERC). This prompted conservative state electric regulators to call for a conference to validate mechanisms like C&T as a market-compatible alternative to high cost interventions. Conservative expert testimony at that conference, invited by conservative FERC leadership, explained that interventions layered on top of C&T merely reallocate emissions reduction under a binding cap, which raises costs, creates no additional abatement, and undermines innovation. This implies that such states might increase abatement and lower aggregate costs by upgrading the role of C&T and downgrading the role of costlier interventions.
In the 2000s, bipartisan interest in federal C&T policy arose, but it failed and has not resurfaced. In its absence, states have supplanted federal policy with subnational C&T programs. However, the durability of C&T beyond progressive states is unclear. Moderate states have sometimes joined a regional C&T program under Democratic leadership, but sometimes departed them under Republican leadership. Conservative state groups typically challenge C&T adoption and seek repeal of C&T programs like the Regional Greenhouse Gas Initiative. This suggests that C&T is at the fringe, but typically outside, an Overton Window across political movements.
Permitting and Siting
Permitting policy can base decisions explicitly on GHG criteria, or they can be based on non-GHG factors but hold indirect GHG consequences. Generally, only progressive states and presidents have pursued the former. Federally, these include the Obama administration’s “coal study” and Biden administration’s “pause” on liquified natural gas (LNG). The LNG pause did not provide any apparent emissions benefit, yet carried substantial foregone economic opportunity and strategic value to U.S. allies. Pragmatic progressive thought leaders expressed concern with the pause, noting the creation of economic and security risks, and suggested lifting the pause in exchange for companies to commit to strict, third-party verified methane emissions standards. Relatedly, some conservative thought leaders have supported policy that enables voluntary participation in certified programs that provide market clarity and confidence to harness private willingness to pay for lower GHG products. This has been buttressed by support from an industry-led effort to advance a market for environmentally differentiated natural gas based on a standard, secure certification process.
Permitting constraints on clean technology supply chains can have perverse economic and emissions effects. A prime example is critical minerals, which are essential components to clean energy technologies. A net-zero emission energy transition, relative to current consumption, would increase U.S. annual mineral demand by 121% for copper, 504% for nickel, 2,007% for cobalt, and 13,267% for lithium. Market forces, unsubsidized, are poised to produce a sufficient amount of domestic copper and lithium supply to satiate a large share of domestic demand, but face undue barriers to entry that restrict production far below its potential. To meet net-zero objectives, permitting reform allowing all currently proposed projects to enter the market would lower U.S. import reliance for copper from 74% to 41%, while dropping lithium import reliance from 100% to 51%.
Expanding domestic mining no doubt carries local environmental tradeoffs. However, the U.S. has some of the most stringent and comprehensive mining safeguards in the world. Thus, foregoing development domestically is likely to push mining toward foreign countries with inferior environmental, safety, and child labor protections. It is therefore critical that domestic permitting decisions account for the unintended effects of denying permits, not merely the direct consequences of approving a project.
Permitting and siting constraints on energy infrastructure also impose major costs and foregone abatement. These entry barriers largely exist as environmental safeguards, yet almost always inhibit projects with a superior emissions profile to the legacy resources they replace. In fact, 90% of planned and in progress energy projects on the federal dashboard were clean energy related as of July 2023. In 2023, the ratio of clean energy to fossil projects requiring an environmental impact statement to comply with the National Environmental Policy Act (NEPA) was 2:1 for the Department of Energy and nearly 4:1 for the Bureau of Land Management. A 2025 study estimated that bringing down permitting timelines from 60 months to 24 months would reduce 13% of U.S. electric power emissions.
Permitting has proven to be a litmus test for the progressive environmental movement, as the movement bifurcates between anti-development symbolists and pragmatic pro-abundance progressives. While a minority of mainstream environmental groups have become amenable to permitting reform, such as The Nature Conservancy and Audubon Society, the core of progressive environmental groups have not. Instead, new progressive groups like Clean Tomorrow and the Institute for Progress filled the pro-abundance void alongside traditional market-friendly progressive groups like the Progressive Policy Institute. This progressive subset has helped influence moderate Democrats to support permitting reform in a collaborative way with conservatives.
