Challenge, November-December 1999

Joseph Stiglitz is senior vice president and chief economist of the World Bank. The following article was the keynote address in the Governance, Equity, and Global Markets segment of the annual World Bank conference on development economics in Europe, Paris, June 21-23, 1999.

The chief economist of the World Bank presents a detailed critique of Western reform proposals for transition economies, notably shock therapy. He believes institutional realities were ignored in light of substantial economic theory that suggests they are important.


Two economic events have dominated the twentieth century--the socialist/communist experiment that began in the Soviet Union in 1917 and ended with the fall of the Soviet empire just under a decade ago; and the transition of these economies to market economies during the past decade. There is widespread agreement about the lessons from the first--central planning cannot replace markets, and even market socialism cannot replace the incentives associated with capitalism--incentives to produce goods more cheaply, to produce what consumers want, and to innovate. But the lessons from the transition to market economies are far more ambiguous, and it is upon these that I want to focus my remarks today.

First, the facts, to say the least, are jarring. With only one exception, the countries under question have done more poorly since the transition to the market economy than before. Indeed, the overwhelming majority of these countries have yet to reach their 1989 gross domestic product (GDP) levels, meaning that, on average, these countries today are worse off than they were before the transition. These results are corroborated by other indicators of well being, such as longevity of their inhabitants. Again, the results are bleak--life expectancy in these countries on average has fallen even while world life expectancy has risen by two years. But even bleaker are the statistics on poverty. For eighteen of the twenty-five countries for which we have data, poverty on average has increased from 4 percent to 45 percent of the population, using the $4-a-day standard, revealing the devastating changes in standards of living. Evidently Russia and several of the other countries of Eastern Europe and the former Soviet Union have managed to repeal one of the long-standing "laws" of economics: the trade-off between growth (efficiency) and inequality. They have managed to experience lower growth and greater inequality since the transition than before it began.  There is one marked success that I have deliberately left out of the data presented so far--China--and fortunately, it is a large success, with a population greater than all the other transition economies combined. Over the twenty-year period since its transition began, it has succeeded in growing at an average annual rate of 9.5 percent, and poverty has been reduced from 60 percent to 22 percent. Three-quarters of the increase in incomes in low-income countries in the past decade was inside China. If the separate provinces of China were treated as separate countries--and some of them have populations of 60 million or more, larger than all but a few of the countries of Eastern Europe and the former Soviet Union--then twenty of the twenty fastest-growing countries during the past two decades would be located in China.(n1)

Not long ago, several of the economists who had played important roles in the early days of the transition expressed their optimism concerning Russia, publishing books with such titles as How Russia Became a Market Economy and The Coming Boom in Russia. The optimism was widely shared. What was surprising was the absence of data to support that optimism. Let us be clear: It is not hard for a country rich in natural resources to find investors abroad willing to exploit those resources, especially if the price is right. Far more difficult, however, is creating an industrial or service-based economy. Consider the investment in natural resources and manufacturing in Russia as a percentage of all foreign investment; in 1994, investment in manufacturing was a mere 7 percent, compared with 57 percent in natural resources. In 1997, while our data remain incomplete, preliminary statistics show the number for manufacturing to have slipped even further, to approximately 3 percent. Investment was attracted for natural resources, but not for manufacturing.

There are other seeming anomalies. The wisdom that emerged in the years immediately after the beginning of the transition was clear: "Countries that liberalize rapidly and extensively turn around more quickly [than those that do not]."(n2) Countries that worked most rapidly to bring about macroeconomic adjustment and lower inflation rates were more likely to do better than those that hesitated. Those that pushed rapid privatization were more likely to do better than those that went more slowly. Among the countries of Eastern Europe, whose history, geography, and prospects are at least more similar to one another than they are to the Central Asian or Baltic states, early liberalization and overall average growth exhibit no positive relationship. If anything, they appear to have a negative correlation. In fact, those countries that grew most quickly in the early years of transition were no more likely to grow quickly in more recent years. Recent evidence also casts doubt on the dogma of low inflation; within Eastern Europe, those economies that grew the fastest were those with higher, not lower, inflation.(n3) At this juncture, then, inferences are far less clear than many asserted in 1996. Differences in initial conditions and geography imply that different countries faced different opportunities, and these differences in prospects might account for the success or the speed of transition (implying that any observed correlations may simply be spurious, an observation that those ready to defend the orthodox prescriptions often seemed to overlook). It is nonetheless remarkable how different the world looks today than many thought it would earlier in the transition. Countries that were castigated a few short years ago for the slowness and incompleteness of their reforms, such as Uzbekistan and Slovenia, are performing rather well, whereas other countries heralded as models of reform, such as the Czech Republic, are now encountering difficulties.

These data are striking, and they should at least serve to undermine confidence in the previous conventional wisdom. But it should be clear that I am not presenting these as definitive results. I am simply arguing that perhaps there has been excessive and unwarranted confidence in the previous conventional wisdom--a conventional wisdom that, unfortunately, continues to unduly influence policymakers in some circles. What is remarkable about these results is that they are not the results of any data mining. In each case, I put forth an hypothesis, looked for standard indicators, for example concerning corporate governance, that had been constructed by others (and had in fact been used earlier to justify the old conventional wisdom), and ran simple correlations. Whether or not coefficients are statistically significant, the results would still serve to cast doubt on the previous views, which argued for contradictory statistically and economically significant relationships.

I am confident that these results will stir up a data-mining industry, providing gainful employment to Western economists, even as conventional mining grows in relative importance in Russia. Refinements are needed. As I comment later, there are inherent problems in many of the key concepts, raising questions about the corresponding indicators: Is a firm really privatized when it issues shares and the shares are owned by a public enterprise? Or when the privatization is financed by a "soft" loan from a bank? And how do we know whether the loan is soft?

