Military Expenditure and Unemployment in South Africa: Evidence from Linear and Nonlinear ARDL with and without Structural Break

Abstract

This study examined the symmetric and asymmetric effects of military expenditure on the unemployment rate in South Africa using quarterly data from 1994Q1 and 2019Q4. The study adopted the autoregressive distributed lag (ARDL) and nonlinear autoregressive distributed lag (NARDL) frameworks and accounted for structural breaks in the military expenditure-unemployment nexus. The ARDL results revealed that military spending reduces unemployment in the long run for models with and without structural break but aggravates it in the short-run without a structural break. For the nonlinear ARDL results, irrespective of whether we account for a structural break or not, a positive change in military spending increases unemployment while a negative change in military spending reduces unemployment, especially in the long run. This implies that an increase in military spending is detrimental to unemployment reduction, whereas a reduction in military spending would reduce unemployment. Thus, the policy implication of the study is that the government need to prioritise other forms of spending over military spending in its drive to abate unemployment.