Increasing the Value of Federal Investigator-Initiated Research through Agency Impact Goals
American investment in science is incredibly productive. Yet, it is losing trust with the public, being seen as misaligned with American priorities and very expensive. To increase the real and perceived benefit of research funding, funding agencies should develop challenge goals for their extramural research programs focused on the impact portion of their mission. For example, the NIH could adopt one goal per institute or center “to enhance health, lengthen life, and reduce illness and disability”; NSF could adopt one goal per directorate “to advance the national health, prosperity and welfare; [or] to secure the national defense”. Asking research agencies to consider person-level or economic impacts in advance helps the American people see the value of federal research funding, and encourages funders to approach the problem holistically, from basic to applied research. For almost every problem there are different scientific questions that will yield benefit over multiple time scales and insight from multiple disciplines.
This plan has three elements:
- Focus some agency funding on measurable mission impacts
- Fund multiple timescales as part of a single plan
- Institutionalize the impact funding process across science funders
For example, if NIH wanted to reduce the burden of Major Depression, it could invest in a shorter time frame to learn how to better deliver evidence-based care to everyone who needs it. At the same time, it can invest in midrange work to develop and test new models and medications, and in the decades-long work required to understand how the exosome influences mood disorders. A simple way to implement this approach would be to build on the processes developed by the Government Performance Results Act (GPRA), which already requires goal setting and reporting, though proposals could be worked into any strategic planning process through a variety of administrative mechanisms.
Challenge and Opportunity
In 1945, Vannevar Bush called science the ‘endless frontier’, and argued funding scientific research is fundamental to the obligations of American government. He wrote “without scientific progress no amount of achievement in other directions can insure our health, prosperity, and security as a nation in the modern world”. The legacy of this report is that health, prosperity, and security feature prominently in the missions of most federal research agencies (see Table 1). However, in this century we have begun to drift from his focus on the impacts of science. We have the strange situation where our enterprise is both incredibly productive, and losing trust with the public, viewed as out of touch or misaligned with American priorities. This memo proposes a simple solution to address this issue for federal funding agencies like NIH and NSF that largely focus on extramural investigator-initiated research. These are research programs where the funding agency signals interest in specific topics and teams of scientists submit their research plans addressing those topics. The agency then funds a subset of those plans with input from external scientific reviewers.
This funding approach is incredibly productive. For example, NIH funds most of the pipeline for the emerging bioeconomy, which accounts for 5.1% of our GDP. From 2010 to 2016, every one of the 210 new entities approved by the FDA had some NIH funding. And yet, there appears to be a disconnect between our funding strategy and the public interest focus of the Endless Frontier operationalized through our federal science agency missions for investigator initiated research.
A fundamental driver of this disconnect might be a slight misalignment of the incentives of academic scientists, who are rewarded for novelty and scientific impact, with the broader public interest. Our federal agencies are highly attuned to scientific leaders, and place equal or even greater weight on innovation (novelty plus scientific impact) than real world impact. For example, NSF review criteria place equal weight on intellectual merit (‘advance knowledge’) and broader impacts (‘benefit society and contribute to the achievement of specific, desired societal outcomes’). NIH’s impact score of new applications is an ‘assessment of the likelihood for the project to exert a sustained, powerful influence on the research field(s) involved’ [emphasis mine], which is only part of the agency’s mission. The practical implications of this sustained focus away from the impact portion of agencies missions become apparent in figure 1, showing tremendous spending in health research unrelated to a key public interest measure like lifespan, especially when compared to other nations’ health research spending.
Perhaps the realization that the federal research investment is not strongly linked to their mission impact is one reason why American science has been slowly losing public trust over time. Among the people of 68 nations ranking the integrity of scientists, Americans ranked scientists 7th highest, whereas we ranked scientists 16th highest in our estimation of them acting in the public interest. And this is despite the fact that the American investment in science is many times higher than the 15 nations who rated scientists more highly on public interest. A more accurate description of our 21st century federal science enterprise might be the ‘timeless frontier’, where our science agencies pursue cycles of funding year in and year out, with their functional goal being scientific changes and their primary measure of success being projects funded. Advancing the economy, health, national defense, etc., are almost incidental benefits to our process measures.
We can do better. In 2024, the National Academy of Medicine called out the lack of high level coordination in research funding. In 2025, the administration has been making drastic cuts and dramatic changes to goals and processes of federal research funding, and the ultimate outcome of these changes is unclear. In the face of this change, Drs. Victor Dzau and Keith Yamamoto, staunch champions of our federal science programs, are calling for “a coherent strategy […] to sustain and coordinate the unrivaled strengths of government-funded research and ensure that its benefits reach all Americans”.
