For this piece, we also spoke with Beza Tesfaye of Mercy Corps (an institutional member of the EGAP network), the organization that implemented the vocational skills-training program that was evaluated in the study. Below, you will find questions and answers from the research team followed by those from Beza on behalf of Mercy Corps.
This study looks at intergroup contact between internally displaced persons (IDPs) and locals in Kandahar, Afghanistan. As part of the project, you evaluated a vocational skills-training program that enrolled locals and migrants in an active conflict setting, where IDPs face daily prejudice and discrimination. What were the primary findings from this project?
Jason Lyall and Yang-Yang Zhou: To give some background on the study, IDPs in Kandahar faced daily prejudice and discrimination. They experienced suspicion by locals for being Taliban recruits, housing discrimination, mistreatment by police, and violence. So although the active conflict was not between locals and IDPs, migrant status was still a very relevant cleavage for social conflict.
INVEST in Kandahar was a vocational skills training program that lasted 3 or 6 months–depending on the course–and was run by our partner Mercy Corps. The program was designed to improve employability for vulnerable youth, not to reduce locals’ prejudice towards migrants. So there was nothing specific in the class curriculum or messaging that had to do with IDPs. But because the program enrolled both locals and IDPs in near equal numbers in a classroom setting (where they worked collaboratively on hands-on projects and, as reported by instructors, did not self-segregate), we realized that INVEST provided a naturalistic way of testing the contact hypothesis (Allport, 1954).
The title of our paper gives away the primary findings: despite a well-designed and well-implemented program that provided 3 to 6 months of daily contact (the most intensive dose of intergroup contact compared to similar studies), we found no difference in attitudes towards IDPs between treated and control locals at endline or eight months later. In terms of secondary findings, when we compared subgroups – for example, younger or older participants, women or men, classes with majority, balanced (50/50), or minority number of IDPs – there were still no differences between treatment and control.
From these null findings, we conclude that in this active wartime context, where biases may feel necessary to hold onto for survival, intergroup contact, even if prolonged and on equitable grounds, is likely insufficient to improve attitudes toward migrants. While we felt disheartened by these results, we also recognized that attitudes didn’t worsen, which was a possibility. This program, which is a very common type of economic intervention, could have exacerbated local-migrant conflict by stoking fears of economic competition, and it’s important and encouraging to learn that it did not.
How do these findings complement the existing literature on intergroup contact?
JL and YZ: Much of the intergroup contact literature from social psychology tells us that it works. But, within this literature, there have been few experiments in the real-world, particularly non-Western contexts, and even fewer in active conflict settings. Recent (natural) experimental studies in Nigeria (Scacco and Warren, 2018), Iraq (Mousa, 2020), and Israel (Ditlmann and Samii, 2016; Weiss, 2021) are notable exceptions. And these were all based around ethno-religious categories (e.g., Muslims and Christians) as the basis for intergroup conflict.
Compared to these studies, ours centered migrants as a relevant outgroup category, and our intervention was much longer. Still, our null findings are generally in line with what these other studies, like the Nigeria and Iraq ones, and a recent meta-analysis (Paluck, Porat, Clark, and Green, 2020) have found; that it’s very difficult for these types of contact programs to change minds. We also learned that migrant status, even though it may not be as deeply entrenched and historic as other social identity categories like ethnicity or religion, may be just as powerful.
What are the implications of the project’s findings for organizations working to improve intergroup contact between IDPs and locals in wartime settings such as Afghanistan and, if applicable, around the world?
JL and YZ: In terms of policy, billions of dollars are spent on these types of economic training programs in conflict settings. And they tend to naturally bring together different groups of people, like with INVEST. While they can offer meaningful opportunities for intergroup contact, that contact on its own is not enough to reduce prejudice. Without explicitly tackling prejudice—which we acknowledge, was not the focus of Mercy Corps’ program—our intervention was likely just too subtle.
Beyond these programmatic interventions, political actors that have a say in hosting policy (host governments and the UNHCR, for example) need to be mindful about the structural incentives and inequalities that hosting policies and institutions can generate. With respect to migration, policies like whether migrants are encouraged to settle and work among locals and have access to local schools and healthcare, are much more in the spirit of intergroup contact theory than what shorter-term aid programs can simulate.
What was the motivation behind collaborating with the research team on this project? Were you already working together, or did this partnership come together in the context of understanding the impact of a specific program that Mercy Corps was working on in Afghanistan?
Beza Tesfaye: The partnership with Dr. Zhou and Dr. Lyall, as well as others involved in this project, originated from our interest to improve our ability to measure the impact of our employment-focused programming on various peace and security outcomes in Afghanistan. In previous years, we had conducted quasi-experimental research around the INVEST program, which led us to believe the program was mature and ready for a randomized controlled trial (RCT). Working with the researchers, we designed and received financial support from a donor to conduct a robust RCT to measure the program’s impact on support for armed violence, as well as inter-group attitudes between IDPs and members of the local community. The importance of testing and measuring program impact, for us, stemmed from the desire to focus on doing more of what was working and change what was not working.
What were the biggest obstacles in getting the project off the ground?
BT: As you can imagine, working in Afghanistan is full of challenges. The security and operational environment is complex. Mercy Corps began working in Afghanistan in 1986, so over time we developed the relationships, systems, and capacity to work effectively there. Nonetheless, starting up and implementing programs remained a complicated task, and the addition of a research arm required additional support and close collaboration across operational and research teams. Mercy Corps and the research team worked very closely towards achieving our shared goal to learn from the program, but in some instances we faced delays, modified our plans, and carefully considered risks to program participants, staff, and enumerators at every step of the process.
How has your organization’s participation in this study affected programming decisions made since the results became available to you?
BT: At Mercy Corps, and elsewhere, research can play an important role in how we design programs, as development actors increasingly understand the value of evidence-based approaches. Because the INVEST program was not explicitly focused on this objective, the findings have not fed into its approach. However, the lessons from the research study—particularly the need to be more intentional and intensive in addressing prejudices—will be valuable in the design of future programs focused on reducing prejudices towards migrants and displaced populations, and programs that are predicated on contact-theory, more generally.