The Life Cycle of South African Manufacturing Firms

Working Paper
Published on 1 June 2015

Abstract

In this paper Andrew Kerr (2015) uses a new source of firm level panel data to describe and better understand the life cycle of South African manufacturing firms. An important point to emerge from recent firm level work is that there are substantial cross country differences in the size of firms and in the relationship between plant age and size (Hsieh and Klenow, 2014), generated by differences in the importance of selection and within firm growth as well as changes in size of entering cohorts (Sandefur 2010, Davies and Kerr 2015). These differences may indicate substantial misallocation of resources and low firm level investment in developing countries. South Africa is an interesting country to study in this regard because the importance of informal firms is small and the problems usually associated with running formal firms in developing countries, such as a lack of contract enforcement and the resultant lack of hiring non-family managerial labour (Akcigit, Alp, and Peters 2014), thin land markets, and financial frictions (Hsieh and Klenow 2014), are not as prevalent. But clearly - given the extremely large levels of unemployment - South African firms are not generating enough employment. A lack of within firm growth, high equilibrium levels of exit or low equilibrium levels of entry are possible reasons for the low employment equilibrium South Africa finds itself in. The author investigates these possibilities further by exploring the life cycle growth of manufacturing firms using a new source of firm level panel data. The Cape Town RSC panel data set is a census of formal firms in the City of Cape Town between 2000 and 2006. Yearly administrative data is available on basic firm characteristics (wage bill and revenue) and operating status and an initial survey collected more detailed data on around two thirds of the firms that were operating in 2000. The author uses the subset of manufacturing firms in this paper.