Permitting reform has long been championed by conservatives for its economic benefits, with climate considerations typically a secondary-at-best rationale. Yet permitting reform has become a priority for the newer climate-minded conservative movement. However, permitting has also proven to be a differentiator between conservatives and right-wing populists. The latter engages in forms of government intervention that sometimes contradict conservative principles. For example, the Trump administration enacted an offshore wind energy pause that followed the same problematic blueprint as the Biden administration’s LNG pause. This elevates the importance of technology-neutral permitting reforms with an emphasis on permitting permanence safeguards.
In recent years, a coalition of Republicans, centrist Democrats, and clean energy and abundance advocates have pressed for reform to NEPA. A broad suite of federal permitting reforms with bipartisan appeal was identified in a 2024 report by the Bipartisan Policy Center. Bipartisan alignment led to the passage of the Fiscal Responsibility Act of 2023 into law and the Senate passage of the Energy Permitting Reform Act of 2024 (EPRA). Although a 2025 Supreme Court decision suggests executive actions alone may substantially reduce NEPA obstacles, plenty of NEPA and other federal statutory reforms remain of high value and hold considerable bipartisan potential.
The positions of leading progressive, conservative, and centrist thought leadership organizations highlight alignment on various federal permitting and siting reforms. These include statutory changes to NEPA, the Endangered Species Act, the Clean Water Act, the Clean Air Act and the National Historic Preservation Act. Substantive alignment includes reforms that reduce litigation risk (e.g., judicial review reform), limit executive power to stop project approvals and undermine permitting permanence, maintain technology neutrality, strengthen federal backstop siting authority for interstate infrastructure, codify the Seven County decision, and streamline agency practices while ensuring sufficient state capacity.
Despite considerable positive momentum at the federal level, the greatest permitting and siting barriers generally reside at the state and local levels and trending sharply in a more restrictive direction. Wind and solar ordinances have grown by over 1,500% since the late 2000s. Oil and gas pipelines and power plants face mounting permitting and siting restrictions in progressive states, which not only raise costs but do not necessarily reduce emissions. In fact, the New England Independent System Operator said that a lack of natural gas infrastructure in the region has raised prices and pollution by forcing reliance on higher-cost resources like oil-fired power plants. The only major power generation resource with a less restrictive trend is nuclear, as six states recently modified or repealed nuclear moratoria to ease siting.
Motivation for opposing energy infrastructure permitting has included the well-known “not in my backyard” concerns, such as noise, construction disruptions, or land use conflicts. Interestingly, much opposition appears to come from perception, as much as substantiated negative effects. Relatedly, permitting resistance rationales increasingly appear to result from ideological opposition to particular energy sources. Finally, much opposition and most litigation of energy projects comes from non-governmental organizations, not the land owners directly affected. Altogether, this underscores the importance of permitting and siting reform that improves the quality of information to agencies and parties, ties decisionmaking to specific harms not speculative claims, limits standing to affected parties, and creates appeals processes for landowners to challenge obstructive local government laws and decisions. A key tension to overcome is that technology-agnostic legislation has been more likely to advance in states with one or more Republican chamber, yet environmental advocates resist “all-of-the-above” reforms.
Policies that reduce permitting and siting burdens are class I: they boost economic output and are increasingly key to emissions reductions. Permitting and siting policies that are restrictive on fossil development are not particularly effective at reducing emissions and often add considerable cost, granted costs vary widely depending on the nature of the policies and implementation. Effective fossil restrictions can range from class II to class IV policy, while ineffective ones actually increase emissions. The political economy of permitting and siting must overcome the lobby of entrenched suppliers, who seek to maintain competitive moats. An ironic example was incumbent asset owners funding environmental groups to oppose transmission infrastructure in the Northeast that would import emissions-free hydropower.
Electric Regulation
The power industry is at the forefront of energy cost concerns and decarbonization objectives. In the early 2020s, electric rates have risen most in Democratic states. These concerns reoriented progressives towards cost containment, even at the expense of climate objectives. In the 2024 election, cost of living concerns propelled Republicans to widespread victories as President Trump vowed to halve electricity prices. A year later, voter concerns over rising electricity rates in Georgia, New Jersey, and Virginia boosted Democrats in gubernatorial and public service commission (PSC) elections.