One further observation concerning interpretation of any results, and opportunities for, shall we say, further data exploration. While the large number of countries involved in the transition might suggest a wealth of variation in policies on the basis of which we could make meaningful inferences, the fact of the matter is that the countries differ enormously in their history, geography, and endowments. Not controlling for these variables throws into question any inferences. Yet even the wealth of such relevant variables still leads to an underidentified system. Worse still, the policies pursued are, at least to some extent, endogenous, affected by the circumstances of the country. Is a country's poor performance due to the fact that it liberalized slowly, or was there some factor leading to both slow liberalization and low growth performance? The fact that some of the slow reformers have, over the long haul, done so well may convey more information about the impact of the pace of reforms on long-term growth than the observation that some country has both liberalized and grown more rapidly in the short run. This is especially the case because some of the faster reformers were in Eastern Europe--countries where the possibility of quick accession to the European Union, with the access to those vast markets, provided an additional impetus for fast reform and, perhaps, additional returns to those reforms, than might have been the case for the landlocked countries of central Asia. Earlier studies that failed to either recognize or take adequate account of the differential position of the different countries failed to reflect on the endogeneity of the policy environment. In such a case, the interactions of liberalization and growth should be treated with a high degree of skepticism.

In any case, whatever results we obtain should be looked at with caution. It is only through looking at the interplay between mutually supporting pieces of theory and evidence that we can attain any degree of confidence. Data on declining GDP and increasing poverty are consistent with independently arrived--at statistics suggesting worsening health conditions.(n4) And, as I shall argue, the observed problems are precisely those anticipated by information economics, with its emphasis on corporate governance. This picture itself is reinforced by the reported regression results. Even the results on inflation are not inconsistent with other bits of data and theory. Inflation is traditionally a problem because it interferes with the working of the price system. But when attempts to suppress inflation are associated with a movement out of a market system and toward a heavy reliance on barter, the price system works even more imperfectly.(n5) Theories of downward wage and price rigidity have contended that pushing inflation below a critical threshold actually interferes with the dynamic adjustment of the economy.(n6) Clearly, the critical threshold should be higher when the required adjustments are greater. For the economies in transition, these adjustments are particularly large.

At one level, the facts speak for themselves: The transition to the market economy has not delivered what its more ardent advocates promised. To be sure, the countries have not done everything right; but no country does, and if capitalism is so frail that it cannot survive normal human fallibility, then its virtues, at least for this world, may surely be questioned. After all, one of the longstanding criticisms of socialism was that it was too utopian. The quandary of the failure of so many experiments is particularly vexing when economic theory was so clear in its predictions: Distorted prices, central planning, and attenuated incentives arising from the absence of clear property rights meant that resources were not efficiently allocated. Reducing those distortions, decentralizing decision-making, and privatizing--even if not done perfectly--should have moved countries closer to their production possibilities curve. Output should have soared--instead it plummeted.

When I put forward this argument at a conference in Washington a couple of months ago, there were several answers.(n7) One was that output was not measured accurately. This may be true, but there are arguments suggesting that output was being overestimated (given the huge amounts of barter trade, market prices may provide an overestimate of true values of many commodities), as well as underestimated.(n8)

A second argument was that such declines were to be expected, as loss-making enterprises were shut down. This kind of reasoning illustrates a level of confusion that should be embarrassing to an economist. The relevant losses should be calculated by looking at the value of the output relative to the opportunity cost of the resources. If a firm is making a loss (in this sense), it means that if its inputs (the resources it uses) are redeployed to other, more productive uses, output (at least correctly valued) should rise. Of course, output is not increased by moving resources from low-productivity uses into idleness. The opportunity cost of such resources is zero (or the value of leisure).

To be sure, the production possibility curves of the economies in transition are lower than they would have been had better investments been made. But that is not the issue. The point is that movement toward a market economy should lead an economy from a point interior to the production possibilities curve toward the frontier and then, as investment is allocated more efficiently, move the frontier outward. Hence, growth should have been higher both because of the rapid movement toward the frontier, and the more rapid movement outward of the frontier, as investment was more efficiently allocated. That did not occur.

In my lecture in Washington, I set forth several hypotheses attempting to explain these phenomena. At one level, there is now a growing consensus: It is far more difficult to create a market economy than was previously thought to be the case. An institutional infrastructure--including not just contract enforcement, but also competition policy, bankruptcy law, and financial institutions and regulations--is required. And although institutional infrastructure requires time to create, without it, privatization may lead more to asset stripping than to wealth creation. A regression run on the effect of privatization alone on GDP growth in the transition economies revealed no statistically significant effect of privatization on growth. If  however, the regression is run with additional data for corporate governance, for those economies receiving high scores for corporate governance and restructuring, privatization did have a positive and statistically significant effect.

In previous discussions, I have focused on several issues:  

° The lack of corporate governance--combined with free capital mobility and the contractionary macroeconomic policies--made it more profitable for those who had obtained control over Russia's huge manufacturing and natural resources to divert their wealth abroad through a process of asset stripping rather than investing at home.

° While well-functioning market economies require both competition and private property, reforms focused more on the latter than on the former, and the differential performance between China and Russia may be explained in part by these differences in emphasis. Prospects for privatization--except through sales to foreigners or restructuring into smaller, more locally controlled units--were inherently bleak.

° Another fundamental difference between China and Russia was the former country's emphasis on enterprise and job creation, rather than an exclusive focus on restructuring existing assets.

° The destruction of social and organizational capital in the process of transition--from an admittedly weak base--and perhaps partly attributable to the speed and structure of reform, without a corresponding emphasis on the creation of new social capital, also may have played an important role in the failures in Russia and some of the other countries of the former Soviet Union.

While all participants in the reform debate recognized the importance of political processes, many of the reform strategies were based on assumptions concerning the political dynamics for which there was neither theory or evidence. The sequence of reforms matters, in part because particular reforms lead to the creation of interest groups that can either facilitate or block subsequent reforms. Underlying some of these misguided views was a naive belief in what are called Coasian processes that once property rights were appropriately assigned, efficient institutional arrangements would evolve. Such beliefs ignored both general theories that suggested that there may be inefficient institutional equilibria and that evolutionary processes need not be efficient.