We can build on the incredible success of the federal science enterprise – inarguably the most productive science enterprise in all history. The primary source of American scientific strength is scale. American funding agencies are usually the largest funders in their space. I will highlight some challenges of the current approach and suggest improvements to yield even more impactful approaches more closely aligned with the public interest.
The primary federal funding strategy is broad diversification, where our agencies fund every high scoring application in a topic space (see FAQs). Further, federal science agencies pay little attention to when they expect to see a fundamental impact arising from their research portfolio. For example, a centrally directed program like the Human Genome Project can lead to breakthrough treatments decades later, but in the meantime, other research that generates improvements on faster timescales could have been coordinated, such as developing conventional drug treatments, or research to optimize quality and delivery of existing treatment.
And yet, the breadth and complexity of broad diversification makes it easy to cherry pick successes. This is a strategic issue, and is bigger than the project selection issues highlighted in the earlier discussion about review criteria. When research funding agencies make their pitch for federal dollars they highlight a handful of successes over tens of thousands of projects funded over many years. They ignore failures, the time when investments were made, and time to benefit. With the goals and metrics we have in place, it is simply too hard to summarize progress in any other way.
Overly diversified science funding supports both good Congressional testimony and bad strategy. If your problem happens to fall into a unicorn space of success, there is a lot to celebrate. But most problems do not, and we experience inconsistent returns. We need to define the success of research funding more precisely, in advance, and in ways that more obviously align with the public interest.
Plan of Action
If we tweak our funding strategy to focus on societal impacts, we can move to a more impactful science enterprise, and help regain public support for science funding. We can focus federal research funding on effective answers to difficult problems demanding both urgency and short term improvements, and fundamental discoveries that may take decades to realize. My solution and implementation actions for agencies, and potentially Congress, are described below.
Recommendation 1. Focus some agency funding on measurable mission impacts.
We should empower our science agencies to step away from broad diversification as the predominant funding strategy, and pursue measurable mission impacts with specific time horizons. It can be a challenge for funders to step away from process measures (e.g. projects or consortia funded) and focus on actual changes in mission impact.
Ideally, these specific impacts would be broken into measurable goals that would be selected through a participatory process that includes scientific experts, people with lived experience of the issue, and potential partner agencies. I recommend each agency division (e.g. an NSF Directorate) allocate a percentage of their budget to these mission impact strategies. Further, to avoid strategic errors that can arise from overwhelming power of federal funding to shape the direction of scientific fields, these high level funding plans should be as impact focused as possible, and avoid steering funding to one scientific theory or discipline over another.
Recommendation 2. Fund multiple timescales as part of a single plan.
Research funders need to balance their investment portfolios not only across problem areas, but over time. Complex challenges will often require funding different aspects of the solution on different timelines in parallel as part of a larger plan. Balancing time as well as spending allows for a more robust portfolio of funding that draws from a broader array of scientific disciplines and institutions.
Note, this approach means starting lines of research that may not lead to ultimate impact for decades. This approach might seem strange given our relatively short budget cycles, but is very common in science, where projects like the Human Genome initiative, the Brain Initiative, or the National Nanotechnology Initiative, have all exceeded a single budget cycle and will take years to realize their full impact. These kinds of efforts require milestones to ensure they stay on track over time.
Recommendation 3. Institutionalize the impact funding process across science funders.
Our research enterprise has become oriented around investigator-initiated, project-based awards. Alternative funding strategies, such as the DARPA model, are viewed as anomalies that must require completely different governance and procedures. These differences in goals are unnecessary. A consistent focus on impacts and strategy in funding across agencies will help the scientific community become more aware of the time to benefit of research, help underscore the value of research investment to the American public, and help research agencies collaborate among themselves and with their partner agencies (e.g. NIH collaborates more closely with CMS, FDA, etc.).
In short, institutionalizing this process can lead to greater accountability and recognition for our science enterprise. This structure allows our funders to report to the public progress on specific goals on predetermined and preannounced timelines, rather than having to comb through tens of thousands of independent funding decisions and competing strategies to find case studies to highlight. In this way, expected and unexpected scientific results, and even operational challenges, can be discussed within an impact framework that clearly ties to the agency mission and public interest.
