New research supports the controversial conclusion that foreign aid can actually harm recipient countries.

A new working paper by researchers including Yale assistant economics professor Nancy Qian has found evidence that American food aid to developing countries may hurt those countries more than it helps. Qian described the unintended consequences as potentially devastating, by spurring conflict between ethnic groups and exacerbating pre-existing strife. The paper also found that the amount of U.S. grain sent abroad depended more on domestic levels of grain production than on the actual need by the recipient countries.

“Before this paper, journalists who spent time in these donation countries had told anecdotes about how bad things can get [as a result of food aid], but we didn’t know whether these stories represented the actual picture,” Qian said. “What we have done with the data is quantified the average effect and proved that these aren’t random stories and that these unintended consequences do exist.”

Qian said rival factions compete to steal food from aid convoys, which they then barter for weapons or use to feed their soldiers. Once armed groups have monopolized aid resources, the rest of the population may be forced to join the conflict simply to get enough food to survive.

An analysis of 35 years of data about food aid to 134 developing countries strongly correlated consistent American food aid with an increased likelihood and duration of civil war. The paper cited examples of anecdotal evidence of food aid causing internal conflict from Afghanistan, Western Sahara, the Democratic Republic of Congo, Ethiopia, Eritrea, Israel, Iraq, Liberia, Rwanda, Sierra Leone, Somalia, Sri Lanka, Sudan and the former Yugoslavia.

“Food aid simply becomes another resource to fight over,” Qian said.

Paul Macek, a food aid expert from the humanitarian organization World Vision said he thought the relationship between food aid and civil wars remains unclear, but agreed that food-based aid programs are flawed because they create a dependency on external goodwill. The root of this dependency, he said, was the inabiity of food aid to generate long-term agricultural development.

Macek added that he hadn’t read the paper, but thought it might understate the humanitarian benefits of food aid. He said that a long-term solution to food insecurity in Africa would require improving farming techniques to enable poor countries to achieve food self-sufficiency, but that food aid would continue to be important in emergency situations.

The paper also found that volumes of American food aid from year to year were affected more by domestic supply levels than global demand for aid. The U.S. Department of Agriculture, in conjunction with the U.S. Agency of International Development, purchases surplus grain to stabilize prices in the domestic market. This extra stock is then donated to developing countries, the paper said.

Qian said that one important remedy to the negative effects of food aid would be to start basing food aid on country-specific consumption levels, not domestic supply.

“I don’t see any reason for why it has to be this way. Hopefully, people will realize this is a serious problem and that current policy needs some meaningful reform,” Qian said.

Nathan Nunn, an economics professor at Harvard, co-wrote the paper with Qian.