In a recent article in the APSR, Katerina Linos, Laura Jákli, and Melissa Carlson partnered with a Greek NGO to conduct a cutting edge text message donation experiment. Their findings are relevant to scholars interested in novel ways of gathering data as well as those concerned with the plight of the most vulnerable among us.

Vulnerable minority groups have experienced outsized economic loss during the ongoing COVID-19 pandemic. The most stigmatized communities face added social discrimination and segregation as the global health crisis places them at risk of scapegoating by panicked ethnic majorities and prejudiced governments. For example, in the early months of the pandemic, a number of local governments across Eastern Europe cast the Roma ethnic minority as a distinct health and safety threat responsible for the spread of coronavirus, imposing police checkpoints and other restrictions on Roma neighborhoods. In Slovakia, the national government even used armed forces to quarantine Roma settlements.

As scapegoating practices mount, NGOs may choose to exclude stigmatized groups from their donation campaigns in order to secure adequate funding. This is a more common practice than many might think. For example, HIV prevention programs in Europe used to exclude those at risk due to sexual activity. Poverty relief programs in Southeast Asia have historically avoided fundraising for the “ultra-poor.” Other NGOs instead try to highlight commonalities like ethnicity, culture, or religion between donors and recipients—but this approach, too, can exclude the most socially isolated and stigmatized groups. In other words, NGOs face a trade-off between fundraising for those in greatest need versus raising enough money to maintain their programming. When these organizations are worried about their own financial viability, as many currently are due to protracted economic hardship, this trade-off is particularly acute.

A text message experiment to study discrepancies in giving

How can NGOs maintain a sustainable fundraising model while increasing aid to stigmatized yet high need groups? In an article recently published in the American Political Science Review, we fielded a nationwide text message donation experiment that provides insight on what works and what does not. We partnered with a Greek NGO in soliciting donations, allowing us to unobtrusively study the generosity gap between Greek and ethnic Roma aid recipients and identify strategies to reduce it. Our results are particularly timely for two reasons. First, we fielded the experiment in Greece in 2016, at a time of great economic austerity due to the Greek government debt crisis. Second, we studied fundraising for the most highly stigmatized ethnic group throughout Europe: the Roma. Despite integration efforts, many Roma still reside in settlements that lack electricity and heating; many Roma kids never graduate from high school.

In the experiment, the NGO Prolepsis requested a small (1.5 euro) donation to the Diatrofi Program, which provides free lunches throughout Greece’s poorest schools. We randomly varied the donation request in a three-by-two factorial design. In one condition, we made an in-group appeal to an ethnically Greek child. In another, we made an out-group appeal to a Roma child. In the control condition, we did not specify ethnicity. We also varied whether we included an appeal to the “right to food” to see if a universalistic appeal to rights could help close the generosity gap. We assigned 79,368 individuals spanning 1,051 Greek ZIP codes to one of these six treatments.

Donations fell by 50% when “a Roma Child” was mentioned

We found that the in-group appeal did not increase generosity in donations relative to the control, but that specifying aid recipients as Roma children halved donations. We conclude that the generosity gap we measure is driven primarily by stigma against the Roma out-group rather than by a preference for the majority ethnicity. We also found that an appeal to universal rights neither influenced donation magnitude nor reduced the ethnic discrepancy in generosity. This suggests that human rights language may be less effective at increasing empathy and altruistic behavior than previous research has assumed.

Proximity to Roma settlements reduced donations to the Roma by 70%

Finally, we used geographic data to examine whether proximity to Roma neighborhoods and settlements influenced the effectiveness of in-group and out-group appeals. Whereas on average, donations fell by 50% when the Roma out-group was referenced, near informal Roma communities, the donation gap reached 70%. This suggests that proximity to highly visible out-group communities activates racial threat and further “otherizes” the stigmatized minority.

We supplemented our main analysis with semi-structured, in-depth interviews with principals from 12 schools that benefited from the Diatrofi program. This qualitative assessment allowed us to examine the stigma associated with the Roma and understand how local Greek communities overcome this to fundraise for them. We found that even in the most “integrated” communities, intergroup interaction is very limited due to out-group cultural stigma and spatial segregation. While in-kind community donations to predominantly Roma schools were common, they were based on the resonance of perceived need, rather than on the resonance of fundamental rights.

Our recommendation to NGOs: emphasize needs, not rights in fundraising campaigns

Based on our study, we recommend that NGOs consider adopting fundraising strategies that are broad-based rather than narrowly targeting any specific group of people. The experimental condition that led to the highest rate of donation specified neither an ethnic Greek nor ethnic Roma aid recipient. We also recommend that NGOs underscore the needs of their beneficiaries, rather than focusing on their rights. The standard practice of invoking rights language in NGO aid campaigns may not resonate with a broad audience.
In the current pandemic environment, refocusing on the core message of need may be particularly resonant to potential donors.

Our recommendation to scholars: text messaging allows for unobtrusive experimentation in diverse contexts

We also believe that a wide range of scholars can benefit from extending our unobtrusive, low-cost text message intervention to a variety of contexts. For example, text message experiments can enhance the external validity and generalizability of research on public goods, altruism, and intergroup bias. More broadly, text message appeals can help test a variety of theories concerning social, economic, and political development. An additional benefit of the text message approach is that field experiments are typically limited to one or few study sites, whereas our experimental model sheds light on what holds and what doesn’t across highly varied institutional and geographic contexts.

– Katerina Linos is a Professor of Law at U.C. Berkeley School of Law.

– Laura Jákli is a Junior Fellow at the Harvard Society of Fellows.

– Melissa Carlson is a Postdoctoral Fellow in the Center for International Security and Cooperation (CISAC)’s Middle East Initiative at Stanford University.