Whence Reform?

A Critique of the Stiglitz Perspective

By Marek Dabrowski, Stanislaw Gomulka and Jacek Rostowski

Fulltext                  Excerpts



1.    Introduction


That the transition from communism is largely over can be seen in the fact that the battle over its historical interpretation has begun. An early salvo in this battle has been fired by Joseph Stiglitz [1999a and 1999b], at the time Senior Vice-President of the World Bank, in a blistering attack aimed at both the Washington consensus (of the IMF, World Bank and US Treasury) which, he claims, foisted tragically mistaken policies on the countries emerging from communism, and at unnamed ‘western advisers’ who misunderstood both the true nature of capitalism and the nature of the reform process. Furthermore, Stiglitz contrasts two models of transition. On the one hand, he says, is China, where institutional reform has become gradual, building on already existing solutions, and economic growth has been rapid. On the other, there is Central Europe (CE) and the former Soviet Union (FSU), above all Russia, where reformers and their western advisers, both institutional and personal, organised a ‘blitzkrieg’ on existing economic structures, pursuing a ‘clean slate’ policy by which they tried to “engineer the new, clean, and pure ‘textbook institutions’ of a private economy.” Some western advisers were just bolsheviks ‘a rebours’, only “using the right textbooks this time” (or so they believed). The result has been a massive fall in economic activity. This occurred, says Stiglitz, partly because superficial modern economics textbooks informed the thinking of the advisers, rather than the deeper and more subtle Austrian approaches of Hayek and Schumpeter, or more advanced and rigorous modern journal articles of such authors as Stiglitz himself.




We do not feel directly targeted by this criticism, for although we have all been advisers (and one of us has been also a decision maker) during the transition, we are all East European. Moreover, we believe that some of the points Stiglitz makes are important for understanding what has happened in the transition (and indeed what occurred during the development of capitalism in West). This is particularly the case when he points, in the context of corporate governance, to the impossibility of quickly building long agency chains”.

However, we shall argue that the picture Stiglitz paints of the transition is fundamentally flawed in three key respects:

1. his mis-interpretation of the facts of the Chinese transition;

2.his mis-description of the facts of the Russian transition; 

3.his failure to consider the implications of the success of the “third model”, which is represented by some Central European and Baltic transitions (particularly the Estonian, Hungarian and Polish).

The result of these flaws is a deep misunderstanding of the real choices open to decision makers in Russia at the beginning of the 1990s, and therefore a profound mis-interpretation of history. …



2.   Stiglitz’s Key Hypotheses. 


Organizational Capital


  1.The organizational capital present in enterprises under socialism is still very valuable under conditions of transition. It needs to be preserved rather than destroyed.

  2.‘Entrepreneurial efforts that arise out of or spin-off from existing enterprises may be particularly effective

  3. A high rate of new enterprise creation was not to be expected in the FSU given its lack of a history of ‘market oriented entrepreneurship’. …

  4. Given the low probability of new enterprise creation and …it is particularly important to promote entrepreneurial restructuring in existing [state] enterprises.


Social Capital


  5. Social capital encompasses the civil norms which allow trust, and this in turn is important for all forms of inter-temporal trade. Such capital is vital to a functioning market economy.

  6.The social norms which existed under socialism are inadequate for a market economy, yet the development of new norms is bound to be time consuming. Imposing laws which do not correspond to existing norms is likely to fail because they will not be implemented in the right spirit (if at all).




Short Agency Chains and Privatisation


7. In the more developed market economies long multi-stage chains of agency relationships have developed…. Long agency chains, such as those implicit in voucher privatisation schemes (where workers are agents for managers, managers for funds and funds for citizens) are likely to generate fraud and inefficiency, contributing to output collapse and impeding subsequent recovery. This is why the method of privatisation matters.

8. In the early transition, therefore, agency chains should be as short as possible. …

9. Another way of establishing short agency chains is to privatise to the various stakeholders in the business (workers, suppliers, customers, representatives of the  local community) who have a long-term relationship to the enterprise. ….



Privatisation, Corruption, Asset Stripping and Capital flight


10. The ‘Coasian’ hypothesis that it does not matter whom assets are privatised to initially because initial buyers will sell out to the highest bidder, and the highest bidder will be the most efficient potential manager of the assets, is wrong in transition economics. The most efficient potential managers cannot buy the assets because they do not have the wealth. …

11. Inefficient new owners will therefore asset strip their companies. …

12. Inefficient new owners will invest much of their wealth in the political system, so as to obtain preferential access to further state property to asset strip. This leads to the corruption of both the political and economic systems.

13. Privatisation carried out in this way will not be accepted as legitimate by society, …. This provides a further reason for asset stripping by new owners and capital flight, the latter made easier in Russia by premature capital account liberalization.




Bankruptcy and Restructuring


14. Bankruptcy is unlikely to result itself in much restructuring. In the West most restructuring takes place before bankruptcy.

15. Pre-bankruptcy restructuring should be fostered by encouraging decentralization of decisions within SOEs, so that the constituent parts can look for customers and suppliers outside the existing firm and engage in ‘horizontal discussions’ with other units so as to rebuild organizational relationships ‘from the ground up’, leading to recombination of the units into new firms.