Another underlying confusion had to do with the interface between macroeconomics and microeconomics. There was a widespread belief in Say's law (an idea that was discredited three-quarters of a century ago) that somehow, if workers who were underemployed in their current jobs could only be released into open unemployment, the market would respond and create jobs for them. Price stability (low inflation) is not an end in its own, but a means to more fundamental goals, such as faster economic growth. And when pushed too far--below a critical level--not only may the costs of pushing inflation lower not be worth the benefits, but the benefits may actually be negative. Macrostability does not, in itself, imply microrestructuring.

I want to focus more narrowly on the issue of corporate governance, and touch on other issues only to the extent that they are related to this issue.

Corporate Governance

In a lecture delivered at the end of the previous century, Alfred Marshall (1897), reflecting on what had been learned during the previous century and the unsettled issues for the one to come, discussed at some length the issue that we now refer to as corporate governance. How do we ensure that the manager of the firm acts in the interests of the "owners"--the shareholders? In nineteenth-century economics, there was simply a single owner-manager, and hence what we now call the agency problem simply did not arise. But even by the time Marshall wrote, it was clear that there was a separation of ownership and management among many, if not most, large enterprises and which represented an ever-increasing share of GDP. Their interests diverged.

Beginning in the late 1960s, I and others put this issue of corporate governance into the contexts of the problems of imperfect information and public goods.

° Owners had imperfect information concerning the opportunities facing managers and could not, by looking at outcomes, infer whether or not managers had made the right decision. indeed, imperfections of information were what necessitated the delegation of responsibility to managers.

° Managers not only knew this, but also could take actions that exacerbated the asymmetries of information, enhancing managers' discretionary authority.(n9)

° These problems would arise even if there was a single owner who delegated management responsibility.

But in most large companies, the diverse ownership gave rise to a public goods problem. If any shareholder works hard to improve the management of the firm, all shareholders benefit equally--management is a public good.(n10)

If managers have the means and incentives to act in their own interests, if their own interests do not perfectly coincide with those of shareholders, and if diverse ownership provides no (or inadequate) incentives for monitoring managers, then how can large firms function? If firm-value maximization is the key to the success of a market economy, and if, instead, firms maximize managerial returns, then what assurance do we have of the efficiency of the market economy?(n11)

There are four possible ways out of this dilemma: The first provides strong controls over the managers (the "firm") vis-a-vis all shareholders, even the smallest stockholder. Actions that are viewed as depriving the small shareholder of value are grounds for legal action against the firm and its management. Management may have a high degree of personal legal liability. Only two countries have gone this route and provided sufficiently strong legal protection so that diverse stock ownership systems have been sustained.

The second approach entails there being a single shareholder with a large enough stake that it pays him to provide managerial oversight himself.(n12) To be sure, there is still a presumption of an undersupply of managerial oversight. But we seek not perfection--simply workable solutions, or, from the perspective of the transition, at least solutions that avoid disaster. But this raises a new agency problem: How can one be sure that the majority shareholder will not take actions that represent his interests, to the disadvantage of the minority shareholders?(n13) One solution the obvious one--is that there be no minority shareholders, a return to the classic owner-managed firm. But this approach has severe limitations--it implies that the only way that firms could expand would be through borrowing. And though the famed model of Franco Modigliani and Merton Miller argued that corporate financial structure didn't matter, their analysis ignored crucial issues such as bankruptcy and asymmetries of information, issues that are at the center of the problem here. I will turn to the role of banks in a minute, but for now, I simply want to note that excessive reliance on bank/debt finance makes firms highly susceptible to fluctuations in demand and costs, significantly increasing bankruptcy probabilities with their high transaction and managerial costs.

This brings me to the third approach--that there be strong laws to protect minority shareholders against majority shareholders. Note that these legal protections are somewhat different from those discussed earlier. Actions taken in the interests of the "firm"--the long-run viability of the enterprise, job protection of its employees--might lower all shareholder values and enhance long-run managerial incomes, but they would not be a diversion of value from minority to majority shareholders. Thus, in a sense, these protections are "weaker" than the first protections. Here, the minority shareholders free ride on the self-interest of the large shareholder to ensure that shareholder values (as a class) are protected.

In a recent study done for the World Bank, Alex Dyck of Harvard has classified countries by the strength of their shareholder protection and by the concentration of shareholders.(n14) The hypothesis was simple: Only countries with very strong legal protections could support diverse ownership. The results confirmed that no country with weak legal protections had diverse ownership. From this perspective, one can perceive the weakness of the voucher privatization scheme--as attractive as the idea of creating a people capitalism might have seemed at first blush. It attempted to occupy a position in the chart that all countries--for good reason--had previously avoided: weak legal protections combined with diverse ownership. I shall comment later about where one goes from here.

There is yet another control mechanism, which stands the theory of the firm on its head: It is not shareholders (the nominal "owners") who really control the firm, but the banks.(n15) Concepts of control and power are, perhaps by necessity, ambiguous. The shareholders have rights, but when ownership is diverse, they must act collusively to exercise those rights; they simply lack the incentive to do so. If there is a single dominant lender, it pays him to monitor--or if there are a few lenders who interact repeatedly, so that they in effect act collusively with respect to monitoring, it pays to assign one of the group to be the lead bank, and hence the chief monitor. And if loans are short term, or if there is a succession of loans so that the borrower must repeatedly turn to the bank, then the lender has control rights, at least in the sense that if the borrower does not do as he wishes, he can force the repayment of the loan or force the firm into bankruptcy. Given that, under these circumstances, most firms cannot access other sources of funds, the threat of termination becomes a viable threat.(n16)

To be sure, in exercising their role as a monitor, banks are not concerned with value maximization; rather, they wish to minimize the probability of default.(n17) But in monitoring against default, they at least ensure that the worst mistakes--and the worst examples of managerial theft--are avoided.

From this perspective, we can see another source of problems in the privatization strategies followed by many of the economies in transition. Banks under the former socialist system were not real banks; they were banks in name only. They were not involved in screening and monitoring; they simply allocated credit as directed by the government. In the early days of the transition, I and others warned against confusing these nominal banks with real banks, and that one of the real tasks in the transition would be to create real banks.(n18) State banks, undercapitalized private banks, or private banks with an implicit guarantee of a bailout, are not, in this sense, real banks. They lack the incentive to screen and monitor on a commercial basis. Indeed, in some cases, they became a new source of soft-budget constraints; in others, they became the vehicle through which state wealth was diverted to the hands of the political cronies.