Example of Planning using an Impact Focus
Here is an example of a mission impact goal Reducing the Burden of Major Depressive Disorder that could be put forth by the National Institute of Mental Health (NIMH), and the process to develop it.
Commence Inclusive Planning: NIMH brings together experts from academia, clinical care, industry, people impacted by depression, and FDA and CMS to develop measures, timelines and funding strategies.
Develop Specific Impact Measures: These should reflect the agency’s impact portion of their mission. For example, NIH’s mission impact of “enhance health, lengthen life, and reduce illness and disability” requires measuring impact on human beings. Example measurement targets could include:
- Reduced incidence of Major Depressive Disorder
- Increased productivity (e.g. days worked) of people living with Major Depressive Disorder
- Reduced suicide rates
Fund Multiple Time Scales: Designate time scales in parallel as part of a comprehensive strategy. These different plans would involve different disciplines, funding mechanisms, and private sector and government partners. Examples of plans working at different timescales to support the same goal and measures could include:
- 10 year plan: Increase utilization of evidence based care
- 15 year plan: Develop and implement new treatments
- 30 year plan: Determine how the exposome causes and prevents depression, and how can be changed
- It is likely that NIMH has already obligated funds to projects that support one of these plans, though they may need additional work to ensure that those projects can directly tie to the specific plan measures.
Implementation Strategies for Impact Goals
Each federal funding agency could allocate a percentage of their budget to these and other impact goals. The exact amount would depend on the current funding approach of each agency. As this proposal calls for more direct focus on agency mission, and not a change in mission, it is likely that a significant percentage of the agency’s current budget already supports an impact goal on one or more of its time scales.
For an agency heavily weighted towards project based funding of small investigator teams, like NIH, I would recommend starting with a goal of 20% of their budgets set towards impact spending and consider increases over time. Other agencies with different funding models may want to start in a different place. Further, I would recommend different goals and targeted funds for each major administrative unit, such as an institute or directorate.
All federal funders already engage in some form of strategic and budget planning, and most also have formal structures for engaging stakeholders into those planning decisions. Therefore, each agency already has sufficient authorities and structures to implement this proposal. However, it is likely that these impact goals will require collaboration across agencies, and that could be difficult for agencies to efficiently conduct by themselves.
Additional support to make this change could come from Congressional Report language as part of the budget process, through interagency leadership from the White House Office of Science and Technology, or through the Office of Management and Budget. For example, the Government Performance Results Act (GPRA) already requires agency goal setting, reporting and supports cross agency priority goals. That planning process could easily be adapted to this more specific impact focus for research funding agencies, and reporting on those goals could be incorporated into routine reporting of agency activities.
Conclusion
We are living through a massive disruption in federal research funding, and as of the fall of 2025, it is not clear what future federal research funding will look like. We have an opportunity to focus the incredibly productive federal research enterprise around the central reasons why Americans invest in it. We can meet Bush’s challenge of the Endless Frontier simply by clearly defining the benefits the American people want to see, and explicitly setting plans, timing and money to make that happen.
We can call our shots and focus our science funding around impacts, not spending. And we can set our goals with enough emotional resonance and depth to capture both the interests of the average American, and the needs of scientists from different disciplines and types of institutions. We already have the legal authorities in place to adopt these techniques, we just need the will.
Inadvertently, the huge scale of federal funding could lead to a monopsonistic effect. In other words, NIH’s buying power is so large, if NIH does not fund a specific type of research, people may stop studying it. This risk is highest within a narrow scientific field if there is a bias in grant selection. A well publicized example being NIH’s strong funding preference to one theory of Alzheimer’s Disease to the diminishment of competing theories, which in turn influenced careers and publication patterns to contribute to that bias.
To increase the real and perceived benefit of research funding, funding agencies should develop challenge goals for their extramural research programs focused on the impact portion of their mission.
Americans trade stocks instantly, but spend 13 hours on tax forms. They send cash by text, but wait weeks for IRS responses. The nation’s revenue collector ranks dead last in citizen satisfaction. The problem isn’t just paperwork — it’s how the government builds.
While it seems that the current political climate may not incentivize the use of evidence-based data sources for decision making, those of us who are passionate about ensuring results for the American people will continue to firmly stand on the belief that learning agendas are a crucial component to successfully navigate a changing future.
In recent months, we’ve seen much of these decades’ worth of progress erased. Contracts for evaluations of government programs were canceled, FFRDCs have been forced to lay off staff, and federal advisory committees have been disbanded.