The economics of military spending, conflict and growth

• Main Author: Tian, Nan
• Other Authors: Dunne, John Paul
• Format: Doctoral Thesis
• Language: English
• Published: University of Cape Town 2016
• Subjects: Economics; Military spending; Economic growth
• Online Access: http://hdl.handle.net/11427/16720

Includes bibliographical references === This dissertation is a collection of studies on the economics of peace and security. Chapter one introduces the roles military spending and conic play in affecting economic growth, while also considering the causes of civil conflict. Chapter two investigates the relationship between military expenditure and economic growth, considering group heterogeneity and non-linearity. Using an exogenous growth model and dynamic panel approach, the results suggest military burden to have a negative effect on growth. Breaking the overall panel down into various sub-samples shows estimates that are remark-ably consistent with the full panel. These results provide strong support for the argument that military spending has an adverse effect on growth. There are, however, some intriguing results suggesting that for certain types of countries military burden has no negative growth effect. Chapter three deals with the transnational spatial spillover effects of conflict on neighbouring countries. It moves beyond using geographical dis-tance as a spillover measurement and allows for economic and political distances. The initial empirical results suggest that conflict has a strong negative spillover effect on directly contiguous countries growth, but no significant impacts were observed for non-contiguous countries. When economic and political factors are considered, this result remains, but the spillover effect is smaller. While the impact of conflict remains devastating, it is important to take other factors into account as studies using only geographical distance may be overestimating the impact on neighbours. The fourth chapter examines the determinants of civil war, using a zero-inflated modelling approach to deal with excess zeroes in the dependent variable. Traditional probit and logit models have limited capacity in dealing with this issue and can create misleading results, which is illustrated through replicating published work. A general greed-grievance model is then estimated giving further support to using zero-inflated models. While the standard probit models tend to emphasise opportunity variables, consistent in other studies, the zero inflated model gives supports both opportunity and grievance variables. In particular, ethnicity, democracy and inequality are found to play a significant role in civil war prevalence. Finally, chapter five summarises the findings of the dissertation, providing some policy recommendations, concluding remarks and discusses future research opportunities.