Empirical evidence for the compensation hypothesis holds that trade openness independent of income risk has no significant effects on government size. Hence, using time series data for the period 1965-2013 from Egypt, Ghana, Kenya, Nigeria and South Africa and applying the bounds test method of cointegration, this paper investigates the nature of the relationship between openness, income risk, terms of trade volatility and income per capita on government size in each of the 5 African countries. Our empirical findings show a long-run relationship particularly in Egypt and Ghana and a limited evidence for other countries. Similarly, after applying both the Autoregressive Distributed Lags (ARDL) and Fully Modified OLS (FM-OLS) on original data and Moving Average (MV) converted data, our findings vary across the countries due to country-specific factors. Hence, the evidence we find does not support the compensation hypothesis. However, we observe significant positive effects of income per capita on government size for each country and some evidence of the mitigating and cushioning effects of governments when exposed to income risks and volatility due to trade openness. We therefore suggest some appropriate policy recommendations for the selected countries.
ARDL-bounds testing FM-OLS trade openness income risk government size Africa.