Domestic terrorism and government duration -An event history analysis

Abstract

There are now several studies investigating empirical relationships between that of terrorism, electoral preferences and electoral accountability. However, very little research has been done on the connections between terrorism and government duration. Most of the latter studies also seems to have too much of a narrow focus on that of international, not domestic terrorism. In this thesis the main focus is on that of domestic terrorism and government duration. The research question is therefore as following: Does domestic terrorism affect government duration? The question is assessed by quantitative hypothesis tests based upon new empirical data on domestic terrorism, government duration and control variables. Observations are governments of Western European parliamentary democracies for the time period 1964-2005. The data used to test the hypotheses are handled by event history analysis. More specifically the event history models used are the semi-parametric Cox and Cox shared frailty models. Empirical findings indicate -in line with the previous research that mostly focuses upon international terrorism- that domestic terrorism decreases government duration. These findings are also robust across different models, sample sizes, and domestic terrorism data. The second hypothesis, predicting increased government duration following domestic terrorism, receives no support. Nevertheless, one finding of the thesis runs contrary to that of an important previous study. While the previous study found evidence for the hypothesis that terrorism has a stronger negative effect on the duration of left-wing governments than that of right-wing governments, empirical tests in this thesis indicate the opposite: domestic terrorism has a stronger negative effect on the duration of right-wing as compared to those governments that are not right-wing. However, this finding is not as robust as those of hypothesis one. Findings contribute to increased confidence in: (1) the decreased duration hypothesis (2) that domestic terrorism, not only international, decreases government duration and (3) that government ideology should be considered an important mediator variable. Despite of findings researchers still need to be cautious: a few studies are not sufficient for establishing an approximate truth. Additional research is therefore needed to improve confidence further