| Summary: | This thesis consists of four papers, each of which helps to understand certain dynamics surrounding political dynasties. The first paper focuses on the role of ‘dynastic identity’ in influencing the behaviour of legislators from the political class of Bangladesh. In particular, it analyses whether dynastic legislators behave differently in comparison to non-dynastic legislators by examining their parliamentary attendance level and the likelihood of them having a criminal profile. The findings from the analysis suggest that ‘dynastic identity’ may influence a legislator’s behaviour. The second paper investigates if there is a systematic relationship between dynasty-politics and corruption in a cross-country empirical analysis. In doing so, the paper produces multiple dynasty indices that try to capture the variation in dynasty- politics across countries. The key findings from this scrutiny are indicative that countries with greater prevalence of dynasty-politics are associated with higher levels of corruption. In the third paper, I study the role of political assassination in facilitating the rise of political dynasties in Bangladesh. More specifically, I construct a data set of political leaders from Bangladesh who faced at least one assassination attempt to exploit the randomness in the success or failure of assassination attempts to identify assassination’s effect on the probability that a leader will start a political dynasty. The results point out that successful assassination increases the likelihood that a political leader will have a posterior relative in office. Lastly, the fourth paper examines if political assassinations have facilitated the rise of political dynasties across countries. To this end, the paper builds on the data used in Jones and Olken (2009), which has information on leaders with at least one assassination attempt. Thus, by comparing national leaders who barely survived an assassination attempt with those who died, the effects of political assassinations on dynasty formation are studied.
|
|---|