16. Making workers unemployed, whether through the bankruptcy or even the restructuring of existing firms, is likely to be economically inefficient. While employed they are at least producing something, whereas - because of a lack of ‘market-oriented skills’ among entrepreneurs - they are unlikely to find employment in new firms. The same holds for the capital made unemployed through bankruptcy.

17. Finally, the wrong firms are likely to be bankrupted since under socialism finance was not allocated on a commercial basis, with the result that the distribution of debt is largely random in the transition, and so is bankruptcy.



Inequality, Liberalisation and Stabilisation


18. The rapid transition attempted in the FSU and Eastern Europe has led to increased inequality, increased poverty and reduced life expectancy.

19. Rapid liberalisation did not lead to rapid growth, at least in Eastern Europe.

20. The same is true of stabilization. Because of the need for relative price adjustments, very rapid reduction of inflation may actually hinder micro-economic adjustment



3. Two Key Definitional Issues



...In his inter-temporal comparisons of China and Russia, Stiglitz uses the official growth rate of GDP as the yardstick of the success of reforms. This standard measure ...t is in our view inappropriate, indeed misleading, in the initial period of transition. The reason is that, at that time, economic developments in the FSU and CE reflected not so much the quality of current reforms, but the pre-reform crisis conditions, which led to the collapse of the Soviet style economic, military and political system…

We therefore propose to measure success of reform in transition countries by their ability to recreate the (institutional, legal and economic) conditions for rapid and sustainable growth. This ability is indicated by the increase of output from the start of recovery. ....

  As regards the concept of ‘privatization’, we use the standard measure of the share of output which originates in the private sector. This concept refers, thus, to the privatisation of the entire economy, and not necessarily to the privatisation of enterprises. Such economy-wide privatisation may occur, as it did in China or Poland, through the fast expansion of a new private sector relative to the rest of the economy, rather than through privatisation of existing SOEs.



4.   Three Models of Transition from Communism.


Much of Joseph Stiglitz’s argument is based on the contrast between the success of the Chinese transition and the economic failure so far of the Russian. We suggest that Stiglitz has misunderstood the reasons for China’s success, and misunderstood the facts (and therefore the causes also) of Russia’s failure. Finally he has ignored the success of some of the central European and Baltic countries, and therefore has misconceived the real requirements for successful transition.


A key reason for China’s success was indeed its underdevelopment, The first way this has helped has been the fact that only one fifth of the workforce was employed in the state sector at the start of the reforms in 1978. It was therefore possible to build a whole new industrial sector from scratch, taking advantage of the underemployed labour ...

... the second feature of China’s underdevelopment: its low productivity and the opportunities which exist as a result for rapid catching-up. This is known as the “Gerschenkron phenomenon” of the “relative advantages of backwardness”.

… the absence of privatisation and of an ‘institutional big bang’ has not inhibited the growth of corruption in China, which seems to be on a scale comparable to that in Russia (see Section 6). The same may be true of asset stripping and the amount of capital flight from China (see Section 8)…

Thus we reject the idea that China’s economic transition has been successful because it: (1) achieved efficiency enhancing restructuring by conserving the organisational capital of its SOEs; (2) sustained useful social norms inherited from socialism and avoided corruption thanks to gradual institutional development; (3) avoided harmful types of asset stripping by privatising to stakeholders and maintaining ‘short agency chains’…


As regards Russia, Joseph Stiglitz makes three important factual errors and, we suggest, one fundamental error of interpretation. The first error is the statement that manufacturing industry was privatised to outside ‘oligarchs’ and the attendant belief that had it been privatised to inside ‘stakeholders’, the Russian economy would have fared better. …. In fact, some 70 percent of manufacturing firms was initially privatised through Variant II of the mass privatisation scheme to workers and managers, who are the classical stakeholders. The subsequent evolution of the ownership structure in the direction of outsider control was the result of free choices made by the insiders to sell out, and was in any case very modest … Thus restructuring failed in Russia in spite of the very privatisation to stakeholders that Stiglitz recommends. Second, the oligarchs made their money not through asset stripping manufacturing firms, but through their preferential access to cheap credits from the Central Bank of Russia, the preferential privatisation to them of natural resource based (not manufacturing) firms, and their ability to arbitrage the difference between controlled domestic natural resource prices and free world market ones. …. Finally, a premature lifting of capital controls did not contribute to asset stripping, as Stiglitz claims, for the simple reason that capital controls were never lifted….

Stiglitz also claims that failure to privatise to insiders was part of a wider desire to destroy existing institutions and norms, which were perceived by the reformers as tainted by communism. This created a moral void, which has led to the flowering of crime and corruption. We have pointed out that….many of the old social norms relating to economic activity, which had existed under socialism, died before the collapse of communism, indeed, their death was a major reason for its collapse. This is the major difference in interpretation regarding Russia between us and Stiglitz. Thus, in our view, in many areas there were no existing norms at all   for the reformers to build new institutions upon. ....