At the same time, electricity is arguably the most important sector for climate abatement given its emissions share and the indirect effects of electrifying other sectors, namely transportation and manufacturing. Ample pathways exist to reduce electric costs and emissions simultaneously, primarily by fixing profound government failure embedded in legacy regulation. Electric industrial organization shapes economic and climate outcomes, with market liberalization an advantage for both.
Electric regulation falls into two basic formats. The first is cost-of-service (CoS) regulation, where the role of government is to substitute for the role of competition in overseeing a monopoly utility. The alternative is for regulation to facilitate competition by using the “visible hand” of market rules to enable the “invisible hand” to go to work.
CoS regulation historically applied to power generation, though about a third of states enacted restructuring to introduce competition into power generation and retail services, in response to rising rates and the recognition that these are not natural monopoly services. Nearly all transmission and distribution (T&D) historically and today remains under CoS regulation. Importantly, CoS regulation motivates a utility to expand the regulated rate base upon which it earns a state-approved return. Generally, the main sources of cost discipline problems in the power industry stem from its CoS regulation segments: transmission, distribution, and the portion of generation that remains on CoS rates.
Generally, restructured jurisdictions see greater innovation and downward pressure on the supply portion of customer bills. The economic performance of restructuring is highly sensitive to the quality of implementation. This includes the quality of wholesale energy price formation and capacity market design. It also includes various elements of retail choice implementation. They have also seen improved governance, whereas CoS utilities are prone to cronyism and corruption given the inherent incentives of their business model. Competitive wholesale and retail power markets hold cost and emissions advantages through several mechanisms:
- Markets accelerate capital stock turnover when it is economic. With the brief exception of nuclear retirements, new entry is dominated by zero emission resources or high efficiency gas plants that displace legacy plants with higher emissions rates. Markets usher in new entry and induce retirements in response to economic conditions. Last decade saw markets outperform in the coal-to-gas transition, and this decade with advances in wind, solar, and storage economics. Texas, the most thoroughly restructured state, leads the country in solar, wind, and energy storage additions while placing second in gas additions. A review of restructuring found that competition worked as intended, facilitating new, low-cost entry while “driving inefficient, high-cost generation out of the market.” A new paper evaluating generator-level data found that from 2010–2023, regulated units were 45% less likely to retire than unregulated units.
- Markets encourage power plant operating efficiencies. Competitive generators adopt technologies and practices that use fuel more efficiently and improve environmental performance. The introduction of competition caused nuclear generators to adopt innovative practices to reduce refueling outage times, boosting operating efficiency by 10%. One study found 9% higher operating efficiencies in the thermal power fleet in restructured states. By contrast, CoS utilities sometimes engage in uneconomic operations because they are financially indifferent to market signals, resulting in overoperation of the fossil fleet.
- Markets reflect customer preferences, including clean power. Footprints with retail choice have seen much higher popularity of voluntary clean power programs. Competition lowers the “green premium” and customer choice allocates it equitably. This is critical as the willingness to pay for clean power varies enormously across customers. Notably, most growing power customers are large companies with ambitious corporate emissions reductions targets, which explains their commercial interest in advancing consumer choice.
- Markets better integrate unconventional resources, namely storage, wind, solar, and demand flexibility. The central planning of monopoly utilities struggles to account for the profile of variable (e.g., wind and solar) and use-limited (e.g., storage) resources. Demand flexibility is valuable to integrate more variable supply sources. Wholesale and retail competition are the only structural pairings that have elicited substantial shifts in demand in response to price signals, because they align the incentives of retailers and end-users to reduce consumption during high price periods.
- Markets induce lower-cost environmental compliance and better environmental lobbying behavior. Restructuring reoriented the incentives to influence and comply with public policy. Notably, competitive enterprises pursue more innovative, lower-cost compliance pathways that tend to deepen abatement. Monopoly utilities have a track record of lobbying for higher cost environmental laws. For example, monopolies have a preference for command-and-control regulation that pads their rate base, and have opposed market-based policies like the 1990 Clean Air Act amendments.
Electric cost increases are multifaceted, prompting many misdiagnoses that blame markets for non-market problems. Utilities have begun pushing campaigns in restructured states to revert back to CoS regulation, whereas the growing consumer segment – namely data centers and industrials – are organizing campaigns to expand consumer choice. Independent economic assessments warn against a return to CoS regulation, and instead encourage state regulators to implement restructuring better. This includes better market design, consumer exposure to wholesale prices, and effective coordination with transmission investment.
