This paper examines the effect of the presence of a military ruler on military expenditure using a panel of sub-Saharan Africa countries. The paper also explores whether the relationship reflects a capture effect, is an outcome of the confrontational climate of the cold war, or is an effort by military rulers for self-preservation. The Pooled OLS and fixed effects OLS estimations show that the presence of a military ruler has a statistically significant negative effect on military spending as a percentage of GDP. The coefficients are also not significantly different before or after the end of the cold war era. This implies that the negative relationship is driven by an effort by military rulers to preempt the ability of their peers to overthrow them from power. We also attempt to deal with potential endogeneity, and consider the possibility of persistence in military spending. The paper uses the Arellano and Bond (1991) estimation technique that shows a negative but insignificant effect of the presence of a military ruler on military expenditure, while military spending shows a high degree of persistence.

中文翻译：

---

非洲军政府和国防开支：俘获效应还是自我保护？

本文使用一组撒哈拉以南非洲国家来检验军事统治者的存在对军事支出的影响。论文还探讨了这种关系是否反映了俘获效应，是冷战对抗气氛的结果，还是军事统治者为自我保护所做的努力。汇总 OLS 和固定效应 OLS 估计表明，军事统治者的存在对军费支出占 GDP 的百分比具有统计上显着的负面影响。冷战结束前后的系数也没有显着差异。这意味着消极关系是由军事统治者努力抢占同僚推翻他们权力的能力所驱动的。我们还尝试处理潜在的内生性，并考虑持续军费开支的可能性。该论文使用了 Arellano 和 Bond (1991) 估计技术，该技术显示了军事统治者的存在对军事开支的负面但微不足道的影响，而军事开支则表现出高度的持续性。