Black Economic Empowerment and economic
performance in South Africa
James A. Robinsonx
We thank María Alejandra Palacio for her outstanding research assitance. Massachussetts Institute of Technology, Department of Economics, E52-380, 50 Memorial Drive, Cambriudge MA02142. e-mail: firstname.lastname@example.org. z
The EDGE Institute, 11th Floor, Braamfontein Centre, 23 Jorrisen Street, Braamfontein, Johannesburg and the University of Witwartersrand, e-mail: email@example.com. x
Harvard University, Department of Government, IQSS, 1737 Cambridge St., N309, Cambridge MA02138; e-mail: firstname.lastname@example.org.
Non-technical Policy Brief
We lay out a conceptual framework for evaluating the impact of BEE on economic growth, both directly through its e¤ects on …rm behavior and indirectly through its impact on political stability. We emphasize that because of the way the phenomena of BEE has evolved since its inception in 1993, the extent and form of BEE are unlikely to be optimal from the point of view of economic growth. This suggests that there may be possibilities for making growth enhancing reforms.
A priori, there are good arguments that suggest that several of the components of BEE could have positive and/or negative e¤ects on productivity and investment and hence on growth. Which of these e¤ects are stronger is an empirical question. We use data on the BEE ratings of companies from Empowerdex along with a dataset of companies listed on the JSE to examine the e¤ects of the aggregate BEE score of a company (according to the BEE scorecard) and the BEE Ownership score, on investment, labor productivity and pro…tability. The heart of this analysis are two …rm level identi…cation strategies using the allocation of the investment portfolio of the PIC and the presence of a single shareholder with 50% or more of the equity, as sources of variation in BEE across …rms. Both of these variables strongly predict the extent to which an individual …rm engages in BEE, and the second plausibly predicts the BEE ownership share. The main result of these regressions is that so far, BEE seems to have had very little impact on …rm behavior. If anything there is weak evidence that both BEE in general, and the BEE ownership score, has a negative e¤ect on investment and labor productivity, though this is not statistically signi…cant at standard con…dence levels. There are several possible interpretations of these …ndings: (1) Data problems make inference di¢ cult to impossible. (2) It is too early to empirically test for BEE’s implications for the economy, (3) The economic costs and bene…ts of BEE cancel each other out so that the aggregate e¤ect is zero, (4) The big issue may be the aggregate e¤ects (avoidance of populism) which we cannot estimate at the …rm level, (5) BEE actually does not have large e¤ects on …rm behavior though it does in‡uence distribution.
Even though it is di¢ cult to say which of these views is correct, there exists a solid basis
to make some policy recommendations on BEE. This is because it is clear that some aspects of BEE are quite unsatisfactory and there clearly are areas where the policy needs to be clari…ed and amended. The arguments in this paper suggest three main policy conclusions. The …rst follows from the fact that a priori arguments suggest that while the emphasis on the transfer of ownership in BEE may have been important in the decade after 1994 for provisionally securing property rights, it is unlikely to have large positive e¤ects on …rm productivity or economic growth. Moreover, enduring property rights security certainly requires a broader base than can be provided by such a policy. We therefore believe that the policy needs to be changed to de-emphasize ownership and increase the focus on aspects of it more clearly linked to increased productivity and the broader transformation of the economy. A simple policy to achieve this would be:
Please join StudyMode to read the full document