T&D costs, generally, are the core driver of electricity cost pressures nationwide. Over the last two decades, utility capital spending on distribution has increased 2.5 times while nearly tripling for transmission. This reflects profound flaws in CoS regulation of T&D, resulting in overinvestment in inefficient infrastructure and underinvestment in cost-effective infrastructure. This projects to worsen, given T&D expansion needed to meet grid reliability criteria as a result of aging infrastructure, turnover in the generation fleet, and load growth.
T&D expansion is also central to abatement. Even partial transmission reforms can reduce carbon dioxide emissions by hundreds of million of tons per year. This explains why progressives have made reforms that expand transmission a top priority. This needs to be reconciled with the cost concerns of consumers and conservatives to result in durable policy. Consumers and conservatives have a budding transmission agenda rooted in upgrading the existing system, removing barriers to voluntary transmission development, using sound economic practices for mandatorily planned transmission, streamlined permitting and siting, and improved governance. A particularly promising frontier is reforms to enhance the existing system, given the expedience of their cost relief and consistency with a Trump administration directive.
Recent federal regulatory actions have demonstrated bipartisan willingness to improve transmission policy and the related issue of interconnection, which has emerged as a major cost and emissions issue. In 2023, FERC passed Order 2023 on a bipartisan basis to reduce barriers to new power plants trying to interconnect to regional transmission systems. Subsequent reforms were motivated by a coalition of consumer groups and the center-right R Street Institute. In 2024, FERC passed Order 1920-A on a bipartisan basis to improve economic practices in regional transmission development. EPRA, a gamechanger for interregional transmission development, passed the Senate with bipartisan support in 2024.
Demand growth has sparked reliability concerns over tight supply margins and recently put upward pressure on wholesale market prices. However, states with the greatest price decreases typically had increasing demand from 2019 to 2024 (Figure 3). This shows the importance of infrastructure utilization on electric rate pressures, as many areas had supply slack previously. The past may not be prologue. Emerging conditions show supply-constrained scenarios where marginal generation and T&D costs increase steeply to meet new load increase. The Energy Information Administration observes steady retail price increases and projects further rises to exceed inflation.
Source: Wiser et al., 2025.
In an era of resurgent power demand growth, the states poised to keep rates and emissions down have wholesale competition, retail competition, efficient generator interconnection processes, economical T&D practices, and low permitting and siting barriers. The only state that reasonably accomplishes all of these is Texas, which is experiencing the most commercial interest among competitive suppliers and growing power consumers. Texas has experienced industry-leading clean energy investment and earned the distinction of Newsweek’s “greenest state” in 2024.
All aforementioned electric reforms are considered class I policy. Despite cost-reduction appeal, power industry reforms have proven challenging for two reasons. First, reforms are highly technical in nature and face limited state capacity among legislative advisors and technocratic agencies, namely PSCs and FERC. For example, recent FERC and PSC activities reveal that these entities do not have the bandwidth or expertise to properly implement existing transmission policy, much less reform it. Secondly, reforms face strong resistance from incumbent utilities who hold concentrated interests in the status quo, creating a strong lobbying incentive. By contrast, the beneficiaries of reform, especially consumers, are dispersed interests that do not organize as effectively as a lobbying force.
Although the Texas electricity experiment and associated federal power market reforms under President George W. Bush is a conservative legacy, most restructured states are progressive. This reflects significant bipartisan historic appeal. However, traditional conservatives have sometimes conflated pro-utility positions as the “pro-business” position, while it is unclear whether right-wing populist influences will catalyze pro-market reforms by challenging the status quo or retrench monopoly utility interests based on technocratic market skepticism (e.g., Project 2025). CoS utilities also commonly oppose cost-effective T&D reform, especially vertically-integrated utilities, which is consistent with their financial incentives to expand rate base and deter lower-cost imports from third parties. Nonetheless, the political economy of bipartisan electric regulatory reform remains promising, given voters’ prioritization of reducing electricity costs.