Finally, Stiglitz largely ignores the successful transitions of Central European and Baltic countries. One of the most successful has been Poland, ... After a short but sharp contraction of about 15 percent in 1990 and 1991, the economy has grown at an average rate of 5 percent per annum, ... Estonia, Latvia, Lithuania, Hungary and Slovenia have also experienced rapid growth in the last few years. What enabled Poland, ... to achieve such success in the 1990s ...?



The Polish model of transition consisted of five main elements: 

1. complete liberalisation of de novo private sector entry into almost all areas of economic activity (January 1989 and January 1990);

2. adoption of the pre-1939 commercial code (formally 1982, really 1988-1989) and abolition of communist party organisations in SOEs (end 1989), the latter giving real power to the workers’ councils that had formally exercised it since 1981;

3. very rapid price liberalisation (during 1989 the share of freely determined prices rose from 25% to 90%);

4. introduction of hard budget constraints on SOEs and a sharp reduction of inflation ... through tight fiscal, monetary and wages policies (January 1990), ...

5. current account convertibility of the currency and almost complete foreign trade liberalisation (January 1990).



This was certainly the most rapid programme of market reforms that had been implemented until that time, although it should be noted that until 1995 privatisation of SOEs was very rapid only with respect to small firms. The results of the programme were:

1. the rapid introduction of market prices based on relative scarcity and world prices for traded goods;

2. a financial squeeze on SOEs, which forced them to quickly release excess labour and physical capital (so called asset privatisation);

3. the maintenance of a minimum tolerable level of effective corporate governance in SOEs thanks to the workers’ councils;

4. very rapid development of the de novo private sector. … De novo private activity, which at first grew rapidly in services, began to expand in manufacturing industry at a rate in excess of 40 percent per annum already in 1991. An average rate of growth in excess of 25 percent was then sustained for about five years. More recently foreign direct investment has begun to make a significant contribution to growth. All this has resulted in a doubling of manufacturing output since l99l.



This Polish experience (and to an extent that of other Central European and Baltic countries) demonstrates:

1. the usefulness of pre-existing rules and institutions as Stiglitz stresses (workers’ councils, a commercial code, ....);

2. the importance of macro-economic stabilisation and ... imposition of hard budget constraints, which Stiglitz questions;

3. the importance of the liberalisation of prices, trade and entry, which Stiglitz challenges;

4. the possibility of rapid growth of new private enterprise, which Stiglitz flatly denies.…..



...In spite of large differences in industrial structure and history between Central Europe and Russia, the differences between Russia and China are much greater, particularly as regards level of development at the beginning of the transition. Comparing Russia with Poland and Estonia in particular, and Central European and Baltic countries in general, is likely to be at least as fruitful as comparing it to China. In examining Stiglitzts 20 theses (arranged in six main groups) we therefore draw on all three experiences.



5. Organizational Capital.



  ... Stiglitz treats organisational and social capital as indistinguishable from reputational capital .... We prefer to deal separately with each .... Organisational capital we take to be the value of a productive organisation over and above the value of its assets. It exists because things can be done more efficiently within the organisation, due to habits, formal rules and trust which have built up. Organisational capital is thus a name we give to the efficiencies which Coase [1937] used to explain the existence of firms. ..... We define reputational capital as generating trust between organisations and between organisations and individuals. Within an organisation., we take reputational capital to generate trust which adheres to an individual, whereas if the trust adheres to a post (rather than its holder) then we may be talking of organisational rather than reputational capital. Social capital ... consists in formal rules and institutions (legal codes, court systems), ... in general rules of behaviour, which we expect even from those whose reputation we do not know..


Our view is that Stiglitz is wrong on both counts in Thesis 1: the organizational capital present in socialist SOEs is not very valuable, and it is not particularly hard to reassemble it in a more efficient way in new private enterprises. The evidence for the former proposition is the fact that, in transition economies, de novo private firms have been found to be far more efficient than all other categories of firm — both privatised and state owned — with the exception of firms run by foreign direct investors (see the large survey-based literature, including Estrin and Wright [1999], Schaffer [1998]). The reasons for this seem to be the following:


1. Much of the organizational structure of the SOE is unsuited to operation in a market — e.g. the almost complete absence of a sales and marketing function, poor ability to evaluate production costs and business risks, and the considerable attention paid by management to production, investment and purchasing (quite rational in a shortage economy) and lobbying the state (rational in a state run economy).

2. The capital structure and the labour skills structure of SOEs are usually unsuited to producing goods profitably at the relative prices which obtain after the liberalization of prices, entry and international trade. Yet ‘whittling down’ SOEs to the activities which might be efficient is a far harder job than creating new tailor made private firms out of selected physical assets bought from various SOEs or imported from abroad, and labour with particular skills. The reason for this is the resistance of stakeholders to restructuring, compared to the free hand which new entrepreneurs have in deciding what goods to produce, what assets to buy and what labour to hire on competitive markets.


The relative ease of reassembling factors of production (including organizational capital) if the environment is right for new firm creation is demonstrated by the extraordinarily rapid growth in the number, size and output of new private firms in Poland since 1990, which we described in the previous section.

