Lessons From Operation
RESTORE DEMOCRACY

by Captain Michael W. Schellhammer, USAR
In the past several years, the Army intelligence community has grappled with how to define the parameters of Stability and Support Operations(formerly known as OOTW) and how intelligence should support them. Threat trends of the last decade indicate that Stability and Support Operations will become more frequent in the near future. Peacetime contingency operations in Grenada, Panama, and Somalia were milestones in the development of effective intelligence support. Recent operations in Haiti also provide valuable lessons on how to manage intelligence in the Stability and Support Operations environment. This article offers some ideas from the bottom-up on managing tactical intelligence during an unconventional conflict.

Background

I participated in Operation RESTORE DEMOCRACY as the S2 for the 450th Civil Affairs Battalion (Airborne), U.S. Army Reserve (USAR), from September to December 1994. Before activation of the battalion, the S2 personnel built a database containing civil-military information on Haiti. After activation, we used the database information to prepare our initial area assessment and preliminary intelligence preparation of the battlefield (IPB) products. As the 82d Airborne Division's Capstone program civil affairs battalion, we conducted all predeployment planning with the division and launched with it on the aborted airborne assault of 17 September. The 450th Civil Affairs Battalion, attached to Task Force (TF) Mountain under Joint Task Force (JTF)-180, later deployed to Haiti on 20 September.

IPB Focus and Products

The IPB and analysis techniques taught at the U.S. Army Intelligence Center work use them. An exhaustive IPB is an absolutely essential foundation for effective intelligence support. TF Mountain's area of operations focused on the capital, Port-au-Prince. The TF had a 360-degree focus, with no traditional battlefield geometry. Although we conducted preliminary IPB before deployment, the city was still unfamiliar. In addition, roughly one million civilians and three potentially hostile political groups occupied the city. Detailed IPB was the key to understanding this confusing battlefield. We used IPB products defined in FM 34-3, Intelligence Analysis, and civil affairs intelligence products from FM 41-10, Civil Affairs Operations(see Figure 1). When combined, they provided a solid picture of the key terrain and possible targets in the TF area.
The TF Mountain G2 provided sufficient information on the Haitian Army, the Forces Armee d'Haiti, (FAdH). We looked to outside agencies for detailed civil-military information. The U.S. Agency for International Development (USAID) had information on the Haitian government, economics, hydrology, and population statistics. The Defense Attache Office (DAO) in the U.S. Embassy provided population and threat information for the Port-au-Prince area.
The large number of civilians in the area, combined with the volatile political situation, made civil disturbance a primary concern. Large demonstrations not only posed a threat to U.S. soldiers and local stability, but also presented large targets for attack by opposition parties. Protecting the friendly population was part of our mission, and we included the civilians as an element of thinking green during IPB.
The likelihood of civil disturbance was one of our priority intelligence requirements. Civilian crowds could gather and disperse quickly with little warning. The crowds were difficult to predict, therefore named areas of interest (NAIs) focused on known population-gathering areas and march routes. Common NAIs were churches, key road intersections, and government buildings where demonstrations were likely. Looting became more frequent as the operation evolved, so we later added NAIs at food storage facilities and in the business district.
When preparing IPB overlays, remember that Stability and Support Operations require a greater level of incident-tracking detail than conventional operations. Several incidents, including violence, enemy human intelligence (HUMINT) collection, population intimidation, and criminal activity occurred within a two-block radius of Port-au-Prince within a few weeks. Low-level collection on individual murders, beatings, and looting is the core of the incident overlay, so the S2s must be ready to manage hundreds of individual incident reports.
Military operations in urban terrain also require a more detailed focus than conventional operations. Key terrain ranged from single buildings or city street intersections to areas of several square kilometers. We used 1:12,500 maps and map enlargements to analyze key terrain. The 1:50,000 scale maps were almost useless in the city; they were useful in the division area of interest outside Port-au-Prince. Friendly strike operations often targeted small buildings or city blocks. Terrain focus rapidly shifted between individual buildings and the entire city. Even with 1:12,500 maps, analysis of such small areas was difficult. We needed a tool to rapidly zoom in on specific areas, and zoom out again.

Tactical Reporting

Size, activity, location, unit, time, and enemy (SALUTE) reports from U.S. patrols were invaluable. Full SALUTE reports are critical to the success of tactical collection. In Stability and Support Operations, every soldier is a collector. Soldiers from the 450th Civil Affairs Battalion conducted daily patrols through Port-au-Prince and reported a wealth of information in SALUTE reports. On average patrols, Haitian citizens reported items such as enemy weapons caches, headquarters locations, or intimidation activity. One civil affairs team found hundreds of Haitian government election and military documents at Baby Doc Duvalier's palace in Petionville. Early in the deployment, soldiers often reported information in free-text spot reports that did not fully explain the activity or incident. The SALUTE format focused the soldiers' reporting to accurately reflect almost any type of observed activity. Once soldiers began using the correct format, we received complete reports of important enemy activities that received attention throughout the division.