We can see, from this perspective, the difficulties facing countries seeking rapid privatization, which would or could not sell to foreigners, at least at any politically feasible price. There were, at the time, no legitimate wealth holders to buy the assets. Thus, a country could either use voucher privatization or have banks lend to someone to buy the asset. I have already discussed one of the key problems with privatization--the public good problem of management. Most of those involved in the voucher privatization were aware of this problem. (Indeed, I participated in discussions in Prague on this issue early in the 1990s.) But most thought that the voucher investment funds would take over the task of monitoring. They would aggregate interests, to the point where the public good problem was mitigated. This made sense, but the thinking needed to have gone further. But I remember asking at the time: Who was going to monitor the monitors.(n19) And so it turned out: The voucher investment funds provided a vehicle for high-powered abuse? The fact that the mutual funds were closed-ended exacerbated these problems. With open-ended funds, investors disgruntled with the performance of these funds could withdraw their money, providing an effective check on the abuses. That is why one of the recommended reform has been to convert closed funds into open-ended funds.

There is considerable evidence concerning the magnitude and significance of these corporate governance problems. In standard theory, closed-ended mutual funds should never sell at a discount (relative to the value of the underlying shares), for there is a simple strategy that would increase the market value--disband the fund and distribute the shares directly. Imperfections in the market for managers (takeovers) explains why in advanced industrialized economies, shares and close-ended mutual funds have often sold at a discount--of up to 10 percent. But in the Czech Republic, discounts soared to 40 percent or more.

Moreover, standard agency theory argues that when there is a controlling shareholder (with 20 percent or more of the shares), his interest in providing the public good of corporate oversight and management should increase the firm's market value. Empirical studies confirm these predictions. But in the Czech Republic, controlling interests gives the controller the right to strip assets, unchecked by legal restraints.

Thus, it appears that often when a single party gains control, market value plummets, reflecting the market's perception that control is more associated with asset stripping than with wealth creation.

As the companies needed cash injections to continue to operate, and as the weaknesses in the securities market made it clear that these injections could not be provided through the securities market (a telling criticism--the facade of capitalism had been created, but functioning capital markets, presumably the sine qua non of capitalism, did not function), they had to turn to banks. And if banks had been real banks, then the banks themselves could have performed some of this monitoring role. But under socialism, the banks functioned largely as vehicles for state-directed credits and so had little experience with ex ante due diligence and ex post monitoring.

But the problems with the banks went deeper. First, owning the firm provided enormous option value. If the firm turned out to be greater in value than it had paid, the owner kept the difference; if it turned out to be less in value, the owner simply went bankrupt. Worse still, owning the firm provided enormous opportunities for theft. The legal structure provided enormous opportunities for diverting the wealth of the firm into the hands of the "owner," with the owner having only limited personal liability. (In many countries, the returns per unit effort to such theft, given the weak legal structure, clearly exceeded the returns to efforts devoted to wealth creation.)

The theft-plus-option values associated with ownership meant that (in the absence of a competitive loan market) there were enormous returns associated with getting loans. The consequences were predictable, one might say almost inevitable: Loans were not necessarily allocated to those who were most likely to use the assets of the firm most efficiently, but to those with political connections, and who otherwise knew best how to "manipulate" the system.(n20) If there had been real banks, in the sense that I defined them earlier, these abuses might not have occurred, or at least might not have been so prevalent. But the absence of real banks meant that there was neither the incentive nor the capacity to perform the selection and monitoring functions so important to the success of a market economy, and so vital, given the corporate governance problems associated with equity.

In short, the process of transition in general and privatization in particular demonstrates an old lesson of market economics--incentives matter. But it also demonstrates a key lesson that was lost on many of the so-called reformers: Only under highly idealized situations do incentives result in efficient outcomes. Misdirected incentives can provide incentives for asset stripping rather than wealth creation. In many of the countries in transition, that is precisely what happened.

I have tried, so far, to argue that there is a nexus of institutions that make capitalism work. It is not just "private ownership," but financial institutions and legal structures, and these were deficient in the economies in transition. Privatizing without effective financial institutions and legal structures was entering into terra incognita--a bold experiment where already existing theory indicated strong reservations. In fact, however, the problems were even deeper, and the predictable pitfalls more profound. I have time here to mention but two aspects. The first has to do with capital account convertibility, a subject that one might think more relevant for a discussion of macroeconomic and exchange rate policy than for privatization. In a typical Coasian model, the mechanisms described earlier, while they clearly affected the distribution of wealth, might not have had much effect on the efficiency of the economy. After all, Coase argued that in the absence of transaction costs, the initial wealth distribution does not matter for efficiency, so long as property rights are clear. In a closed economy, however, these new wealth holders would have had an incentive to invest their wealth in activities that yielded the highest return. Coase, of course, ignored agency costs. And those who had demonstrated the greatest ability to garner for themselves political favors were not necessarily those most suited to wealth creation, or even to selecting managers who were most suited for doing so. Hence, entrusting the country's wealth to these new "robber barons" was not necessarily the best way of maximizing growth rates.(n21) But allowing these robber barons to take their cash flows out of the country changed their opportunity sets dramatically. For now, they faced a simple calculation: Were their expected returns higher investing their money at home or abroad (given the widespread perception of the illegitimacy of their wealth, there might well be attempts to confiscate it, or at least to tax it at "punitive" rates)? A robber-baron advocate might argue that the real problem was not with the oligarchs, but with the government: if only they could commit themselves to treat as legitimate wealth that which was begotten in these sometimes nefarious ways. But no democratic government can make commitments for its successors. It is through a process of social consensus that stability is maintained. It is hard to attain such social consensus if the process of privatization (and transition more generally) does not have at least a minimal degree of legitimacy.

