Replication Data for: An Events-Based Approach to Understanding Democratic Erosion

Abstract for paper for which data is used: In this paper we introduce and demonstrate the utility of a new event dataset on democratic erosion around the world. Through case studies of Turkey and Brazil, we show that our Democratic Erosion Event Dataset (DEED) can help resolve debates about the extent to which democracy is backsliding based on prominent cross-national indicators, focusing in particular on the V-Dem and Little and Meng (L&M) indices. V-Dem suggests that democracies are deteriorating worldwide; L&M argue that this may be an artifact of subjectivity and coder bias, and that more “objective” indicators reveal little to no global democratic backsliding in recent years. Using DEED, we show that, at least in these cases, objective indices may underestimate the extent of democratic erosion while subjective indices may overestimate it. Our analyses illustrate the ways that DEED can complement existing indices by illuminating the nature and dynamics of democratic erosion as it actually occurs on the ground.