The Spatial Dimensions of Economic Activity in South Africa: The Role of Regional Policy, Crime and the Business Environment

Abstract The spatial distribution of economic activity is unevenly distributed across regions within countries. Regional development policies and the incidence of crime rates, an important aspect of the quality of the local business environment, may both impact spatial disparities in economic activity. This thesis examines the relationship between regional development policy, the quality of local business environment, that is, security or lack thereof, captured by crime incidence, and regional economic outcomes. The study argues that regional development policy incentives such as labour and tax regulations influence the local cost of doing business and hence are an important part of local business environment. In addition, the incidence of violent crime rates such as robbery and murder are an important dimension of the local business environment. Regional development policies and variations in the quality of the local business environment may drive regional inequality within a country through its effects on factors related to economic activity such as firm entry and labour migration. South Africa provides a good opportunity to study the impact of regional policies and local business environment, given the high levels of spatial inequality due to historical factors and the country had the first experience of regional development polices in Sub-Saharan Africa. In addition, the country is characterised by high rates of crime incidence. The first analytical chapter of the thesis examines the relationship between regional development policy incentives and regional disparity in economic activity. Chapters two and three, then examines the impact of crime incidence on various measures of economic activity, captured by firm entry and migration of labour, particularly skilled labour. These three chapters are linked by the focus on the determinants of the spatial distribution of economic activity across local municipalities in South Africa. The first main chapter uses a new dataset on business registrations that spans from 1800 to 2011, to examine the relationship between the creation of Regional Industrial Development Programme zones in 1982 and their removal in 1991, and the spatial distribution of firm entry in South Africa. The creation of these zones marks the introduction of Sub-Saharan Africa’s first Special Economic Zones. However, little is known about the effects of such programs. Since incentives in these policies are expected to reduce the costs of registering and doing business to attract industries in targeted regions and create long lasting effects that make the 3 regions economically sustainable, it is expected that the creation of these zones will increase firm entry and the impacts will persist after the removal of policy incentives. Using merged data on the location of Regional Industrial Development Programme zones and the business registration database, empirical results show that the creation of RIDP zones was positively associated with firm entry when the policy incentives were still present, and after the removal of policy incentives, firm entry decreased. However, these results show that in the manufacturing and services sector, the reduction of entry after the removal of RIDP zones did not completely offset the positive effect of the policy on entry. This finding is consistent with the presence of agglomeration economies in the manufacturing sector. Overall, results from this chapter suggests that regional policy incentives were important in encouraging private sector development in marginalised regions of the country, although the policy did not create long-run economic benefits. The second main chapter empirically examines the impact of an important aspect of the local business environment that affect regional economic incentives, that is, crime rates, on entry of firms across local municipalities in South Africa. South Africa’s crime rates are high by international standards and surprisingly, little is known about how crime rates affect business activity in the country. Since crime rates increase the fixed costs of entry and the costs of doing business, it is expected that high crime rates in a region will lead to fewer firms entering the market, because only more productive firms will have expected profits high enough to justify paying the entry costs. This chapter merged crime data with the business registration database and found out that crime rates, particularly, property crimes have a deterring effect on firm entry and the effect is large for firms in the wholesale and retail sector. These results are robust to using rainfall shocks as an instrumental variable for crime rates, to control for the fact that crime might be a consequence not a cause of firm activity. The third main chapter empirically examines whether crime rates affect another aspect closely related to firm performance, that is, the availability of labour through migration across local municipalities in South Africa. Since high crime rates increases the costs of living in a region and this will reduce net income and utility associated with living in that region, it is expected that high crime rates will reduce in-migration. The chapter merged crime data used in the previous chapter with migration data created from 2011 population census and showed that contact crime rates reduce migration of labour into municipalities. 4 These results are also robust to using rainfall shocks as an instrumental variable for crime rates, to control for reverse causality. Empirical results also show that the effects of contact crime differ by population group and skill level, with the effects stronger for male, unskilled and black African migrants. Since unskilled and African migrants will have lower expected wages to cover the costs of crime because of lack of high paying work opportunities and labour market discrimination. In addition, it is expected that the effect of crime is high for these population groups compared to skilled and white migrants, who have the income and the mechanisms to mitigate the risks of crime, like picking better neighbourhoods to live in. Estimated results by gender also reveal that the effect of contact crimes is high for male migrants as compared to female migrants suggesting that increases in contact crimes reduces in-migration of male individuals by a higher magnitude as compared to females. This result is consistent with the view that male African workers may be more mobile than females, and most of these migrants are unskilled. Important policy messages from these findings are that any industrial development strategy and economic policy in the country that seeks to create jobs in marginalised municipalities or improve integration of local labour markets should also consider other measures, besides targeting disadvantaged regions with policy incentives. This thesis provides empirical evidence that crime prevention measures are likely to be one of the important tools for industrial policy. This implies that the government should put measures that improve policing and security. In addition, the private sector and international development organisations should continue to engage the government to put in programs and projects that reduce crime rates in the country. These may include, providing support for research that examines the main causes of crime rates in South Africa.