Development aid can exacerbate violence in war-torn countries

Although development aid is commonly seen as an important tool in the quest to reduce poverty in conflict-riven countries, new research co-written by a University of Illinois expert in development economics concludes that large-scale foreign aid programs can backfire and actually exacerbate violence in some long-running conflicts.

An analysis of data from 2002-06 of a major government-sponsored development program in the Philippines suggests that an influx of aid caused a spike in casualties in the country's long-running, low-level civil conflict but did not affect the levels of violence associated with armed criminal groups that lacked political motivation, says Benjamin Crost, a professor of agricultural and consumer economics at Illinois.

"It turns out that an unintended consequence of development aid is that, in the war for hearts and minds, insurgents have a strong incentive to pre-empt a successful program from coming in," he said.

The paper estimated the aid program's effect on violence by studying the eligibility for the program, which was restricted to the poorest 25 percent of municipalities in participating Philippine provinces. The eligibility criteria created a discrete cutoff point to compare the number of casualties in municipalities just above and just below the poverty threshold, Crost said.

"Normally, it's very difficult to estimate the effect of the flow of aid because it often comes in response to something," he said. "What makes our research unique is that we have this cutoff where aid was targeted. If you were to compare places that received a lot of aid with places that didn't, and find that the ones that got a lot of aid are more violent, that doesn't really tell you anything. What makes the paper unique is that we have this novel way of getting around it at the margins."

The implementing agency generated a poverty index and ranked municipalities from richest to poorest within each province, with only the poorest 25 percent of municipalities eligible to receive aid.

"That allowed us to really compare like with like," Crost said.

According to the paper, after the program's start, barely eligible municipalities experienced a large and statistically significant increase in casualties compared with barely ineligible ones.

In other words, a municipality at the 25th percentile for poverty was markedly more violent than one at the 26th percentile, even though they're both broadly similar in poverty levels and other factors that affect conflict.

"With all observable characteristics, they're basically the same – except that municipalities eligible for aid became markedly more violent," Crost said. "Comparing them allowed us to isolate the causal effect of aid on conflict."

According to the paper, the increase in violence was concentrated in the development program's early stages, before funds were disbursed and before eligible municipalities committed to participating in the program.

"The uptick in violence is all concentrated in the six months after local governments find out if they're going to get aid, but it's happening well before the resources actually comes through," Crost said. "So it's not simply a matter of the insurgents grabbing the money, because we see the violence go up before any money starts to flow to municipalities."

The effect was strongest for casualties suffered by government forces as a result of insurgent-initiated attacks. The hypothesis in the paper is that the developmental aid poses a threat to the insurgents, Crost said.

"The insurgents don't want the aid to come at all, because they don't want the government to have a success story," he said. "They don't want citizens to think that the government is trying to do something good for them, because that might undermine support for the insurgents. It all goes back to the idea of winning hearts and minds. If people think the government is doing something to help the population, then they'll stop supporting the insurgents."

What makes the Philippines interesting in its own right is that the country has this long-simmering civil conflict but "it's not like Iraq and Afghanistan, where in certain places it's so violent that it's basically impossible for humanitarian organizations to operate," he said.

For aid organizations, the challenge becomes how to help without exacerbating regional violence.

"What that means for policymakers is that foreign aid isn't necessarily bad, it's just not as effective at reducing conflict as a lot of people were hoping it would be," Crost said. "Is there a way to give aid without increasing conflict? There's very little research on that. Maybe cooperating or coordinating with security forces – but in certain circumstances, you don't want aid agencies aligning themselves with security forces. You don't want aid agencies being seen as part of a counterinsurgency effort."

The paper was published in the American Economic Review and was co-written by Joseph Felter of Stanford University and Patrick Johnston of the Rand Corp.