Public Spending
Government spending occurs through direct spending outlays or indirect spending through tax expenditures. Spending takes the form of industrial policy or innovation policy. The economics literature is historically critical of industrial policy, while positive literature on industrial policy usually conflates it with innovation policy. A distinguishing element is that innovation policy selects policy instruments suited to specific market failures, namely the positive externalities of knowledge spillovers and learning-by-doing. These generally apply to research and development (R&D) and early stage technologies, including those in demonstration stage and infant industries that have not achieved economies of scale.
Predictably, progressives have been consistent backers of robust innovation policy, while conservatives typically scrutinize such expenses closely. Although differences of opinion exist on optimal funding levels, historically conservatives and progressives have agreed on a role for the government in supporting R&D. There is also a history of good governance agreement, such as a joint project between the Center for American Progress and the Heritage Foundation in 2013 on improving the performance of the national lab system. Improving outcomes-based Department of Energy program performance may have broad appeal, including better performance metrics, stronger linkages to private sector needs, and program reevaluation to determine government investment phase-out. Improvements to state capacity are paramount in this regard.
Conservatives are often critical of public spending on infant industry, where government failure can outweigh market failure. For example, policymakers often struggle to identify when to end industry support, while industry engages in rent-maintenance behavior even after it has achieved maturity. Historic evidence indicates that direct subsidies and tax exemptions for infant energy industry continue well after the targeted technologies mature. Conservative and progressive scholars have historically framed the merits over subsidies for infant industry as a debate over government versus market failure.
Since innovation policy targets non-climate market failures (e.g., knowledge spillovers) it may have a high static abatement cost. However, it is an inexpensive abatement policy when accounting for dynamic effects, because of induced innovation and learning-by-doing. Importantly, innovation policy holds massive climate benefits, because achieving abatement cost parity between clean and emitting resources is central to clean technology market adoption. Efficient R&D policy can be classified as class I policy, because the upfront cost of the policy is outweighed by long-term cost savings. Demonstration and infant industry support falls into class II-III range, depending on its implementation, and often exhibits substantial durability.
In recent years, climate-minded conservatives have shown stronger inclinations of public spending for innovation policy. However, there is a stark difference between conservatives and right-wing populism on innovation policy. Conservatives note that the adverse consequences of Department of Government Efficiency’s “gutted, ineffective government” approach to the Department of Energy is inconsistent with limited, effective government practice. The economic self-interest benefits of innovation policy may induce a course-correction with MAGA, which has not deliberately targeted innovation policy insomuch as sacrificing it amid a rash government downsizing exercise.
In contrast to innovation policy, industrial policy aims to directly promote a given industry, typically using mature technology, with interventions untethered to any underlying market failure (e.g., negative emissions externality). This generally takes the form of public spending on mature industries. For decades, traditional conservatives and climate-minded conservative scholars have been critical of green industrial policy for carrying high costs with modest emissions reductions.
The most relevant case study in climate industrial policy versus innovation policy is the Inflation Reduction Act (IRA) of 2022. IRA represented the “largest federal response to climate change to date.” It consisted mostly of subsidies for mature technologies, especially wind, solar, and electric vehicles (EVs). It also contained subsidies for infant industry. IRA was passed exclusively by Democrats, with Republicans voicing concerns over its cost. Republicans then passed the One Big Beautiful Big Act (OBBBA) in 2025, which phased-out subsidies for mature technologies, but generally retained those for infant industry. This underscores the political durability of innovation policy and the fragility of industrial policy.
A broader debrief on IRA and OBBBA reveals:
- Disregard for cost considerations preceded passage of the IRA. All known ex ante modeling of IRA’s abatement benefits before it passed ignored costs. This left Congress unequipped to weigh the merits and tradeoffs of the policy. A simplistic abatement cost technique in 2022 yielded a cost of $72/tonne for the renewable energy subsidies. A more sophisticated modeling exercise in 2023 projected an average abatement cost of $83/tonne. IRA could have been identified as a high abatement cost policy (class IV) before it passed. Before passage, R Street Institute analysis suggested meager additionality from subsidies and identified permitting and electric regulation flaws as the determining factors of energy emissions trajectories, yet Congress neglected those reforms.
