Clustering Opposition: A Data Driven Analysis of Gezi Park Protestors

• In Event: 066 - Section on Political Sociology Roundtables (one-hour)
  In Refereed Roundtable: Section on Political Sociology Roundtables (one-hour). Table 01. Social Movements

Abstract

In this article, we analyze the political fragmentation and polarization of Turkish opposition based on survey data collected from the participants of the June 2013 mass protests in Turkey. In particular, we use the data on past political activities of the participants to show the divide in the political concerns of the opposition. In our analysis, the past political activity of an individ- ual is represented by the attendance to past protests that happened within the time frame of the current ruling party. We treat ‘attended’ and ‘not attended’ as categorical variables and use the K- mode clustering algorithm to attain the distinctions in the opposition. The results indicate that the two main opposition parties, CHP and BDP, are fragmented between each other and within themselves. We discuss the resulting clusters in the context of the voter base of the two opposition parties in Turkey.

Authors

• Defne Over, Cornell University

• Ceyhun Eksin, University of Pennsylvania