Making matters worse, of course, was the fact that as each of the oligarchs decided not to risk reinvesting their funds in the country, the return was lowered for others doing so--and a powerful signal was provided to outsiders. Thus, while there might have been an economic equilibrium in which all invested in the economy, the equilibrium that was established entailed few doing so. Capital account convertibility was thus an essential ingredient in the failures--in establishing the overall incentive structures leading to the disastrous outcomes described earlier.(n22)

Alternative Theories of the Firm

Underlying this entire discussion are two quite contrasting theories of the firm.(n23) One, a mild modification of the traditional owner-manager theory, viewed shareholders as the true owners of the firm, and the other viewed stakeholders as having clearly defined "rights" of control and returns. Thus, "normally," bondholders and other lenders, workers, and local authorities have no real say in the actions of the firm. Bondholders and lenders have rights to certain flows of income. Workers have rights to be paid contractual wages. Local authorities have rights to be paid taxes due and to require conformity to general local zoning and other regulations.

But with these other potential claimants on the returns to the firm highly circumscribed, the residual claimant--with respect to both control and income--are the shareholders. The theory thus had a clear prediction: Shareholders maximize shareholder values, presumably subject to well-defined restrictions on the claims of other claimants, and in doing so, they maximized social efficiency. This shareholder-focused theory, when it is the clear residual holder of control and ownership, is usually called "shareholder primacy" or "shareholder sovereignty."

There is an alternative theory (set out in my 1985 paper), which I refer to as the multiple principal agent theory and which is sometimes called the "stakeholder theory." In this theory, there are many stakeholders of the firm. (It is perhaps worth observing that within the legal structure of many countries, the objective of the firm is not the maximization of share value, but a broader set of objectives which takes into account the interests of other stakeholders.) In this theory, the behavior of the firm is described as the outcome of a perhaps complicated bargaining process involving all the firm's stakeholders.

I believe this model provides a better description of the situation facing most economies in transition (and many advanced industrial economies as well). Among the important stakeholders that were ignored in the exclusively shareholder-focused theory were the local government units. Given the high (potential) levels of taxation and the role of local authorities in discretionary regulation (including zoning), local authorities have control and income rights, no less than do shareholders. Similarly, workers, if they can organize and are not subjected to legal restraints, can exercise effective control rights. Metaphorically, in such a situation, ownership and control claims can exceed 100 percent--or, perhaps more accurately, more than one party can exercise veto rights ("hold-up power") associated with any proposed plan and allocation of the firm's income. We can now better see the failure of the privatization model that was pushed in the immediate aftermath of the reforms. There was a naive belief in the shareholder theory. The ownership fights of the state would be transferred to a new owner, and that owner would then act to maximize the stock market value of the enterprise. In fact, the multiple principal agent model provided a better model of the firm. Transferring so-called ownership from the central authority to a private owner left the other "owners" intact. Having been left out of the new dispensation, the other stakeholders responded in noncooperative ways (e.g., predatory behavior from local officials and unproductive sullen workers) when, in fact, their full cooperation was needed to actually restructure the firms in the new environment. Many of the new shareholder-owners then just turned to looting--grabbing what they could while they could.(n24) I should, at this point, say a word about China, whose remarkable growth I referred to earlier. It did not solve perfectly the corporate governance problem (just as it would be hard to claim that the United States or any other Western government has provided a perfect resolution). The great innovation of China--the source of its new enterprises and much of its growth--was the township-village enterprises. China eliminated the layered agency problems that were the source of the problem in Czech privatization, and it avoided the corporate governance problems that were the source of the difficulty in Russia.

China demonstrated that one did not have to have a perfect legal structure, with property rights perfectly clarified, in order either to attract foreign capital or to induce domestic investment. the experience in China brings to the fore another issue too often forgotten in the rush to privatization the importance of competition. Standard competitive theory emphasizes that it is only with competition (and a high degree of competition at that) that markets, private property, and the profit incentive yield efficient outcomes. To be sure, if one could have both competition and private property, one should try for both. But the contrasting experiences of China and Russia suggest that, if one has to make a choice, competition may be more important than private property, especially the form of ersatz privatizations that actually occurred. It is competition that provides the driving force for greater efficiency and lower prices. But competition is also an important part of corporate governance: It is the absence of competition that creates rents that so often get diverted to inefficient uses.(n25) Bankruptcy Financial turmoil--from East Asia, to Russia, and other countries in transition--has focused attention on bankruptcy and the laws that govern it. Bankruptcy rightly belongs in any discussion of corporate governance for a simple reason: Bankruptcy is the means by which control rights get shifted. Bankruptcies and takeovers together can be viewed as the central ingredients in the market for management. And the rules governing bankruptcies and takeovers determine how effectively that market works--and therefore how effectively the market economy works.

Simple textbook models have been as misleading here as they have in the general theory of the firm. In the simple theory; the management team that is the most effective manager of a set of assets takes over the firm; the firm, in effect, is constantly on the auction block. Grossman and Hart (1980,1981) long ago showed the fallacy of that theory, in the context of firms with widely dispersed owners: Each shareholder, believing that a value-increasing takeover was about to occur, would retain ownership, becoming a free rider on the value-enhancing activities of the manager taking over the firm. Only if a shareholder believed that the person attempting to take over the firm would be successful in his bid, and upon success would strip the assets and reduce the market value, would he be willing to sell his shares. Thus, value-reducing takeovers would be successful. This is a dramatic example of the public-good nature of management to which I referred earlier. In fact, of course, managers of firms have within their discretion a wide range of actions that impede takeovers and increase their own bargaining power as a result, including increasing the asymmetries of information between insiders (the current management) and outsiders (not only shareholders, but those potentially taking over the firm) (Edlin and Stiglitz, 1995).

