Uncertainty and Growth in Transition Economies

Abstract

The paper investigates the relationship between fundamental uncertainty, a recurrent theme in post-Keynesian economic literature, and economic performance in transition economies. Uncertainty in the transitional economic environment is enhanced by factors such as institutional transformation, political and social instability, and legacies of the past. To capture the changes in the levels of transition-specific uncertainty, the authors have designed the uncertainty index, based on a weighted selection of Heritage Foundation and Freedom House data. The correlation between the uncertainty index and growth is strong and clearly negative. Panel data analysis based on a growth model, supplemented by variables to simulate transitional cycle, and performed on a sample of transition economies for the period 1995–2002, confirms that high levels of transition-specific uncertainty had a negative impact on economic growth.

Keywords:

• uncertainty
• economic transition
• institutions
• economic growth

Acknowlegements

Helpful comments by Professor Sheila C. Dow and by two anonymous referees of this journal are gratefully acknowledged. The usual disclaimer applies.

Notes

1 The term was first used by Mlčoch (1992), for whom transition is the final stage of “institutional wandering” of ex-socialist countries, bringing the evolution of institutions in these countries back to the normal growth path. In many postsocialist countries, however, the institutional “wandering” persists.

2 See also Glickman (2003).

3 See also Keynes (1921). For a survey of early theoretical views on uncertainty see Wubben (1993).

4 See also Eichner (1985: 64–67).

5 Davidson (1991b: 69) ironically describes economic subjects in neoclassical theory as “robot-decision-makers”.

6 For a critique of the neoclassical maximization hypothesis and of the rationalist conception of action see Hodgson (1988: Chapters 4 and 5).

7 According to the classification in Davidson (1995), however, the bounded rationality theory is placed among the immutable-reality (ergodic) models of uncertainty.

8 See again Keynes (1937: 213–214). Dow (1995) is an excellent discussion of various levels of uncertainty as perceived by economic decision-makers. See also Dequech (1999: 417n).

9 These sources may well be taken as a collateral to the postulates of the Washington consensus (although the justification of the latter is far from definitive; see Lavigne 2000).

10 Alston (1994: 55) correctly observes, that in transition economies “(p)oliticians can change (…) formal institutions (…) quickly, but they have little direct influence on informal institutions such as norms of behaviour or culture, which have an equal if not more important impact on market behavior”.

11 For a discussion about these issues from the perspective of the new institutional economics see Feige (2003).

12 “Conventions play a large part in the formation of judgements under uncertainty” (Dow 1996: 19). For a list of procedures (conventions) followed by economic agents in conditions of uncertainty see Lavoie (1992a: 56). See also Earl (1995: 124–126).

13 See Johnson et al. (2002) and Linz (2002) for an analysis between property rights protection and investment strategies in transition economies.

14 For details on these reforms see Lavigne (1999).

15 On the impact of transaction costs on growth see also North (1987).

16 The Heritage Foundation Index of Freedom is available for transition economies for the time period 1995–2002. It runs on a scale from 1 to 5, 1 denoting very good quality of institutional environment, 5 denoting a very poor quality. The overall score is comprised of 10 subscores (subindices) evaluating: (1) trade policy; (2) fiscal burden of government; (3) government intervention in the economy; (4) monetary policy; (5) capital flows and foreign investment; (6) banking and finance; (7) wages and prices; (8) property rights; (9) regulation; and (10) informal market activity. The index is liberally based, which does represent a slight drawback to our analysis: (1) because the liberal agenda might be more suitable for developed than developing economies; and (2) in any case the “varieties of capitalism” issue can thus not be considered. But it is very systematic and evaluates a broad range of categories.

17 Table A1 in the Appendix presents subindices which were used to create uncertainty index for 2002. The index is based on Heritage Foundation indices for government intervention, banking and finance, property rights, regulation and informal market activity and on Freedom House indices for political rights and civil liberties. In the first column are data on per capita GDP for 2002.

18 The span from 1 to 5.4 results from combining indices that run on different scales, one from 1 to 5, the other from 1 to 7. The weighted average of both, which is our measure of uncertainty, runs consequently from 1 to 5.4.

19 We are aware of the potential drawbacks of the index, especially its inability to capture country specific transition paths and the subjectivity of weights. An additional potential future development of the idea would be to consider the importance of the three main sources of uncertainty and change the respective weights in time. Also, the data published by Fraser Institute, Freedom House or any other similar indices (also Economist Intelligence Unit, Heritage Foundation, Transparency International, World Bank evaluate the quality of various institutional aspects) are based on a specific definition of what a good institutional framework is. Therefore the data also fail to recognize to some extent the potential of different varieties of capitalism to work equally well. But on the other hand, a different approach would be too subjective.

20 The improvements in the uncertainty levels would be even more visible had the data set spanned to the beginning of transition. Unfortunately, Heritage Foundation publishes indices only from 1995 on.

21 Due to lack of data for uncertainty index, either for 1996 or 2002, the figures exclude Macedonia, Serbia and Montenegro, Georgia and Tajikistan.

22 The outliers represent Slovenia which has by far the highest GDP per capita among transition economies. If measured in purchasing power parity, the wealth of Slovenia compared to other transition economies is significantly reduced.

23 Georgia, Moldova, Serbia and Montenegro and Macedonia were excluded from the panel data analysis due to a large number of missing data.

24 Data were obtained from Economist Intelligence Unit and Transition Report (2004).

25 On the basis of this theoretical foundation, we considered an extended production function: Y = f(K,L,H,A, uncertainty, variable used as a cycle proxy, variables capturing regional characteristics), where Y stands for GDP, K for total capital, L for labour, H for human capital, A for technology. Given that GDP per capita was used as a dependent variable, the extended production function was divided by L to give the final theoretical foundation for empirical evaluation: y = f(k,h,A, uncertainty, variable used as a cycle proxy, variables capturing regional characteristics), where y stands for per capita GDP, k for gross fixed capital per capita. As a proxy for human capital per person (h) a share of secondary school enrolment in total population was used, given that years of schooling per person were not available.

26 Cummulative inflow of foreign capital per capita indicates total inflow of foreign capital since 1989 till the end of each year in the observed period (1995–2002).