Military Expenditures, Income Inequality, Welfare and Political Regimes: A Dynamic Panel Data Analysis

Abstract

The goal of this paper is to investigate the relationship between type of welfare regimes and military expenditures. There is a sizeable empirical literature on the development of the welfare state and on the typology of the welfare regimes. There appear to be, however, no empirical studies that examine welfare regimes with special attention to military spending. This study aims at providing a comprehensive analysis on the topic by considering several different welfare regime typologies. To do so, we use dynamic panel data analysis for 37 countries for the period of 1988–2003 by considering a wide range of control variables such as inequality measures, number of terrorist events, and size of the armed forces. We also replicate the same analyses for the political regimes. Our findings, in line with the literature, show that there is a positive relationship between income inequality and share of military expenditures in the central government budget, and that the number of terrorist events is a significant factor that affects both the level of military expenditure and inequality. Also, the paper reveals a significant negative relationship between social democratic welfare regimes and military expenditures.

Keywords:

• Military spending
• Welfare regimes
• Political regimes
• Income inequality

JEL Codes:

• C33
• D30
• H56
• I30

Acknowledgments

The authors thank Hasan Dudu, Nadir Öcal, Hamid Ali and participants of the 16th Annual International Conference on Economics and Security (June 21–22, 2012, The American University in Cairo, Egypt) and two anonymous referees for their valuable comments.

Notes

¹ In this context, only three studies that are worth mentioning are Moon and Dixon (1985), Dixon and Moon (1986) that analyzes different welfare spending with respect to Blondel’s (1969) political regime categorization (cited in Gough, 2000), and Huber and Stephens (2001) that considers military spending as a percentage of GDP as an explanatory variable for social security benefit expenditure in analysis of the development of the welfare state.

² Initial number of countries was 44 based on our welfare regime categorizations. Nevertheless, as showed in the Table AI in the Appendix due to missing values of some variables the number drops to 37.

³ However, it is worth noting that the results do not change remarkably when the regressions are iterated with the share of military expenditures in GDP or GNP.

⁴ Since the findings do not change significantly we do not report our results to save space.

⁵ For instance, although we acknowledge that Spain, Italy, Portugal, Greece, and Turkey can be categorized as a distinct group of so-called Southern/Latin Rim countries (see Table AI in the Appendix), we prefer to categorize them under general corporatist regime since we believe their differences with these regimes (i.e. higher role of family in provision of welfare) are not relevant in our context. As a matter of fact, our regressions yield very similar results when we take the Southern welfare regime as a separate category.

⁶ The original classification of Hsu (2009) includes more classifications, namely, European Colony, One-party Democracy, Islamic Republic, Military Dictatorship and Dictatorship.

⁷ All regressions are repeated for the case where income inequality is dependent variable. To save space those results are not reported.

⁸ Roodman (2006) develops ‘the xtabond2’ command for use with STATA.

⁹ On the other hand, when the model is controlled for the welfare regimes the coefficient of ‘arms imports to total imports’ becomes positive only in two models (i.e. 1.6 and 2.5). It can be argued that depending upon the variables included in these two regressions, the mechanism of comparative advantage may not work for some countries. In fact, out of a total of 20 models in Tables III and IV, the variable has a positive sign (and significant) only in models 1.6 and 2.5. This result suggests that the relationship between military expenditures as a share of central government budget and arm imports in the context of welfare regimes is complicated, and it requires a more detailed analysis. However, such analysis is out of the scope of this study and might be a topic for future studies.

¹⁰ We acknowledge that it is possible to analyze the relationship between political regimes and military expenditures for more countries, rather than 37 as in our case. However, the main goal of this paper is to analyze the relationship between military expenditures and the welfare regime types defined in the existing empirical literature. Therefore, we consider political regimes just to compare and check our findings for the countries in question. A special focus on political regimes and military expenditures for a wide range of countries might be the interest of future studies.

¹¹ Here we would like to see if there is a significant differences between developed and developing countries in terms of military expenditure. In order to save space, we do not provide the results here; however, one may refer Töngür and Elveren (2012), where our findings show that developed countries have a lower share of military expenditures in budgetary spending compared to developing ones.

¹² We acknowledge that inequality is only one of several other determinants of terrorism, and our results are not a evidence for a direct causality between the variables but rather a supportive evidence to the literature that contends that inequality is one root causes of terrorism (see inter alia Krieger and Meierrieks, 2009 and Derin-Güre and Elveren, 2013 for further discussion).