While it is true that FSU entrepreneurs developed skills in evading and profiting from government regulations before the transition, and that many operated on the interface between legal and illegal activity, the same was true of Polish and Estonian entrepreneurs. This is illustrated by the figures for ‘economic crime’ in Poland, which show a dramatic fall when the transition began, government intervention in the economy was reduced and many commercial activities were decriminalised. The result was that the number of such ‘economic crimes’ fell from 26,741 in 1985 to 6,042 in 1990. These facts throw doubt on Stiglitz’s Thesis 3, that entrepreneurs at the start of transition do not have the skills needed for creating and developing new businesses. The relevant question is therefore not: how could Russian reformers, and their Western advisers, have been so naive as to think that entrepreneurs would create a host of new firms to restructure the economy? but rather: what were the conditions which allowed entrepreneurs to do so in Poland or Estonia but not in Russia? We shall return to this question in the Conclusion.




6.    Social Capital.


We agree entirely with Stiglitz on Thesis 5 about the critical role of social capital, and also largely on Thesis 6 that the social norms under socialism are inadequate for a market economy. Indeed, we suspect that Thesis 6 may have been one of the factors which explains why rapid de novo private sector development occurred in Poland and other Central European and Baltic countries, but not in Russia. Of course, ex ante policy makers had no reason to suppose that this would prove to be the case, as we have argued in the previous section. Our main problem, however, is with the un-stated hypothesis that there existed social norms in the Soviet Union which would have allowed the maintenance of the existing structure of industry and the existing level of output for a considerable time. During this time, Stiglitz assumes, new norms could have been developed which would have provided the basis for the new institutions of an efficient form of capitalism. As stated earlier, our understanding is, rather, that the social norms which sustained the communist economic system died or decayed before the system itself collapsed. We can see this, in the case of Poland, in the increase in the number of ‘economic crimes’ which occurred during the period when belief in the socialist system collapsed between 1980 and 1985. In 1980 there were 7,659 such crimes reported, but by 1985 the number had increased to 26,741


Thus, governments were faced not so much with the task of transforming existing socialistic norms into new capitalist ones, but rather with a state of anomie, or of an absence of norms relevant to economic activity. ... waiting for the new norms to develop spontaneously before introducing new laws was not a practical option. New laws and institutions had to be introduced quickly in the hope that they would help society to develop the new norms it needed. ...

There have been a few backward or very isolated countries where the old socialistic norms continued to hold sway for a long time, Belarus and Turkmenistan being examples. .... In China, corruption (which we take to be a measure of the degree to which norms are not suited to the needs of a market economy) seems to be on a scale second only to that in the Commonwealth of Independent States (CIS), and ahead of Central Europe and the Baltics which experienced far more institutional shock therapy.

Social capital should be thought of as including not only unwritten norms and trust but also laws and formal institutions. In this respect Poland and Hungary were relatively well placed. Poland had a commercial code, which had been in operation since 1982. while Hungary built up a body of company laws from 1985....


  7. Short Agency Chains and Privatisation.


We agree very strongly with Theses 7 and 8 which link the optimum length of agency chains to the quality of institutions, social capital and markets, and partly with Thesis 9 about the best way of establishing short agency chains. Indeed, we consider that here Stiglitz has provided an important insight into both the development of capitalism in the West and the transition process in central and eastern Europe. The rapid growth of owner-operated businesses has been the hall-mark of those central European and Baltic transitions which have been successful. We also consider that workers’ self-management played an important part in limiting the fall in output and the amount of criminal asset stripping in the state sector in Poland. Indeed, two of us argued at the time in favour of the advantages of workers’ management as a form of corporate governance, against the advice coming from the World Bank. Thus we also agree with Stiglitz that the method of privatisation matters, and that methods involving long agency chains (such as the Czech variant of voucher privatisation) are not likely to be very successful for the reasons of breach of trust which Stiglitz describes.


Indeed, we would suggest that there is a very good reason why long agency chains require a long time to develop. This is the need for providers of agency services to build up the ‘reputational capital’ which will induce the potential purchasers of these services to trust them. Typically, providers of monitoring and screening services come to markets with a certain amount of net worth in the form of physical and financial assets, which stand as guarantee (i.e. implicit collateral) for the implicit liabilities which they undertake in the process of monitoring and screening. Given a reasonably well functioning legal system, breaches of trust will lead to a loss of physical and financial assets. As providers of agency services build up reputational capital, their need for physical and financial capital to reassure customers declines. As a result, their capital costs fall and they obtain a higher return on equity. This causes a virtuous circle, as such cost savings on capital further increase the value of the reputational capital of the firm, which customers know would be lost in the event of a breach of trust. This is why even in mature capitalist economics, new forms of agency services are often provided by firms which are already well established as providers of other services which require them to have a large amount of ‘reputational capital’. In the early transition, of course, hardly anyone has any reputational capital to loose.


