Read on Twitter
Now available for FirstView at @APSRjournal “Does Aid Reduce Anti-Refugee Violence? Evidence from Syrian Refugees in Lebanon.” With Christian Lehmann.
APSR version here: https://www.cambridge.org/core/journals/american-political-science-review/article/does-aid-reduce-antirefugee-violence-evidence-from-syrian-refugees-in-lebanon/18C5224DD575F6E9E484CB5BC42CD49A
Pre-print here: https://www.dropbox.com/s/xmzygtlm3p4yh2j/lehmann_masterson_anti-refugee_violence_may2020.pdf?raw=1
We need to know more about the links between refugee crises, humanitarian aid, and conflict in hosting countries. Much humanitarian assistance goes to refugees, and may generate resentment from natives. We ask: *Does humanitarian aid to refugees drive anti-refugee hostility?*
How can policy manage a potential trade-off between assisting the most vulnerable and the risk that they are targeted for violence? If aid worsens anti-refugee violence, humanitarian aid could be exacerbating the plight of exactly those it seeks to help.
We know from existing research on inter-group conflict and anti-immigrant sentiment that politicized perceptions of resource scarcity/distributional concerns drive hostility. It’s clear to see how aid to refugees could cause resentment among locals
Right-wing politicians frame aid to refugees as a source of tension, which they use as a excuse to cut aid. Other times (or at the same time!) leaders use aid tensions as a bargaining chip w international agencies to get more aid distributed to locals, i.e., rents to constituents
The humanitarian community also often works under the assumption that resentment about refugees receiving aid drives anti-refugee violence. So…Is humanitarian aid a source of resentment? Is there a trade-off between aid and violence?
We build on great work on hostility in refugee crises, which examined effects of exposure to refugees & proximity to camps on hostility – @yangyang_zhou, @tiffany_s_chu, @theBraith, @EliasDinas. But the work hasn’t identified impacts of aid — a key policy response — on hostility
Anti-Syrian hostility and violence are widespread in Lebanon, often attributed to sense that aid is “unfairly” distributed, and fears of resource competition; negative economic impacts; increased demand for housing, schools, and health centers.
eg
1. http://m.naharnet.com/stories/en/149765-in-lebanon-syria-refugees-face-growing-discrimination
2.
eg
1. http://m.naharnet.com/stories/en/149765-in-lebanon-syria-refugees-face-growing-discrimination
2.
As in many refugee crises, humanitarian orgs delivered significant amounts of aid and development assistance to host community — in part to mitigate tensions https://www.thenewhumanitarian.org/analysis/2013/01/28/avoid-tensions-refugees-lebanese-hosts-need-support
Design: We use an RD to study a UN cash transfer program for Syrian refugees in 2013–14. The UN delivered $575 cash ($US PPP 1,000) for 5 months. This was a large amount of money, about two thirds of control-group household monthly income ($910)
This map shows the treated (blue) and control (black) villages across Lebanon. We surveyed 1,358 households in 89 towns, villages, and cities across Lebanon
Eligibility was based on altitude and poverty proxies. *Why altitude?* The goal was to keep people warm and dry during the cold winter months, especially those at high altitudes in the mountains.
We leverage an RD around the program’s 500-meter altitude eligibility cut-off, and compare aid recipients just above 500 meters altitude to non-recipients just below. People were not able to move to sort into the program.
Results: In our sample Syrians in Lebanon face widespread hostility, we find that 5.2% of Syrians had been verbally assaulted by Lebanese in the past 6 months, and 1.3% had been physically assaulted in the same time period.
These are *very high* rates of assault — much higher than crime rates in most countries, also much higher than assault rates on refugees in other places where anti-refugee crime has been studied (e.g., Germany).
RD results: What do we find about the effect of cash aid on recipients likelihood of having suffered verbal and physical assault by hosts? RD results provide evidence that ...
**aid did not increase violence, and may have reduced violence**
**aid did not increase violence, and may have reduced violence**
The RD gives us estimates of the effect of aid on assault at the 500-meter altitude cutoff. We run regressions across a range of bandwidths, restricting the sample to observations closer to the cutoff.
First we look at the impact of aid on *verbal assault*. Point estimates are consistently negative and we can rule out meaningful positive effects. The black lines shows the point estimates using different bandwidths. Light (dark) gray bands show 95% (90)% confidence intervals.
And the estimated effect of aid on physical assault looks similar:
Across the range of bandwidths we can rule out large increases in violence due to aid, suggesting that aid either had a very small (null) effect or a large negative effect on violence. See paper and app. for more modeling choices and robustness checks
Why doesn’t aid increase violence? We find evidence for a number of possible mechanisms, although future work should provide more rigorous tests. We see that aid may not increase hostility because recipients can:
(1) Aid leads to higher demand for local goods and services, benefiting locals; (2) recipients can directly benefit locals with gifts and helping; and (3) when people have more resources, they decrease contact w potential aggressors, and increase positive social contact w locals
NB: despite widespread racism and troublingly high rates of violence, there are many Lebanese of conscience who reject racism and xenophobia, and fight for the rights of refugees living in their country.
Lebanese protest in support of Syrian refugees: https://www.thenewhumanitarian.org/feature/2016/07/28/syrians-make-easy-scapegoats-lebanon
Lebanese protest in support of Syrian refugees: https://www.thenewhumanitarian.org/feature/2016/07/28/syrians-make-easy-scapegoats-lebanon
During Lebanon’s revolution starting in October 2019, a common chant was “Bassil out out, refugees in in,” a chant of solidarity with refugees and in opposition to Lebanese politician Gebran Bassil, a figurehead of anti-refuge/anti-Syrian refugee narratives in the country
Further evidence is needed to inform humanitarian policy: We need more on the relationship between hosts’ perceptions of exclusion from aid, services, and employment and conflict. And how different aid modalities (e.g., cash vs. CDD) impact hosts’ attitudes and conflict.
We thank the International Rescue Committee (IRC, @RESCUEorg) their for collaboration on this project in spring 2014!
