Abstract
Economic theory postulates that trade liberalisation influences the location patterns of manufacturing activities across regions. However, the direction of this impact remains theoretically controversial. While few authors suggest that trade openness causes dispersion of economic activities, others argue that it strengthens agglomeration tendency. Against the background of ongoing debates in both mainstream economics and in geography, the paper explores the role of trade policy in shaping the patterns of geographical distribution of manufacturing activities across South Africa’s provinces over the period 1993-2016. Contrary to the previous literature, we employ a new measure of trade openness which is able to account for both South Africa’s trade share of her GDP and her relative size of trade compared to the world trade in a given year. Using this measure of openness, the study draws an inference based on the Eicker-White robust covariance Poisson Pseudo-Maximum Likelihood (PPML) which performs well in the presence of heteroscedasticity and data with zero values. The findings suggest that trade openness matters considerably for explaining the industrial patterns across South Africa’s provinces. In particular, industries facing trade liberalisation are most likely to settle at proximity to the metropolitan cities surrounding Gauteng. The implication is that South African authorities can use trade policy reforms to complement appropriate redistributive spatial development policies. Such can ensure a fair inter-provincial distribution of industries and mitigate inequalities, especially the spatial dimensions.