- IRA abatement cost estimates escalated sharply after passage. The total abatement cost of IRA subsidies to taxpayers rose from $336/tonne in 2024 to $600/tonne in 2025. The initial 2022 IRA renewables subsidy cost estimate of $72/tonne rose to $142/tonne in 2024 and $208/tonne in 2025. The EV subsidy came in at $1,626/tonne. It is possible that this is understated, since the direction of the emissions effect of EV subsidies may depend on recipient qualifications, especially when accounting for the behavioral tendencies of EV adopters. The subsidies also undermined developer cost reduction in two ways: 1) motivated development in the least efficient areas and 2) weakened incentives for innovation that lowers costs, which translates into long-term cost increases relative to an unsubsidized baseline.
- Government failure precluded most of the anticipated climate benefits of the IRA. IRA abatement was overstated in 2022, because models understated artificial constraints on the core abatement driver: wind and solar deployment. The Energy Information Administration’s renewables projections in 2025, which reflected IRA subsidies, were close to their no-IRA estimates from 2022. Risk, not cost, has consistently been the barrier to wind and solar. A Brookings Institution analysis found that artificial barriers to entry were the leading causes of wind and solar project cancellations from 2016-2023, whereas the lowest cause was “lack of funding.” Renewables subsidies primarily constituted a wealth transfer from taxpayers to suppliers. One analysis suggested 80-90 percent of clean energy backed by the IRA would have occurred anyways. An S&P Global forecast projected OBBBA to cause a 15 percent decline in wind, solar, and battery storage capacity by 2035.
- Wind, solar, and EV tax credit phaseouts should lower costs and increase economic productivity, despite increasing electricity prices. Price and cost are related, but not the same thing. The phase-out of subsidies under OBBBA will put upward pressure on electricity prices. However, it will likely lower costs by restoring dynamic cost management incentives and removing distortions so investment reflects economic fundamentals. Electricity subsidies shift cost burdens from power generators and ratepayers to taxpayers. Because taxpayer funding is expensive – tax collection imposes considerable deadweight loss on the economy – the net effect of taxpayer subsidies tends to shrink economic output. The Tax Foundation projected that IRA would reduce U.S. gross domestic product by 0.2 percent, while OBBBA would increase long-run GDP by 1.2 percent, granted energy tax credits were only one factor in these analyses.
The takeaway from IRA and OBBBA is that subsidies for mature technologies are high cost, likely to erode social welfare, and not politically durable. Efficient public spending for RD&D, however, enhances social welfare and falls in the Overton Window due to its value for economic self-interest. Late-stage infant industry is at the fringe of the Overton Window. It is the area where conservative and progressive scholars have historically had contrasting views on whether market failure outweighs government failure, yet political outcomes have largely supported infant industry.
Generally, the literature finds strong evidence of opportunity cost neglect in public policy, which “creates artificially high demand for public spending.” The IRA was a case-in-point. Meanwhile, the opportunity cost of public spending is rapidly rising given the dire fiscal trajectory of the United States. In 2025, moderate experts emphasized a pivot away from unsustainable and ineffective “Green New Deal thinking” for clean technology subsidies in favor of an innovation-driven strategy.
Takeaways
This analysis finds chronic flaws of cost considerations in ex ante policy analysis. Many medium and high-cost policies have passed without any robust accounting of costs at all (e.g., IRA, fuel bans). Interventions with cost-benefit analysis have had a tendency to underestimate costs (e.g., regulation). These flaws contribute to public misconception and play into political economy dynamics that tend to incent policies with hidden costs over those with transparent ones.
High-cost policies have typically only been enacted by progressive governments and have come under greater scrutiny as energy costs escalate. This calls their social welfare effects and durability into question. It has cast climate action in the public eye as requiring deep economic sacrifice.
Conservatives have been hesitant to engage on climate policy outright, largely over dire economic tradeoff perceptions. Such concerns have instigated a conservative backlash to climate policy, including to policies that are compatible with U.S. economic interests. This has been exacerbated by right-wing populism, which often strays from limited government conservatism in pursuit of cultural identity objectives. For example, in a 2024 piece promoting energy affordability, the Heritage Foundation correctly attributed cost increases to renewable energy mandates, but incorrectly presumed that a broad shift towards renewable energy and away from fossil fuels would always increase costs.