Turning to Thesis 9, we have already expressed our view that giving corporate governance rights to workers can be quite efficient in the early transition. However, we have to enter three important caveats. First, workers’ management is no different from other forms of agency arrangement, in that it requires some pre-existing tradition (social norms) and reputational capital, which will allow it to function. Representatives elected to the workers’ council need to feel obliged to defend the interests of theft electors, and not to just further their own. This is particularly important in a context in which privatisation is expected quite quickly, and with it the disappearance of the workers’ council, so that representatives may not have the incentive of prospective re-election. In Poland this tradition existed in Russia it did not. Russian reformers could no more create such a tradition ab ovo than they could have implanted Quaker business ethics in the new capitalist class. This point is confirmed by the fact that, as we pointed out in Section 3, most SOEs in Russia were privatised via Variant II of the voucher privatisation scheme predominantly to the workers. Yet these workers generally proved unable to ensure effective corporate governance for theft firms, or to prevent the asset stripping which Stiglitz so deplores.


Our second caveat relates to the inclusiveness of Stiglitz’s proposal to privatise to ‘the various stakeholders in the business’. He subsequently goes on to identify workers, suppliers, customers and representatives of the local community as the relevant stakeholders. We suspect that such an approach would lead to paralysis and endless conflict between the various entrenched interests within the firm. A vital part of transition is switching suppliers from traditional to new, more efficient, often foreign ones. Equally, sales often need to be shifted to new customers who are prepared to pay more. Entrenching the position of existing interest groups within the decision making structure of the firm would thus be a recipe for failure to restructure, and therefore for failure of the transition. While agency chains should be short, the corporate governance structure needs to be clear, with power concentrated in the hands of those whose interests are most closely aligned with an efficient use of the capital embodied in the firm. This may, under certain circumstances be the local community [Wiles 1977], or as in Poland the workforce (in spite of the well known drawbacks of workers’ management as compared to mature capitalism).

Finally, since we do not think that socialist SOEs had much organizational capital of value, we do not consider it important to ‘save’ it in its pre-transition form by entrenching the rights of the various stakeholders. Stiglitz’s proposal for ‘government by discussion’ between the various stakeholder groups within firms thus seems to us a striking example of the ‘communitarian romanticism’, which pervades much recent writing on the importance of social and organizational capital in the transition [e.g. Grey 1995].




8. Privatisadon, Corruption, Asset Stripping and Capital Flight.



  There are aspects of Theses 10-13 on these four interrelated issues with which we agree, and aspects with which we strongly disagree. Clearly, it matters whom assets are privatised to initially, if only because the re-allocation of assets in the market takes place in time, and not instantaneously. The question posed by Thesis 10 is therefore: do assets reach efficient owners faster by being privatised to the population as a whole, to the friends of top politicians, by being kept longer in state hands and then being privatised to more efficient owners (such as foreign investors), or by being privatised rapidly to ‘stakeholders’ as Stiglitz suggests?

In the previous section, we accepted that “voucher privatisation” has not been efficient, largely for the reasons Stiglitz expounds. We also agree with his [l999b] view that sale to foreigners cannot be a widely applied solution in the early transition. Others have made the point that state ownership does not prevent asset stripping when the state is weak and, above all, when there are many price and trade controls [Aaslund 1999]. We have already expressed our doubts regarding ‘multi-stakeholder’ privatisation, and pointed out that privatisation to managers and workers (the classical stakeholders) is largely what actually happened in Russia, so that the barrier to the transfer of assets to more efficient users must have lain elsewhere.


An important question is: to what phenomena can the term ‘asset stripping’ legitimately be applied in transition? In the West, the term is usually applied to so-called ‘corporate raiders’, who seize management control of a company and sell its most easily realisable assets for prices which together are higher than they paid for the firm, but are said by objectors to be less than the ‘true’ value the firm would be worth (i.e. could be sold for) if’ the raiders were willing to spend time on its restructuring and resuscitation, Implicit in the concept is the idea, so dear to Stiglitz, that there exists in the firm organizational capital which is somehow undervalued by the market for corporate ownership. However, although the old organizational capital might be lost, the Polish, Estonian and Hungarian experiences suggest that the new organizational capital, which is created in its place, might be more valuable.

The biggest weakness of Stiglitz’s perspective on this topic is the link he makes between privatisation as it was carried out in Russia, corruption, and the lack of restructuring due to bad corporate governance and asset stripping (Theses 12 and 13). The greatest fortunes in Russia were made not through privatisation of manufacturing (where the restructuring was most needed), but through privatisation of the seignorage stream generated by the Central Bank of Russia (Dabrowski and Rostowski 1993, Boone and Hoerder, 1998], through arbitrageing the differences between controlled prices for raw materials within Russia and free market prices abroad [Aaslund 1999], and through favouritism in the privatisasion of natural resource based industries, which began in 1995. It was the Russian treasury which was directly asset stripped, not the manufacturing firms, and it was the fortunes made at that time which have sustained the political power of the oligarchs.


We have already pointed out in Section 4 that capital flight from Russia has not, as Stiglitz claims, occurred as a result of premature liberalization of capital movements, but in spite of the absence of such liberalization. This is a mistake that Stiglitz extends in [1999b, p14) to most transition economies except, strangely, to China, where in fact it appears to be of similar magnitude, relative to exports, as in Russia. The presence of such large capital flight in China suggests that it is paying a price for the lack of clarity in its property rights and corporate governance structures, and for the very reasons that Stiglitz describes with regard to Russia.

