The Political Economy of Institutions, Stability and Investment: A Simultaneous Equation Approach in an Emerging Economy. The Case of South Africa

Abstract

The modern theory of investment identifies the importance of uncertainty to investment. A number of empirical studies have tested the theory on South African time series, employing political instability measures as proxies for uncertainty. This paper verifies that political instability measures are required in the formulation of the investment function for South Africa. It also establishes that there are distinct institutional factors that influence the uncertainty variable such as property rights and crime levels. We find that rising income and property rights lower political instability, and that rising crime levels are positively related to political instability. The inference is that political instability in South Africa may not represent uncertainty directly, since it is systematically related to a set of determinants. Instead, uncertainty would have to be understood as being related to a broader institutional nexus that in concert may generate uncertainty for investors. The paper highlights the significance of getting institutions right to ensure that uncertainty is kept to a minimum by providing a predictable long-term environment. Stability at a systemic level appears crucial if investment rates are to rise in South Africa and this paper demonstrates that stability in turn is driven by a sound institutional environment that has multiple dimensions.

Notes

 1. Perhaps most closely associated with the work of Jorgenson – see for instance Jorgenson (1963). The discussion in Bertola and Caballero (1994) and Ferderer (1993) is also instructive.

 2. See the discussion in Abel and Blanchard (1988: 249), and Sensenbrenner (1991: 819).

 3. Note that in Abel and Eberly (1994) both of these alternatives are combined in a general model.

 4. See Tobin, 1969; Tobin and Brainard, 1968, 1977– and see also the discussion in Mussa, 1977; Hayashi, 1982; Abel, 1983; Sensenbrenner, 1991.

 5. This explains why equilibrium q > 1. Under irreversible investment and uncertainty marginal q must rise above 1 before investment occurs.

 6. This result is consistent with one strand in the literature on uncertainty – see Abel (1983) and Hartman (1972). In these analyses, firms face constant returns to scale production technology, while the marginal value product of capital is again a convex function of an uncertain price faced by the risk neutral competitive firm. By Jensen's inequality greater uncertainty then raises the marginal valuation of one additional unit of capital, thereby increasing investment.

 7. In effect, opportunities to invest are akin to financial call options, which expire (hence lose their value) when exercised. Dixit and Pindyck (1994) show that investment can be treated as a call option with symmetrical results to those obtained from the dynamic programming problem reported above. Further, Drazen and Sakellaris (1999) show that while good news about future events boosts investment, news that additional information is on its way tends to depress (delay) investment, emphasising the value of information.

 8. In the South African instance, see Wood (2000).

 9. For a fuller exploration of the theoretical underpinnings of this view and related empirical evidence, see Diamond (1992), and the wider comparative discussion in Fedderke (1997).

10. In traditional South African growth debates, this view underpins the interpretation of the South African growth path that is present in the contributions of Bromberger, 1974, 1978; O'Dowd, 1974, 1978 for instance. Presentation of international evidence is myriad with both affirmative (see for instance Muller, 1997) and more sceptical views present in the literature (see for instance Huntington, 1984 though the scepticism is revised somewhat in Huntington, 1991).

11. The quintessential example of such change is perhaps the American war of independence. Economic interests were if anything harmed by severing colonial ties, yet the principle of ‘no taxation without representation’ identified a conflict of principles deemed (on this interpretation) sufficiently fundamental by American citizens to merit war. In the South African debate, this position has been consistently argued as a (perhaps the) fundamental force for political change by De Kadt (2001).

12. The contributions here are too numerous to mention, and the present discussion is not aimed at an exploration of alternative conceptions of justice and their welfare implications. Yet, certainly the work of Rawls (1971) and his respondents has had a substantial impact on conceptions of welfare in economics.

13. See for example the discussion of development as freedom, in Sen (1999).

14. See Engerman and Sokoloff, 2003; Fedderke, 2004; Luiz, 2008 and the seminal work by North, 1990.

15. See for instance the discussion in Acemoglu, 2003: 27.

16. See North, 1990; Acemoglu, 2003; Fedderke, 2004; Luiz, 2008.

17. See for example, Luiz (2001) who explores the dynamics of crime in South Africa and its relationship with economic variables.

18. Note that this accords well with the finding in Fedderke and Klitgaard (1998) that social, institutional and political variables hang together by means of ‘webs of association.’

19. See Fedderke et al. (2001a).

20. See the more detailed discussion in Johansen, 1988; Johansen and Juselius, 1990, 1992.

21. See Johansen and Juselius, 1990, 1992; Pesaran and Shin, 1995a, b; Pesaran et al., 1996; Wickens, 1996.

22. See Greenslade et al., 1999: 3ff.

23. The user cost of capital is introduced into the model in order to assess the marginal cost of capital.

24. We also undertook extensive bounds testing of both the base-line models specified in this paper, as well as the mixed strategy estimations reported below. The full set of bounds tests results is reported in a separate paper, Fedderke and Luiz (2006). The core findings are consistent with the specifications tested structurally in the present paper. In particular, Fedderke and Luiz (2006) note that bounds testing implies: (i) that various measures of social cleavage hang together; (ii) in general, social cleavage does not impact on growth but is driven by economic growth instead; (iii) with the exception of racial cleavages which drive distributional conflict; (iv) together with property rights which also impact on distributional conflict; and (v) finally, distributional conflict in turn impacts on economic growth. It is these patterns of association, together with the literature on the political economy of South Africa that in effect inform the specifications tested in the present paper.

25. We also widely tested the alternative specification in which only political rights appears in the instability relationship, while property rights are excluded. The implausibly large elasticities of the weak test uniformly remain present in all these specifications, with output elasticities of approximately 20 in the investment relation, and 10–20 in the instability relationship. Full results available from the authors on request.