If takeover mechanisms worked perfectly, of course, bankruptcies would never occur or at least would be rare. Control, at each moment, would be exercised by those managers most able to maximize the value of those assets. If it turned out that the value of those assets (less the compensation required by the management team, which would be capped by the difference between the value of the firm under that management team and the next best management team) exceeded the claims on the assets, then those other claimants would be willing to take a write-down, in order to maximize their ongoing value--the write-down being better than what they would receive under some alternative management team. The frequency of these voluntary reorganizations and conversions of liabilities might depend on the volatility of the economy and the magnitudes of the uncertainties at the time the original contracts were entered upon; and one might have expected many such reorganizations in the aftermath of the initial privatizations. Indeed, this was one of the arguments proffered as a rationalization for an imperfect first round of privatizations. Once the market economy was in place, reorganizations of this form would occur, ultimately ensuring that each asset was "optimally" managed. But such arguments ignored the imperfections in the takeover mechanism, and it is these imperfections in turn that necessitate the use of bankruptcy procedures. Some thirty years ago, I turned my attention to this puzzle: Under the standard paradigm, there would be unanimity about which actions would be value-maximizing, and the kinds of conflicts that appear so frequently in court disputes would never arise. At one level, the answer was easy. The standard theory assumed a complete set of prices, including prices in each state of nature and at each date, and (implicitly) there was perfect information--in the sense that there was agreement about the profits (inputs, outputs, prices, markets, etc.) in each state (this could be thought of as part of the definition of the state of nature, or at least as a consequence of assumptions concerning rationality, understanding the nature of the equilibrium in that state). Given the obvious unreality of those assumptions, different market participants could have different judgments about the consequences of different actions, and therefore about what actions were appropriate--what actions maximized the stock market value (or the expected utility of each participant). The conditions for unanimity are highly restrictive.

Under those circumstances, existing "owners" might not be willing to surrender their control rights to creditors unless forced to do so. From their point of view, the present discounted value of the income flows to equity owners might be highly positive, even though the market (reflecting the value on those income streams by the marginal outsider) believed that creditor claims exceeded the market value of the firm. In such a situation, equity owners have to be "forced" to surrender their rights.

The traditional textbook model of bankruptcy describes a simple procedure: Creditors must have their claims fully satisfied, and equity owners get the residual. If there is no residual, then the creditors take a write-down. Those with collateral get the value of the collateral. Creditors effectively seize control, choosing the management team that maximizes their returns. They may choose to continue existing management, and even give the existing management an equity claim, if by doing so they will maximize their own return.

In fact, however, matters typically are not so simple. The interests and perceptions of the various creditors may differ--there does not exist unanimity among them, just as there does not exist unanimity among shareholders. But, more important, courts and the legal system present important obstacles in many countries. Given the possible conflicting interests, there need to be referees to ensure that various interests are protected (according to the law) and the priorities of claims (reflecting the various contractual arrangements) are honored. But while textbook models view courts as honest referees, they become, in many countries, especially those with weak judicial systems, active participants in the process. For instance, in many countries, the law recognizes the standing of other stakeholders, such as workers;(n26) the court may rule that, to protect their interests, current management must stay in place. Or courts may demand further evidence to weigh the merits of the various claims. Any actions that continue existing management in place--without appropriate oversight--are an invitation to asset stripping. The potential for corruption of the judicial system should be obvious. The problems become even more confusing when some of the debts are held in the hands of state-owned banks--the banks that lend money to political cronies in the first place. They have a double incentive--not to recognize the loss that would ensue upon bankruptcy; and to keep their political associates in power Moreover, the owners of the firm, or their friends, can buy some of the debts, and thus become creditors. These creditors clearly have incentives conflicting with those of the "true" creditors, but they can use their position to appeal to the courts and to block resolutions requiring unanimity or large supermajorities.

The courts themselves have played an ambiguous role in many countries. I use the term "ambiguous" deliberately ambiguously. The consequences of their actions--to delay transfer of control to creditors--is apparent. The fact that corrupt judges may have incentives to delay resolution is also clear: As long as the firm remains under trusteeship, the court-appointed trustee has a source of income, some of which may be diverted to the judge. Even if the court is not corrupt, but relies unduly on the advice of the trustee, the resolution may be delayed. But under the jurisprudence of many if not most countries, the court has an obligation to look after the interests not only of the creditors, but of other stakeholders, and delay may be consistent with this broader mandate. A quick resolution in terms favorable to the creditors might lead to a shutting down of the plant and high levels of unemployment.

In short, the economic system created in many of the countries in the beginning of the transition process was one in which incentives did play an important role--but they were incentives for asset stripping rather than wealth creation. These incentives were based on a misguided model of the theory of the firm, with residual control and income rights vested in the equity owner, where that owner was supposed to have guided the restructuring process in ways that should have led quickly to ever-increasing efficiency. A Coasian theory--in which as long as property rights were well defined, it mattered little how they were assigned (as long as one was solely concerning with efficiency)--provided the intellectual underpinnings  In fact, the more appropriate model of the firm was a multiple-principal agent model, in which privatization entailed the transfer of only some of the property rights. Bargaining under imperfect information does not, in general, lead to efficient outcomes. And while repeated interactions might have (in a more static environment) eventually allowed both a reduction in the imperfections of information and a gain in efficiency, the fast-changing scene meant that market participants acted much more myopically than is required for long-term efficiency.(n27) A destructive dynamic was put into place: The deteriorating confidence in the reforms led to less investment (or perhaps more accurately, a failure of investment to take off), in turn reinforcing the lack of confidence. The lack of confidence in the reforms undermined confidence in long-term property rights of those who had come by their wealth in ways that lacked social legitimacy. The banking system not only failed in its functions of monitoring; it was itself part of the problem. The failure of bankruptcy to protect creditors would, however, have put even well-functioning banks in a difficult position.

Social Capital

Decision-making in the owner-manager theory of the firm and its derivatives is simple: The owner-manager takes those actions that maximize market value. To be sure, in doing so he takes account of the impact of those decisions on others, for example, on the workers--through the wages that they will insist on in order to be willing to work for the firm. But these other parties are entirely passive. They are not actually part of the decision-making process. By contrast, in the multiple-principal-agent model (or the stakeholder model), decisions are made collectively, through an implicit or explicit bargaining process, with, in effect, various parties having veto powers. Workers, if they do not like what is being offered, can go on strike. The fact that there might be other workers willing to take the job at the offer may make little difference. The outcome of such bargaining processes may or may not be efficient. Some economies have, for instance, been characterized by frequent and long strikes, which both weaken the firm and hurt the worker.

