The Counterinsurgency Debate

What does the research tell us?

Eleven years ago this coming month, a coalition of armed forces led by the United States moved decisively against Saddam Hussein’s forces, declaring victory just over one month after they had invaded the country. The insurgency that emerged shortly afterward left coalition forces embattled for nearly a decade.

With U.S. forces now out of Iraq and readying themselves to withdraw from Afghanistan in as little as ten months, most Americans desire to move on from these deadly, divisive, and costly conflicts. Yet, important questions about the effectiveness of counterinsurgency programs carried out during both conflicts remain unanswered, which we ignore at our own peril.

Over the course of both conflicts, coalition forces and government civilians carried out a broad range of counterinsurgency programs, some for the first time. Determining whether these initiatives were effective in reducing violence levels and, if so, how, is not just a matter of writing accurate history. The lessons learned have implications for U.S. counterinsurgency programs still in effect in Afghanistan and, perhaps more importantly, are immediately useful to the Iraqi and Afghan governments, which now bear responsibility for their countries' national security. As we have seen this year, instability in these countries breeds opportunity that can be exploited by violent political organizations like Al-Qaeda, which remain intent on carrying out attacks against the United States and its allies.

In both conflicts, the United States sought to combat insurgents by improving local economic conditions. A variety of local development programs were carried out under the Commander's Emergency Response Program (CERP). Billions of dollars were committed to major infrastructure revitalization projects, and millions more were dedicated to restarting idled factories.

From a counterinsurgency perspective, such programs were believed to reduce violence levels by serving two functions. First, local citizens who benefitted from these programs were expected to share information with U.S. forces about the identity of members of the insurgency and details about their operations. Second, many of these programs created jobs for local citizens. Employed citizens, the logic held, are not the type to support or participate in insurgent violence. Unemployment, the White House’s 2005 National Strategy for Victory in Iraq described, simultaneously “fuels popular dissatisfaction and may generate sympathy for the insurgency” and makes “Iraqis more vulnerable to terrorist or insurgent recruiting.”

The logic of this approach is intuitively compelling, but were these expectations actually met? Working with detailed data on these conflicts, academics have begun to tease out a far more complicated story of what effectively reduces violence when governments face insurgencies. Researchers from the University of California San Diego, Princeton, and Stanford, for instance, asked whether working men rebel. What they found defied expectation: analyzing periods of insurgent violence in Iraq, Afghanistan, and the Philippines, they discovered that areas of those countries with higher employment levels tended to experience higher levels of violence.

In seeming contradiction, researchers at the London School of Economics and Political Science (LSE) have since determined that CERP programs that created local job opportunities in Iraq were effective in reducing violence levels. Can changes in the employment level of a country experiencing conflict both increase and decrease violence levels? The answer may be that they can and do—and that the key factor lies with who is actually employed.

Authors of the first study theorize that higher levels of violence may occur in areas with more available jobs because employed citizens may be less willing than their unemployed counterparts to provide information on insurgents. By this logic, an unemployed Iraqi struggling to provide for his children and who has information about insurgents operating in his neighborhood may be willing to sell that information to the government, in spite of possible retribution. For an Iraqi with a job who faces the same decision, however, the possible costs of being discovered and punished by insurgents may simply exceed the expected reward. Aggregate these cases by communities, and a pattern emerges in which neighborhoods with relatively high levels of employment generate less actionable intelligence than those with comparatively lower levels and thus suffer greater levels of insurgent violence.