Pillar 1 uses the two historical experiments generated by forced mass movements of ethnic groups.
I.1. After WWII, Polish borders were redrawn. Poles were forced to move from the Eastern Borderlands to Western Territories. We study how self-identification and preferences adjust as a result of forced displacement. We provide the first convincing test of the “uprootedness” hypotheses that displacement leads to a shift in preferences away from material possessions toward investment in mobile inalienable assets. The historical setting allows bypassing typical confounding factors, as the forced migrants and locals are all Poles. Using digitized historical censuses combined with surveys that we designed and conducted, we show that Poles with a family history of forced migration are more educated today. Forced migration led to a persistent shift in preferences from material goods toward investment in education. We reject alternative explanations that Kresy migrants did not have differential access to resources and that local conditions, fertility, or selection of migrants and non-migrants could not confound the results.
I.2. To study cultural diffusion between different groups, we focus on another historical episode: Stalin deported over 2 million people from Western parts of the USSR to Siberia and Central Asia during WWII. Entire ethnic groups were displaced. As a result, native local population at destinations was exposed to deportees with drastically different cultural norms. We combine newly collected archival data on ethnic deportations, digitized historical censuses, and contemporary survey data to show that one cultural trait—gender norms—diffused from deportees to local population, resulted in changes in attitudes and behavior. The results cannot be explained by pre-deportation differences or selective in- and out-migration. Both norms of equality and discrimination were transmitted horizontally. Yet, the norms of gender equality diffused more, due to their economic benefits and to higher political costs of adopting norms that go against official ideology.
Pillar 2 explores the economic determinants of ethnic animosity, the effect of market interactions between different groups and specialization of ethnic minorities in middleman occupations.
II.1. We examine anti-Jewish violence (pogroms) in the 19th and 20th century the Russian Empire. Using data that we collected on the timing and locations combined with archival data, we document that anti-Jewish pogroms broke out solely when economic shocks coincided with political turmoil. Pogroms started in places where Jews dominated moneylending and grain trading. We tested it and conclude that the scapegoating hypothesis, according to which Jews were blamed for all misfortunes, is inconsistent with evidence. Instead, we show that Jewish middlemen served as providers of insurance against economic shocks to peasants and urban grain buyers in a relationship based on repeated interactions. During political turmoil, debtors could not commit to paying in the future, and middlemen were forced to demand immediate (re)payments. This fuelled ethnic violence. The break in the relationship between the majority and Jewish middlemen was the igniting factor.
Pillar 3 focuses on political determinants of inter-group attitudes. We study how TV coverage of conflict affects viewers’ attitudes toward it and, as a result, strategic choices of actors in the conflict to minimize negative effect of their actions. In the context of the on-going Israeli-Palestinian conflict, strategic media considerations are important for military actions of Israeli Defense Forces and narratives used in conflict TV coverage are important to determine the viewers attitudes toward the conflict actors. Then, we look directly at how populist politicians can manipulate public opinion towards immigrants using misleading false statements and whether fact checking can undo these effects using context of the 2017 French presidential elections. We conduct a randomized online experiment where we expose randomly subgroups of a large sample of voting-age French to quotes from Marine Le Pen (MLP) containing misleading information on immigration or to facts from official sources. We find that political narratives that use misleading statements are highly persuasive: voters exposed to MLP rhetoric moved their policy conclusions toward more stringent anti-immigration policies and voting intentions toward MLP. Fact checking did nothing to undo these effects despite improving factual knowledge of voters. Without fact checking, exposure to MLP’s narratives moved opinions of respondents about facts toward more extreme views away from the truth. Being exposed only to official facts also increased political support for MLP while moving factual knowledge toward the truth.
