Suchergebnisse

The literature holds that coalition-building parties prefer the policy distance of coalition partners to be as small as possible. In light of continued importance of religion in electoral politics cross-nationally, the distance argument is worrisome for minorities seeking political access because many minorities are of different religion than the majority representatives forming coalitions. The authors suggest plurality parties' objectives to demonstrate inclusiveness outweigh the concern over policy distance. They test their hypotheses on a sample of all electorally active ethnic minorities in democracies from 1945 to 2004. The authors find support for their hypothesis that ethnic parties representing minorities that diverge in religious family from the majority are more likely to be included in governing coalitions than are ethnic minorities at large. It is interesting, however, that they also find that minority parties representing ethnic groups that differ in denomination from the majority are less likely to be included in governing coalitions.

Is ethnic social diversity relevant to cross-national variation in economic development, or is the inclusion or exclusion of said groups in political decision-making the more salient factor? We argue that deleterious policy effects resulting in diminished economic growth are caused by exclusion of mobilized ethnic groups from the policy process and not just ethnic social diversity per se. Conversely, a positive impact of ethnicity as more groups are included in the policy process with increasing access to cabinet is due, first, to the fact that a population finding its preferences represented in the policy process likely supports implementation of resultant policy. Second, the policy quality likely improves with greater variety in input. Third, a greater number of included ethnic groups in cabinet increases the number of ethnic partisan veto players in the policy process — thereby generating increased policy stability in the long term. We test this idea first on long-run growth in democracy and, second, on annual indicators of growth. We find that increasing ethnic social fragmentation still negatively impacts on the economy. However, cabinet diversity offsets some of these negative effects as it improves growth of GDP per capita.

The article introduces the All Minorities at Risk (AMAR) data, a sample of socially recognized and salient ethnic groups. Fully coded for the forty core Minorities at Risk variables, this AMAR sample provides researchers with data for empirical analysis free from the selection issues known in the study of ethnic politics to date. We describe the distinct selection issues motivating the coding of the data with an emphasis on underexplored selection issues arising with truncation of ethnic group data, especially when moving between levels of data. We then describe our sampling technique and the resulting coded data. Next, we suggest some directions for the future study of ethnicity and conflict using our bias-corrected data. Our preliminary correlations suggest selection bias may have distorted our understanding about both group and country correlates of ethnic violence.

The article introduces the All Minorities at Risk (AMAR) data, a sample of socially recognized and salient ethnic groups. Fully coded for the forty core Minorities at Risk variables, this AMAR sample provides researchers with data for empirical analysis free from the selection issues known in the study of ethnic politics to date. We describe the distinct selection issues motivating the coding of the data with an emphasis on underexplored selection issues arising with truncation of ethnic group data, especially when moving between levels of data. We then describe our sampling technique and the resulting coded data. Next, we suggest some directions for the future study of ethnicity and conflict using our bias-corrected data. Our preliminary correlations suggest selection bias may have distorted our understanding about both group and country correlates of ethnic violence.

Abstract
While machine-coding of data has dramatically advanced in recent years, the literature raises significant concerns about validation of LLM classification showing, for example, that reliability varies greatly by prompt and temperature tuning, across subject areas and tasks – especially in "zero-shot" applications. This paper contributes to the discussion of validation in several different ways. To test the relative performance of supervised and semi-supervised algorithms when coding political data, we compare three models' performances to each other over multiple iterations for each model and to trained expert coding of data. We also examine changes in performance resulting from prompt engineering and pre-processing of source data. To ameliorate concerns regarding LLM's pre-training on test data, we assess performance by updating an existing dataset beyond what is publicly available. Overall, we find that only GPT-4 approaches trained expert coders when coding contexts familiar to human coders and is more consistent coding across contexts. We conclude by discussing some benefits and drawbacks of machine coding moving forward.