To win hearts and minds, focus on small projects, study finds

U.S. efforts to bring stability to Iraq and Afghanistan in recent years have focused less on killing insurgents and more on gaining the cooperation of the local population. But does this population-centered approach to counterinsurgency actually work?

A study published today (October 4, 2011) in the Journal of Political Economy finds evidence that it does.

The study, by economist Eli Berman (University of California, San Diego) and political scientists Jacob Shapiro (Princeton) and Col. Joseph Felter (Stanford), focused on the Commander's Emergency Response Program (CERP) in Iraq. CERP empowers junior officers to spend reconstruction dollars on small, local projects like digging wells or paving rural roads.

The theory driving CERP is that when government provides basic services, the population will turn towards the government and away from insurgents. With less popular support, insurgents will find it harder to operate, as they are vulnerable if the local population chooses to report on their activities. While this theory is a cornerstone of the U.S. counterinsurgency strategy currently practiced in Afghanistan, there has been no systematic empirical evidence showing that aid spending by foreign forces actually helps quell violence.

To test the theory, Berman and his colleagues analyzed the effect of nearly $3 billion in CERP spending in Iraq. Using geospatial data, they looked at whether changes in CERP spending in a given area were correlated with changes in rates of violence over a six-month period.

"We found that CERP projects, especially the smaller ones, are effective in reducing violent attacks on Coalition forces," Berman said. "The research provides strong evidence in favor of the population-centric approach to counterinsurgency currently practiced by the U.S. military, which necessarily involves economic and political development."

Despite the success of CERP, however, the program represents only a fraction of reconstruction spending in Iraq. Berman and his colleagues also tracked additional $26 billion in reconstruction spending that focused on larger, less community-oriented projects. The researchers did not find evidence that those larger projects reduced violence locally. Berman surmises that this is because such projects were administered by authorities that had less direct contact with the population, and thus less understanding of the needs of each community.

"Our research suggests that development money best reduces violence when projects are small and selected by consulting community members," Berman said. "Communities respond when you think small and local."