Nevertheless, Stiglitz does bring home the point that those of us who thought that rapid privatisation was needed to prevent asset stripping were probably excessively anxious about a phenomenon which, in the light of experience, seems to have been unavoidable and not necessarily undesirable.

Indeed, an important question is: why did asset stripping of the typical Western kind, and indeed of the kind carried out perfectly legally by Polish worker managed manufacturing firms (i.e. sale of assets for which outsiders have better use), not in fact happen in Russia? If it had, then the dynamic private sector development which so benefited Polish manufacturing, and which was largely absent in Russia, might have occurred. We shall return to this question in the Conclusion.




9.  Bankruptcy and Resfructuring



We agree with Thesis 14 that bankruptcy is unlikely to result itself in much restructuring. We would point out, however, that a mechanism for the seizure of assets by creditors is very important both for developing the credit market by encouraging lenders to lend (among other things for restructuring), and for imposing discipline on borrowers. Furthermore, much restructuring does indeed occur before bankruptcy, but this is because the parties know that bankruptcy exists. While decentralising decision-making in SOEs, so that the constituent parts of the firm can recombine into new firms (Thesis 15) is probably harmless, we know of no successful application of this approach. In Poland, at the beginning of the transition constituent plants of multi-plant SOEs were given the right to secede, and hundreds of them did so during 1989-91. Although this was healthy for the economy, it may have looked a lot like asset stripping to the workers left behind in the less profitable plants of a multi-plant SOE! Also, we know of no cases of “recombination” of such “secessionist” plants31.


While Thesis 16 is true in the short run everywhere, if its logic were accepted very little restructuring would happen anywhere. We have already shown that transition economies have not in general suffered from the shortage of ‘market­oriented entrepreneurial skills’ that Stiglitz attributes to them. The key thing, of course, is the imposition of hard budget constraints (something which Stiglitz himself stresses as important in [1999a]). Zero unemployment was maintained under central planning through price controls and generalised excess demand for both labour and goods (in the latter case leading to generalized shortages). Since prices did not reflect relative scarcities under such circumstances, restructuring and privatisation are impossible until prices are liberalised and generalised excess demand is abolished. Furthermore, with a guarantee of full employment and soft budget constraints facing firms, there is nothing to prevent workforces (whether organised in trade unions or not) from demanding ever higher wages. Under such conditions, full employment can only be maintained by an accommodative monetary policy, leading to accelerating (and ultimately hyper-) inflation [Rostowski 1989]. Indeed, inflation can only be stopped by allowing unemployment to reach the natural (or NAIRU) level, and this is certainly not zero! High inflation makes restructuring and privatisation much harder, because: (1) rapidly changing nominal prices make it harder to observe the underlying relative prices on which restructuring and privatisation via sales have to be based; (2) accounting and taxation systems are not inflation adjusted so that the lack of transparency caused by high inflation makes tax avoidance and asset stripping of outside shareholders, stakeholders and the state much easier; (3) in an economy with a large number of remaining price controls, inflation provides huge opportunities for enrichment through arbitrage. For all these reasons full employment, be it of labour or capital, is simply not an option in transition.


Furthermore, it is striking that in Central Europe hard budget constraints led large SOEs to shed a large amount of labour, to change their product mix etc. This is so-called “defensive restructuring”. (Grosfeld and Roland [1996] found that a sample of 600 large SOEs in the Czech Republic, Hungary and Poland shed about 30% of their workforces in the first two years of the transition. This indicates that the dilemma Stiglitz outlines in [l999b pp.25-6], according to which privatisation is impossible without prior restructuring, but restructuring by the state was held by the reformers to be impossible, has in fact a very simple solution.

As regards Thesis 17, that wrong firms are likely to be bankrupted, its validity depends on the country concerned. Those countries that suffered from very high inflation in the pre-transition period and during the early transition35, found that this tended to reduce the real value of enterprise debt owed to banks very considerably (thus in early 1990 such debt was the equivalent of about 7 percent of annual GDP in Poland, compared to 70 percent in Czechoslovakia). As a result, debt was not a major factor in the failure of most firms in such countries in the early transition. Far more important was low cash flow due to inefficiency and bad product mix36. Furthermore, in those countries that had experienced some marketisation before the fall of communism (Hungary, Poland, Yugoslavia, the USSR), SOE debt was often the result of lobbying of central planners and/or banks by firms. Firms were thus, to a certain degree, responsible for their debt. Stiglitz’s point does apply to countries in which strict central planning was maintained until the very end and where inflation was relatively low, but there were only two of these: Czechoslovakia and the German Democratic Republic.



10. Inequality, Liberalisation and Stabilisation



  It is certainly true, as Stiglitz points out, that the transition to capitalism in the former Soviet block and Yugoslavia has led to a large increase in inequality. This is, perhaps, not entirely surprising to those who have read Marx. However, there is no evidence that, as Stiglitz implies, inequality has increased because reformers adopted “shock therapy”. Nor is there evidence that where life expectancy has fallen this has occurred because of rapid transition strategies, as is again implied by Stiglitz.


