The role of institutions in Shaping foreign capital: Evidence from South Africa and Zimbabwe

The purpose of the study is to investigate the impact of institutions (particularly property rights) on foreign capital in two Southern African countries (namely Zimbabwe and South Africa). This is motivated by the recent theoretical emphasis on the role played by institutional factors such as property rights protection and the rule of law in determining economic outcomes. The thesis addresses a critical issue concerning the measurement of institutional variables in empirical work. While the political science and political economy fields have produced several institutional indices, most are time truncated. Most existing indices are therefore only useful in cross-country and panel studies, but not useful for country-specific studies. This limits any path-dependency exploration on the link between institutions and economic outcomes. To address the measurement challenges, a new set of institutional indicators measuring de jure property rights, de jure political freedoms and de facto political instability are constructed for Zimbabwe for the period 1946 to 2005 by this dissertation. The Fedderke et al (2001) Delphi technique used in the construction of the de jure property rights and political freedom indices fits in well with the North (1990) institutional theoretical framework employed here. Making use of the newly constructed property rights index in a multivariate cointegration framework, the study establishes the impact of property rights on foreign direct investment (FDI) in Zimbabwe for the period 1964 to 2005. The empirical results indicate that property rights are consistently an important explanatory variable of FDI, even after controlling for periods when there are no significant new foreign capital inflows. Other significant variables ii that explain FDI in Zimbabwe include the real gross domestic product (GDP), capital intensity, the external debt to GDP ratio, political instability as well as the educational levels. In the case of South Africa, the thesis investigates the impact of property rights, domestic risk and neighbourhood effects on the absolute levels of FDI and portfolio investment stocks as well as the relative share of FDI in total foreign capital stocks. Domestic risk is measured by the South African-American sovereign spread. This risk measure captures both the default and currency risks since the sovereign bonds are denominated in different currencies. Neighbourhood effects are captured by introducing the property rights index for Zimbabwe as an explanatory variable for foreign capital stocks in South Africa. The empirical evidence for South Africa shows that while domestic risk reduces the absolute levels of FDI and portfolio investment, secure property rights positively affect both FDI and portfolio investment. These results are in line with the theoretical proposition of the portfolio diversification literature. In terms of neighbourhood effects, the results indicate that weak property rights in Zimbabwe lead to an increase in the absolute level of FDI stocks but reduce the absolute level of portfolio investment stocks in South Africa. The results suggest that when property rights in Zimbabwe deteriorate, long-term foreign investors may relocate their investment from Zimbabwe to South Africa. However, weak property rights in Zimbabwe have a negative spill-over effect on the short-term portfolio investment flows in South Africa. Regarding the composition of foreign capital stocks in South Africa, the results show that the relative share of FDI in total foreign capital stocks is positively related to property rights but negatively related to domestic risk. This suggests that FDI in South Africa is not inalienable. As such, when property rights weaken or domestic risk goes up, the relative share of FDI in total foreign capital stocks decreases. This occurs because foreign direct investors tend to iii reduce the levels FDI relative to other foreign capital inflows when faced with expropriation risk. The results also show a negative relationship between the property rights index for Zimbabwe and the relative share of FDI in total foreign capital stocks in South Africa. This indicates that deteriorating property rights in Zimbabwe result in an increase in the relative share of FDI in total foreign capital stocks in South Africa. In both the cases of South Africa and Zimbabwe, our empirical results confirm the existence of feedback effects from FDI to output. This supports the notion that FDI has some output-enhancing effects in the host country. Overall, the study shows that property rights, domestic risk and neighbourhood effects in the recipient country are critical in shaping the absolute levels of foreign capital stocks as well as the composition of foreign capital stocks. The main policy recommendation for Zimbabwe and South Africa is that, to increase the levels of foreign capital inflows, the host country governments should ensure sound institutions both at home and in the region. Another recommendation of the study is that, if the South African government wants to shift the composition of its foreign capital stocks from portfolio investment to FDI, they should put in place policies that promote secure property rights and low domestic risk