10.7910/DVN/PT2IB9Cebotari, VictorMaastricht UniversityVink, Maarten PeterMaarten PeterVink0000-0001-7143-4859Maastricht University / European University InstituteHarvard Dataverse2015Arts and HumanitiesLawSocial SciencesCausal complexity; ethnic protest; Europe; fuzzy-set analysis; QCA; minorities at riskVink, Maarten PeterMaarten PeterVinkMaastricht University2015-09-252015-12-05108410904953819168text/plaintext/tab-separated-valuestype/x-r-syntaxapplication/pdf2.0CC0 1.0This article analyzes the conditions under which ethnic minorities intensify or moderate their protest behavior. While this question has been previously asked, we find that prior studies tend to generalize explanations across a varied set of ethnic groups and assume that causal conditions can independently explain whether groups are more or less mobilized. By contrast, this study employs a technique – fuzzy-set analysis – that is geared toward matching comparable groups to specific analytical configurations of causal factors to explain the choice for strong and weak protest. The analysis draws on a sample of 29 ethnic minorities in Europe and uses three group and two contextual conditions inspired by Gurr’s ethnopolitical conflict model to understand why some ethnic minorities protest more frequently than others. We find that two group-related factors have the strongest claim to being generalizable: while territorial concentration is a necessary condition for strong protest, national pride is a necessary condition for weak protest. The contextual factors of level of democracy and ethnic fractionalization, which are often emphasized in the literature, and the perceived political discrimination of a group, are neither necessary nor individually sufficient conditions for either strong or weak protest. Hence, they help understanding some cases, but not all, and only in combination with other conditions. Such causal complexity, inherent in the phenomenon of ethnic protest, underscores the need for a case-sensitive, yet comparative, approach.For the original analysis in the paper published in 2013 we used the software fs/QCA 2.0. However, we have also replicated these analyses with the "QCA" package for R (with minor differences due to slightly different calibration algorithms in fs/QCA and QCA. The R script and a txt file are included here, in addition to the csv file we originally used for the IJCS publication.