High abatement cost policies not only risk reducing aggregate social welfare, but they create distributional concerns. Policies that raise energy costs tend to be regressive. This has challenged the social justice narrative of progressives, prompting a rethink by progressive leaders to take a “cost-first approach to [the] clean energy transition.” Although subsidies are a common response to lower burdens on low-income households, the most popular green subsidies pursued have exacerbated distributional concerns. Specifically, renewables subsidies favored by progressives have been challenged by conservatives as “green corporate welfare.” Progressives have also faced criticism for EV tax credits for disproportionately benefiting wealthy households.
Encouragingly, negative- and low-cost policies comprise a rising share of the abatement curve. The Overton Window for pursuing such policies has grown remarkably for “abundance progressives” and conventional conservatives. However, populist subsets within both movements challenge the potential for political alignment. Enacting negative-cost policies also faces the collection active problem of dispersed beneficiaries versus a concentrated incumbent supplier lobby favoring the status quo. Mobilizing consumer and taxpayer groups is an underappreciated strategy to enact these policies.
This analysis is far from comprehensive. A notable omission from this paper is transportation policy, the largest GHG sector in the U.S. A scan of the transportation literature underscores major abatement potential for negative and low-cost policies, including reducing government barriers to efficient heavy-duty transportation like railways, shipping, and heavier trucking. Further, the electrification of transportation requires extensive fixes to government failure, such as liberalizing markets to enable competitive charging infrastructure, which lowers costs. The merits of innovation and GHG transparency policy, previously discussed, also appear to hold promise for transportation applications such as aviation fuel. The transportation sector has also been the target of GHG regulation, mostly in progressive states, which warrants close assessment of costs. For example, one study identified a vast abatement cost range for fuel standards ($60-$2,272/tonne).
A shortcoming of this analysis is that it only characterizes costs by their efficiency (i.e., $/ton). Political decisions are highly sensitive to aggregate cost and its visibility to the public, which our taxonomy does not characterize. It is possible that efficient, transparent, and higher aggregate cost policies (e.g., C&T) fare less favorably in some political settings than inefficient, opaque, and sometimes lower aggregate cost policies (e.g., RPS solar carveouts).
Despite the limitations of this analysis, the sample of policies evaluated is sufficient to support the thesis. That is, a retooled climate policy agenda that prioritizes cost considerations should elevate social welfare and achieve greater abatement by selecting more durable policies.
Conclusion
Abatement costs have huge bearing on whether climate policies benefit society, their likelihood of passage, and whether they prove politically durable. Most abatement need not come from dedicated climate policy, per se, but rather sound economic policy that carries deep climate co-benefits. Chronic disregard for cost considerations has led to an overselection of high-cost policies and underpursuit of low- and negative-cost policies. This has undermined policy durability and exacerbated political polarization over climate change abatement.
This paper finds extensive abatement opportunities within negative-cost policies. These largely constitute fixes to government failure and include permitting, siting, and power regulation reforms. This analysis also finds considerable low-cost policies that are compatible with U.S. economic self-interests. These policies primarily spur voluntary private sector abatement through efficient innovation policy and GHG transparency.
We offer three sets of recommendations moving forward for influencers of the climate policy agenda:
- Focus on results. Climate change abatement is a function of global GHG concentrations. Too much attention pursues symbolic objectives, like preventing fossil fuel infrastructure. This tends to undermine abatement goals and impose high costs.
- Emphasize cost considerations in policy agenda setting, formulation, and maintenance. Negative abatement cost policies should take top priority, with an emphasis on mobilizing beneficiaries. Robust cost-benefit analyses should precede all cost-additive policies and be reconducted periodically to guide policy adjustments.
- Prioritize quality state capacity. The net benefits of abatement policies are sensitive to government capacity and performance. Public management is in great jeopardy in an era of institutional decay. Negative-cost policies are often highly technocratic and require sufficient staffing expertise and accountable management at public institutions like DOE, FERC, PSCs, and permitting and siting agencies.
In an era of energy affordability precedence, a reset climate agenda should anchor itself in good policy basics. That is, a sober-minded return to results-driven, net-benefits prioritized policy. This should improve the durability of climate policy and ensure it enhances social welfare. Executing reforms well requires a recommitment to improving the quality of institutions as much as the policy itself.