In market economies, firms can be seen as local nonmarket solutions to collective action problems where transaction costs inhibit coordination by market contracts (Coase 1937). In the new postsocialist market economies, as in the established market economies, the primary example of extensive (i.e., beyond the family) social cooperation in daily life is indeed found in the workplace. More broadly, recent economic literature has stressed the extent to which economic relations are not mediated through impersonal markets, and the role of social capital in ensuring at least a modicum of efficiency in such relationships.(n28) A key aspect of such relationships is that they involve exchanges that do not occur at the same time and in which there are not enforceable contracts. Rather, one person does a "favor" for another, trusting that it will be returned at a later date. In this sense, such transactions are actually quite similar to many other market transactions--transactions that do not involve the contemporaneous exchange of two precisely defined commodities, about which there is nearly perfect information. When transactions do not occur contemporaneously, one party promises to pay (or deliver) some commodity to the other at a later date. When the properties of a commodity are not immediately transparent at the time of the transaction, the seller may make warranties concerning those properties; that again is a promise that has to be made good in the future.

There are two ways by which such contracts are enforced--through reputation mechanisms and through courts (sometimes referred to as implicit or explicit contracts). Both mechanisms are costly and imperfect. While the direct costs of legal enforcement are clear, implicit contracts, to be effective, require rents--a deviation of price from marginal cost--and this too has a social cost. The effectiveness of reputation mechanisms is attenuated when interest rates rise and when the probability of firm survival is decreased (or individual mobility is increased). The transition into a market economy is thus, in a sense, harder than the maintenance of a market economy, since initial investments in establishing reputation have to be made, at a time when reputational capital is particularly difficult to create. On the other hand, legal mechanisms were also weak in the economies in transition, particularly so since courts had no experience in being the arbiters of market economic relations. Worse still, given the immediate history of courts, there was probably little confidence in their impartiality (and even in Western countries, it has only been relatively recently, in historical terms, that they have gained a reputation for honesty).

The information requirements for, and transactions costs involved in, implicit and explicit contract enforcement are typically different, so that the two should best be thought of as complements rather than as substitutes. The problem in the economies in transition was that both enforcement mechanisms were weak: The state's legal and judicial capacities were limited, while the very process of transition--high institutional turnover, high shadow interest rates, and short time horizons--impairs the effectiveness of implicit contracts. Thus, even if new institutions did not need to be created, the very process of transition itself would have provided impediments to the workings of a market economy.

The weakening of social capital contributed to the problems posed by inadequate systems of corporate governance. Individuals undertook actions that, if they had been detected, would have led to the "loss of reputation" in a more stable environment, or jail sentences in an environment with better legal systems. Perhaps an alternative transitional strategy, paying more attention to the preservation of what social capital currently existed, and to the creation of new social capital, might have led to fewer abuses. We know little about how to preserve and create social capital. But this much seems clear. Breaking what is widely viewed as part of a social contract--such as not paying the elderly the pensions they believe they have earned--undermines social capital, especially if at the same time the government is transferring vast amounts of wealth to a few individuals. (Indeed, the destruction of social and organizational capital is one way of making sense of the seeming anomaly of reduced output, even as physical and human capital remains unchanged and supposed efficiency of resource allocation is increased.)(n29) It also seems to be the case that, once dissipated, social capital--like organizational capital--is hard to reassemble. Given that what social capital exists within a firm typically exists at the plant level, there is an argument that entrepreneurial efforts that arise out of or spin off from existing enterprises may be particularly effective in postsocialist societies in preserving "lumps": of social and organizational capital.(n30)

Political Dynamics 

In the discussion so far, we have seen how politics and economics are intimately intertwined. The court systems, local authorities, state banks--all of these played a role in the failures, exacerbating or failing to address the market failures associated with corporate governance. Economists did take into account politics in their analysis of the reform process--yet it was political considerations as much as economics that provided the rationale for the reform strategy advocated by several of the reformers.

Political dynamics are complicated. Reforms entail vested interests giving up their rents (what they may view as property rights). By definition, self-interested individuals never do so voluntarily. How, then, is reform possible? Sometimes, Pareto improvements can be found, grand bargains where all (or all major) groups are persuaded that they are better off.(n31) In some cases, realignments of political powers--and changes in political perceptions--make reform possible. Political coalitions themselves are altered in the process of reform. A recognition of this may itself serve as an impediment to reform, if existing power relationships worry that changes will lead to new coalitions, with uncertain long-run implications. But often, participants in the political process play chess, thinking through one or at most a few moves ahead. With a high enough discount rate, such myopia may even be rational. Those attempting reform may play on this myopia.

In the case of the economies in transition, there were large vested interests that had much to lose by the transition to a market economy. There was a belief by some reformers that the momentum of transition was strong enough to carry forward an agenda of privatization, over the interests of those who were acting as effective owners; but curiously, that momentum was not strong enough to overcome opposition to restructuring (Boycko et al. 1996). In this view, then, the desired (or only feasible) sequencing was: privatize, restructure, and then regulate. With privatization would come strong economic incentives for restructuring. Diseconomies of scope would lead to the disagglomerations (presumably, as it had done in the United States in the years following each merger wave), and, once restructuring had occurred, political pressures for competition and regulation would succeed. More broadly, the privatization would set in motion a process of legal reforms that would eventually lead to efficient systems of corporate governance. The architects of the Russian privatization were aware of the dangers of poor enforcement of property rights. Yet because of the emphasis on politics, the reformers predicted that institutions would follow private property rather than the other way around. (Shleifer and Vishny 1998, 11) There is, to my knowledge, no theory--and scarce historical evidence--underlying this optimistic assessment of institutional evolution. Indeed, there are general results showing that the equilibrium for institutions (including the interaction between markets and nonmarket institutions) are generally not Pareto efficient.(n32) Coase himself emphasized that transactions costs were important and that with transactions costs and imperfect information, outcomes of bargaining processes are not in general Pareto efficient. But latter-day "Coasians" have closed their eyes to these problems particularly in Russia, and have wagered that the secondary market redistribution of assets would work "well enough." It now seems that the wager was lost, and that the Russian workers and taxpayers will foot the bill.

