Democracies under Threat: Explaining Democratic Backsliding in Central Eastern Europe

This thesis aims to explain democratic backsliding in the European Union member states of Central Eastern Europe (CEE). Despite a successful transition to a democratic regime in the post-communist countries, multiple democracy indices have shown a decline of democracy in the region. The EU aims to reverse this process by measures such as sanctions, yet so far without success. To tackle the issue of backsliding, it is necessary to understand why the process has been happening. Studies on explaining backsliding, however, are dispersed and disagree with each other. Therefore, this thesis has combined the different theories explaining backsliding in the previous literature into two different models. This thesis will test which model is most likely to explain democratic backsliding. Therefore, the following research question is formulated: !Does the topdown model or bottom-up model better explains democratic backsliding in EU Central Eastern European countries?”. To investigate this, a congruence analysis is performed. The two countries with the highest levels of democratic backsliding according to the V-dem index are taken as cases: Hungary and Poland. One model is tested by examining the populist rhetoric of political leaders and the proportionality of the electoral system. The other model is tested by analysing citizens’ satisfaction with democracy and economic development. According to the analysis, the top-down model is most likely to explain backsliding in CEE. This means that the reasons behind backsliding are mainly found in the organisation of the state. Therefore, the EU is advised to revise measures and make sure sanctions specifically target the leader, or change the electoral system in a country. This research is limited in both its external and internal validity. Although a correlation is found between the propositions and backsliding, information is lacking to establish a causal relationship. Future research should test both models more extensively and with more cases.