The determinants of unemployment: a case of South Africa
Abstract
Unemployment is a serious challenge throughout the world and a pervasive problem for developing nations. Besides other countries, South Africa has not escaped from the scourge of high unemployment. The unemployment rate in South Africa is complicated by extremely serious socio-political issues confronted by the South African government. As a result, this research investigates the determining factors which cause unemployment in South Africa and proposes a policy recommendation to alleviate unemployment in South Africa. The determinants of unemployment are analysed using Johansen cointegration analysis technique from an econometric perspective with annual time series data from 1986 to 2016. This study initially provides an overview of the South African unemployment and factors that cause high unemployment. Variables are chosen from empirical literature. The study identifies the potential factors such as population growth, government expenditure on education, foreign direct investment, and gross domestic product as the determinants of unemployment in South Africa. The data of this study is obtained from the South African Reserve Bank, International Monetary Fund, and the World Bank. The Johansen cointegration test established that there is a long-run relationship between unemployment and chosen variables. In addition, the outcome of vector error correction model revealed that the speed of adjustment coefficient is 112.52 percent showing the variation in unemployment rate from its equilibrium level are corrected within a year. The value of adjustment is a relatively higher compared to those from previous studies on South Africa and high-speed of adjustment to long-run equilibrium. This implies that short-run shocks or disturbances in the unemployment rate would quickly move the economy towards the long-run equilibrium. Diagnostic and stability tests results show that the residuals behaved well. Impulse response outcomes were steady with the long-run dynamic model. The outcomes of the variance decomposition showed that most of the forecast error variance in unemployment rate is explained by government expenditure on education, whereas limited proportion of variation was explained by population growth and foreign direct investment in unemployment rate. The Granger causality test results shows that variables such government expenditure on education, and foreign direct investment provide unidirectional relationship and gross domestic product reinforce the inverse relationship suggested by economic theory. Therefore, in order to reduce and possibly eradicate the unemployment level, means to draw up and implement policies that will create an enabling environment for economic growth those are required. The results that have arisen from this study confirm the theoretical predictions and are also supported by previous researchers. Policy recommendations were made by utilizing these outcomes. One of the recommendation is that government should give education first priority.