To verify Stiglitz’s conclusions we divided the transition countries into “early liberalisers”, “late liberalisers” and “non-liberalisers” according to the EBRD classification [EBRD 1999, p64]. We then looked at the changes in Gini coefficients between 1987-9 and 1993-5 reported in Milanovic [1998], and at the change in life expectancy reported by Stiglitz himself in [1999b] following EBRD [1999]. It turns out that early liberalisers experienced a much smaller average increase in their Gini coefficients than late liberalisers and non-liberalisers (Table 3). Similarly, early liberalisers experienced an improvement in life expectancy between 1989 and 1997 (by 0.7 years on average, with no early liberaliser experiencing a decline), while late and non-liberalisers experienced equivalent average declines (although six of the 18 countries in these two categories also experienced an increase in life expectancy). We have a similar story when we look at early stabilizers (although we now have one country, Latvia, where life expectancy fell.


Things become less clear when we look at privatisation. On our more restrictive definition of “early privatisation” (a combined score of 7 or 8 in the EBRD [1995] privatisation index4° plus over 50% of GDP in 1994 generated by the private sector), early privatisers perform far better than late privatisers, both in the deterioration of the Gini coefficient and in life expectancy (only one early-privatiser, Lithuania, has a fall in life expectancy, and that is only 0.3 years). On the other hand if we adopt a looser definition of early privatisation, requiring only that over 50% of GNP be generated in the private sector by mid-1994, then the difference between the two groups becomes smaller as regards increases in Gini coefficients, and early privatisers have on average only a tiny increase in life expectancy (also four of the nine countries in the group register declines)”. Nevertheless, even in this case, there is no evidence that early privatisers performed worse on inequality and life expectancy than late privatisers! It is worth noting that Russia is not classified as an early liberaliser, as an early stabiliser or as an early privatiser (except on the looser definition).


Turning to Thesis 19, that rapid liberalisation did not lead to rapid growth, the first thing to note is that Stiglitz’s position is quite contradictory. In [1994 and 1999a] he stresses that while both competition and private property are required for a well functioning economy, if a choice has to be made competition is the more important.

Yet in [l999b] Stiglitz argues that countries that have liberalised fast have performed worse than those that did so slowly. How then is competition to be achieved without liberalisation? The issue is not addressed directly, but from the texts it seems that, for Stiglitz, competition is mainly to be achieved by “building up the regulatory framework for competition” [l999a, p19]. While such a framework is useful in an economy in which price, trade and entry controls have been abolished, it is not much use on its own. Poland between 1985 and 1989 provides a good example of an economy with a fair degree of de-monopolisation in manufacturing, which suffered from severe stagnation as a result of ubiquitous controls.


Stiglitz’s second point is that in transition, where large relative price changes are required, inflation should not be pushed so low that such changes can only be achieved through nominal falls in some prices (since nominal prices are said to be sticky downwards). Reformers’ attempts to eliminate inflation therefore interfered with the dynamic adjustment of the economy and were misguided. Whether Stiglitz is right on the downward stickiness of prices and wages in transition economies or not, he is certainly wrong on what reformers actually did. Rightly or wrongly, with one or two exceptions such as Croatia in 1993 and Bulgaria in 1998, rates of inflation have been reduced relatively gradually, and cumulative price increases since the beginning of the transition have been more than adequate for the necessary relative price changes, registering in the hundreds of percent at the least (with the sole exceptions of the former GDR., the Czech Republic. Hungary and Slovakia).

As regards liberalization, along with EBRD’s Transition Reports, 1995-1999, we interpret the evidence as supporting the proposition that, in the countries which liberalized faster, outputs not only stopped falling earlier, but also started to recover earlier, so that rapid price and trade liberalization accelerated, but did not magnify, the transformational recession.



11.  Conclusion.



The three central gradualist theses which Stiglitz has put forward have been demonstrated to have either no empirical basis or be contradicted by evidence. Firstly, protecting the existing organisational capital of SOEs, and building the restructuring of the economy on it is not vital for a successful transition — as is shown by the experience of China, Central Europe and the Baltic countries. Secondly, privatising initially to insiders - as actually happened in Russia - need not be an effective mechanism for protecting this organisational capital and enabling successful enterprise restructuring. And thirdly, it is new private enterprise which is vital for a successful transition and this new private sector can grow very rapidly during transition, as both the Central European and Chinese cases show.

We also argued that the idea that the new post-Communist economic system should have been built in such a way as to exploit pre-existing (communitarian) social norms rather than copying real existing western capitalist institutions is mistaken since one of the effects of the total crisis of the old system was that such norms no longer had binding social force (Section 5), The problem facing governments was anomie, or an absence of the relevant norms. As regards asset stripping, we suggested that the relevant question is: why did it not happen in Russia in such a way as to re-allocate physical capital to new private firms, while it did in some other countries, particularly in the Baltic countries and Poland? We argued that the key point is the extent to which hard budget constraints are imposed on existing SOEs and entry costs are lowered for new firms, which means the extent of price stabilization and price, entry and trade liberalization.