All-Source Operations

Stability and Support Operations requires all-source intelligence. Each intelligence discipline has valuable contributions to make to the overall picture. In Haiti, HUMINT dominated the collection. HUMINT consistently provided reliable information on enemy and civilian attitudes and intentions. Counterintelligence (CI) teams targeted the opposition parties and subversive FAdH members. In addition to force protection, CI teams also conducted positive collection missions. Maneuver brigade commanders used their direct support CI teams in a wide variety of missions. CI teams investigated routine criminal activity and performed low-level psychological operations and civil affairs liaison missions.
Targetable imagery intelligence was in high demand down to the company level. Oblique photos of suspected threat locations or targets were invaluable to planning patrols or strike operations. The TF Mountain G2 had a large library containing hardcopy imagery of key facilities and terrain in Port-au-Prince; we distributed them on request. For other targets not located near key areas, the G2 requested specific imagery collection missions from theater assets. Remote sensor imagery was very useful for updating maps.
No intelligence discipline was completely reliable on its own. Correlated information from all sources provided an accurate battlefield picture. For example, HUMINT sources often indicated enemy weapons caches at various residences throughout the city. Since opposition groups often communicated over nonsecure systems, signals intelligence was also a valuable collection tool. However, only consistent all-source collection at the possible threat location could positively confirm or deny enemy activity.

Threat Analysis

Throughout the operation, threat analysis was a difficult process. The armed opposition parties presented our greatest threat. At least three well-armed political factions opposed President Jean-Bertrand Aristide and the U.S. intervention. Despite their expressed overt cooperation, they covertly planned attacks on U.S. forces and pro-Aristide civilians. Although many of their leaders were FAdH members, these hostile groups essentially had no doctrine or established patterns to integrate into analysis. We had no historical example for developing a threat model. All of the threat's previous activity involved intimidation against the unarmed populace of Port-au-Prince.
Threat group operations were exceptionally decentralized. The activities or plans of one cell in one area of Port-au-Prince usually had no relation to the activities of any other cell. Cells may have followed their recognized national leadership, some did not. The absence of threat tactics, doctrine, and training made predictive analysis extremely difficult. Several planned attacks on U.S. forces never took place for unknown reasons. Their unpredictability and decentralization contributed to difficult threat integration. Over time, limited trends became apparent. For example, their decentralization allowed us to place less value on threat leaders' covert plans for attack. When an information source reported a planned attack, we realized that the probability of local cells carrying out the attack was low. Friendly demonstrations also developed trends. Friendly political leaders were sometimes as unsuccessful as the opposition with controlling their followers. Over time, we were mindful that spontaneous gatherings or planned peaceful demonstrations had the potential to turn violent with no warning.

Common Picture

Achieving and maintaining common IPB or threat situation products was also difficult. Below the TF Mountain G2 level, no common picture of the battlefield or the threat existed. I compared IPB and threat information with several S2s in the JTF-180, and each of our battlefield pictures varied greatly. There appeared to be a gap between information available at the joint level and that at the battalion level. At the battalion level, we received the bottom-up feed from patrols, and pulled information from the JTF. We acquired most of our tactical threat information from civil-affairs team patrol reports, the TF Mountain G2, and the 519th Military Intelligence Battalion (CI). We received little tactical information from the JTF. An information push from the JTF could have ensured that all staffs had the same picture of key terrain, possible targets, and threat activities.

Automation

Combat information and intelligence must move faster than the threat. The threat in Port-au-Prince was more familiar with the terrain than were friendly forces. There was also an active threat HUMINT collection net throughout the city. To overcome such obstacles in the future, friendly forces must have the capability to rapidly report, analyze, and disseminate intelligence before the threat can move. Hand-carrying hardcopy reports or imagery through multiple staff levels is too slow for Stability and Support Operations. We need local area network dissemination and reporting capability. Analysis and dissemination must take minutes, not days.

Conclusion

There is no great mystery or secret code to intelligence in Stability and Support Operations, but there is a different focus than in conventional operations. The S2 should selectively apply standard intelligence tools and skills to read the battlefield. Follow the guidance in FM 34-3 and FM 34-7, Intelligence and Electronic Warfare Support to Low-Intensity Conflict Operations. Other lessons learned include-
The U.S. intervention in Haiti achieved rapid success through maneuver combined with effective intelligence support. By employing these lessons learned, we can win again and save lives.
Captain Schellhammer is currently the S2, 450th Civil Affairs Battalion (Airborne), USAR, in Riverdale, Maryland. He was an Active Component officer for eight years in infantry and intelligence. As a civilian intelligence research analyst, he is supporting the Army's Project Management Office, Intelligence Fusion, in McLean, Virginia. Readers can reach him at (703) 893-9095, through the PM Intelligence Fusion at DSN 235-8140, or the author's E-mail [email protected]