PREMIA – A Prediction Market with Integrated Algorithms (funded by the Digital Society Initiative)
Prediction markets are virtual stock markets, which use the information contained in market values to make forecasts. By today, these market platforms exclusively rely on the implicit knowledge of groups of humans either in the form of large crowds or expert panels. Hence, this project aims to develop and implement a prediction market with integrated trading algorithms that combines human expertise with artificial intelligence and apply it to a broad range of research questions.
The Effect of Campaign Events on Direct Democratic Choices; Evidence from Prediction Markets (funded by the Swiss National Science Foundation)
This project investigates the effect of campaign events on direct democratic votes on the basis of two major innovations. It is the first project to design and systematically apply a prediction market to direct democratic ballots. First analyses of data on more than 30 Swiss referenda show that the effect of direct democratic campaigns has substantial effects, but that it mostly activates voters to vote as expected rather than leading to opinion swings in the electorate. These analysis also show that the effect of campaigns vary heavily between proposals. It is these different campaign effects across different proposals that I am studying now.
The electoral participation and vote choice of immigrants
The goal of this project is to describe and explain the political behavior of immigrants, in particular their electoral behavior. So far, I have analyzed variance in party vote choice in Switzerland and Germany, vote choice on a referendum on immigration policy in Switzerland (with Javier Polavieja), and reasons for the (lack of) success of immigrant parties in Western Europe. I have found that political socialization with regards to politics in the sending state (e.g. experience with Communism in Eastern Europe) and being a member of a so-called “outgroup” go a long way in explaining the specifics of vote choice among first and second generation immigrants. In contrast, preferences on immigration policy seem to be surprisingly similar to those of the natives and explain less variance in vote choice than one might think. Finally, with regards to immigrant parties, I find that relative success of such parties is only likely if there is substantially large immigrant “outgroup” and a highly proportional electoral system.
Ethnic mobilization in Europe
The goal of this project is to describe and explain variance in ethnic mobililzation mostly in Western Europe. The project mostly focuses on strucutral determinants of mobilization by ethnic parties. Together with Michal Kotnarowski I have developed indicators that allow to compare levels of mobilization across time and space. In substantial terms I have identified deindustrialization as a major cause for ethnic mobilization in Western Europe. Also, with Raphael Leonisio I have shown — against other claims — the dominance of the nationalist issue dimension relative to secondary issues for voting behavior in such a divided polity as the Basque Country. Finally, with Nenad Stojanovic I have demonstrated that majoritarian electoral formula in multi-member districts can soften ethnic division in voting.
Ethnic voting and identities in Latin America
This projects seeks to understand changes in ethnic demography and (latent) ethnic conflict in Latin America. Some first analyses have already been conducted: First, in a paper on ethnic boundary crossing in Ecuador I have shown that ethnic identities are crossed by a substantial share of the population, but that these identity shifts almost exclusively involve mixed categories and in particular the mestizo category. Also, ethnic self-categorization is strongly anchored in cultural attributes (e.g. language) rather and “racial whithening” due to social mobility is rather exceptional. Second, Aline Hirseland and I show that in Bolivia (latent) ethnic conflict—as measured with ethnic voting—has increased with the Movement for Socialism (MAS) of Evo Morales. This rise shows that the potential of ethnic parties in ethnically dividing the electorate, a finding that I am testing comparatively in joint work with Manuel Vogt. Finally, with Siri Voelker I show that ethnic voting is not the main reason for why Afro-Brazilians are heavily underrepresented in parliament, but that this lack of descriptive representation is mostly due to ethnic inequality in the past and present.
The Political Sociology of Cosmopolitanism and Communitarianism (concluded)
As a senior researcher at the Berlin Social Science Center (WZB), I was involved in this project on the conflict over borders in Germany, Mexico, Poland, Turkey, and the US. More specifically, I took over from Céline Teney and Sarah Carol the responsibility over an elite survey on attitudes towards globalization and investigated with Pieter De Wilde and Joshua Helmer whether the globalization cleavage present in Germany and the US could also be found in other countries around the world. Insights from my bits of the project are 1) that elites are clearly more cosmopolitan than masses around the world and that this can be better explained with cultural than economic or political explanations, and 2) that the cleavage between Cosmopolitans and Communitarians is still more present in the highly developed West, most likely because these countries are globally more integrated.