What, then, remains of value of Stiglitz’s views on economic transition to capitalism? Certainly his recommendation of ‘short agency chains’ and the criticism of voucher privatisation schemes to outsiders which follows (i.e. the Czech case, but not the Russian one as he mistakenly asserts). Also, his support for worker management of SOEs and privatisation to workers and managers (with, however, the caveat that, as is shown by the very different Russian and Polish experiences, this will only succeed in those countries where the existing social norms provide support for such institutions). Finally, there must be a minimum legal framework of commercial and property law and law enforcement, without which markets will not develop and market competition will be very limited. However, Stiglitz may have overstated the amount of framework needed because of his tendency to stress regulation and competition enforcement, rather than the protection of property rights and the enabling of competition. Indeed, it seems unlikely that China satisfied Stiglitz’s requirement for competition enforcement and regulation of markets in the early years of its transition.


If the Russian transition did not fail for the reasons Stiglitz puts forward, why then did it not succeed until 1999? We suggest it was for the very reasons that the Russian reformers themselves explained - in advance - at the very beginning of the transition. Stiglitz seems unaware that there was a fierce debate in Russia during 1991-2 about the sequencing of the reforms. On the one hand were those who wanted to start with the hardening of budget constraints, and the liberalisation of prices, trade and entry, and disinflation, while on the other were those who argued that SOEs by their very nature could not respond to such changes in the economic environment, and that therefore privatisation had to come first. The Russian key reformers and most of their Western (and Eastern) advisers were in the former camp, The opponents of reform, including the Communist Party of the Russian Federation, in the latter. The 1992 attempt at budget hardening/disinflation cum liberalisation, the Gaidar plan, failed because of opposition in the Supreme Soviet and insufficient support from President Yeltsin. It was this failure which impaired macroeconomic stabilization and enterprise restructuring, and which led the reformers to switch in the autumn of 1992 to the ‘privatisation first’ strategy which they had previously opposed.


As had been predicted by the reformers themselves, this meant that enterprises were under less pressure to divest physical assets and labour they did not need, which would have provided the resources needed by de nova private firms to expand. The failure to disinflate and liberalise thoroughly made it hard both for new firms wishing to set up, and old firms wishing to restructure, to obtain clear information on relative prices. The continued practice of large flows of subsidised credit and entry bariers undermined the credibility of the whole transition strategy. In particular, the Russian reforms failed to convince citizens that Russia is the best place to invest their savings. A large flight of capital has been the result. The reform strategy adopted also made privazization for large amounts of cash to foreigners almost impossible and, through the opaqueness it created, enabled the privatisation of central bank seignorage revenue on a massive scale, as well as facilitating criminal types of asset stripping of companies by managers or controlling shareholders (through the lack of transparency and high inflation introduced into company accounts). By the time disinflation began to succeed in 1995 many of the destructive processes described by Stiglitz had either been completed, or had at least begun.


Transition happens in time, hence it is likely to be path dependent and, as Stiglitz stresses, sequencing therefore matters. However, the sequencing error which was made in Russia was not so much, as Stiglitz claims, in placing privatisation before the creation of a legal, competitive and regulatory framework for market processes, but rather the much more important error of placing privatisation before liberalisation cum hardening of budget constraints and disinflation. In an environment of near hyperinflation, massively distorted prices and soft budget constraints, the best designed privatisation scheme, occurring in an environment with the requisite legal structure (commercial and company law) and good regulatory and competition frameworks, could not have succeeded in allocating firms to efficient managers who would have successfully restructured them. On the other hand, as the Slovak and Croat experiences show, given free markets, predictable prices, hard budget constraints, and the basics of commercial law (this last requirement is Stiglitz­consistent), even outrageous degrees of ‘crony privatisation’ need not lead to the huge disorganisation and social costs experienced in Russia.


This sequencing problem is related to what appears to be the most important difference between Stiglitz and us. In our understanding of the evidence, most Russian SOEs, with their immense inherited problems, were not capable of growth whatever the regulatory framework. Given the financial, managerial and other constraints, and poor positive incentives, they were capable merely of defensive restructuring. In our perspective, the success of transition depends above all on a rapid creation of conditions — institutional, legal, microeconomic and macroeconomic — which are conducive to the development and growth of a new private sector.

The adjustments in policies which Russia adopted in the aftermath of the August 1998 financial crisis are creating the right macroeconomic conditions. With the exception of some authors, notably Kornai [1990], the early conventional view probably over-estimated the positive impact on performance of fast privatization of SOEs and, by the same token, failed to appreciate sufficiently the key role that a completely new private sector would play in recovery and growth.











OK Economics was designed and it is maintained by Oldrich Kyn.
To send me a message, please use one of the following addresses: ---

This website contains the following sections:

General  Economics:

Economic Systems:

Money and Banking:

Past students:

Czech Republic

Kyn’s Publications

 American education

free hit counters
Nutrisystem Diet Coupons