There is considerable evidence that authoritarian regimes have tended to spend more on the military than democracies. However, the direction of causality of this relationship has not been seriously investigated. The literature tends to assume that causality runs from regime type to military expenditure, but one might also expect military expenditure to influence regime type: history yields numerous examples of countries whose democratisation process was reversed by a powerful military unwilling to give up its privileged position in society. Successful democratisation, amongst other things, requires the reform of civil-military relations. In this paper, I build on the empirical literature on democratic transitions and examine whether lowering military expenditure in a democratising country increases the chances of successful democracy consolidation. I use a number of techniques, including two-way fixed effects model and a panel VAR to examine the linkages and pattern of Granger causality between measures of regime type, GDP and military expenditures.