To be fair, it is only a decade after the beginning of transition, and patience may be required. Perhaps eventually the story will develop in the way anticipated. Surely, those who advocated the reform strategies did not fully describe the vicissitudes through which the countries would go in reaching the ultimate stage. There is little evidence that they really anticipated the lack of investment in manufacturing, the flight of capital abroad, the drying up of support for market reforms. Had they focused their analysis on economics, they might have said in defense that these adverse outcomes came from political processes that were beyond the focus of their concerns. While they might have been rightly accused of ignoring politics, at least their defense would be intelligible. But when the strategy advocated was not based on economic analysis, but on political judgments, as it was such defenses seem less legitimate.

There is another political dynamic: Democratic processes recognize that it is concentrations of power, economic and political, that represent the strongest threat to a viable democratic society. Reformers might have used the commitment to democratic processes to motivate a process of restructuring and devolution, breaking the power of the industrial ministries. The commitment to privatization might itself have weakened resistance, since the ministries' powers would, in any case, have been short-lived. In the new-found spirit of freedom, new programs to support new enterprises might have found widespread support, and these new enterprises themselves would eventually begin to be a countervailing power to the old and established monopolies. Creating regulatory authorities over a monopoly (especially a natural monopoly) after privatization may be far harder than before. After privatization, there is a private interest seeking to maintain high prices and bar competition, and it can use its profits to further its interests. Before privatization, potential buyers do not know who will be the monopolist. There is a collective action problem--it is not in the interests of any one potential buyer to devote much energy and resources to resisting regulation. If the government is committed to privatizing, then there may be little interest in the ministry to resist such regulation--it will only affect the private sector that follows.

More broadly, I believe that once the political process opens up those within the countries to what is going on outside, questions will inevitably be raised: Why is our standard of living lower than other countries' with similar levels of education and natural resources? If market economies deliver higher growth with greater equity, the demand for market reforms will be irresistible. The problem is that those in the economies in transition look around, and those who are Bayesians--who are not totally committed to a view of the world unaffected by experience--see the dismal experiences of transition, and many are left skeptical. This is a reality that those committed to democratic market reforms will have to confront in the coming years.

Shibboleths, Confusions, and Misunderstandings

Looking over the experiences of the past decade, it is remarkable how confident statements were made about the appropriate paths for transition--and how quick some analysts were to come to the judgment that their predictions were correct. In fairness, while there was a wealth of experience across a multitude of countries, there was still insufficient data to come to firm conclusions: Did the countries that privatize and liberalize faster differ in other ways--differences that might themselves have accounted for their success? Presumably, the speed of liberalization was an endogenous variable, affected by history and prospects. And the countries differed markedly in both. For the countries of Eastern Europe, the prospect of speedy accession to the European Union provided both an incentive for quick liberalization and opportunities that were far different from those of the land-locked countries deep in Central Asia. The experiences provided a Rorschach test--with different observers seeing in the data what they wanted to see. It is only over time, as applying the same lens has led to contradictory results, that analysts have been forced to face up to the facts and the ambiguities that they present.

Among these ambiguities are those concerning what is meant by privatization itself. When the United Kingdom sold the shares of one of its nationalized companies to the public, it was fairly clear that this was a privatization. But when Poland sold shares in government-controlled national investment funds on the Warsaw stock market, did that mean that all the industrial companies in the funds' portfolios had been "privatized" (as was argued for many years)?(n33) Or are the portfolio companies only privatized when they are sold off to strategic buyers or are themselves floated on the stock market (as is now being claimed)? The answer is important for an empirical analysis, because how we answer that question determines whether we view Poland as a fast or a slow privatizer. Or consider another example: If the Czech Republic uses voucher privatization, but then has state banks provide soft loans to the privatized companies, is that real privatization?(n34) Privatization was supposed to provide hard budget constraints--yet the soft loans ensure that there are soft budget constraints. But this example presents the analyst with even deeper problems: How do we know that the loans are soft? Presumably repeated bailouts of the banks provides some evidence ex post. But then even so-called privae banks may provide soft loans, particularly if they are undercapitalized, are gambling on redemption, and are counting on the past history of bailouts.

In a sense, as we go forward, as important as the answers are to the econometric findings, these are semantics. The lessons are the same: Corporate governance matters, and prospects for privatizations that fail to adequately address corporate governance issues (whether because of failures in banking or problems on the equity side) are bleak.

The concept of restructuring presents similar ambiguities. One of the early debates was whether the government should make an attempt at restructuring or leave that to the new owner after privatization. The "first wave of wisdom" was the theory that restructuring should be left to the new owners--after all, the state had not previously shown itself very adept at restructuring. The problem with this theory was that the enterprises were seriously overstaffed with huge agglomerations of outdated and mismatched assets. The governments usually did not want to undertake mass layoffs, and would rather pass that sensitive problem on to a foreign buyer who could subsidize the workers out of their "deeper pockets" or who could take the political heat for the mass layoffs. However, the foreign buyers hardly wanted to enter the country and take the recriminations for such politically sensitive acts (which repaired past mistakes not of their making).(n35)

Another problem in the "first wave of wisdom" was that the balance sheets were stuffed with useless assets and with "liabilities" resulting from subsidies directed to the firm in the past. As soon as a whiff of foreign money was in the air, all those very soft loan liabilities suddenly hardened into "legal obligations" of the enterprise to the state banks. The new buyer would have to, of course, accept the "legal obligations" of the firm being purchased. On the asset side of the balance sheet, the assets useless to the buyer's business plan usually had no secondary market but were valued by the selling privatization agency either at cost or at the imagined "earning potential." Need I mention cleaning up for past environmental neglect (or hidden liabilities of the firm for the effects that emerge in the future)? Leaving aside labor problems, if the buyers accepted all the liabilities (explicit as well as hidden) and useless assets of the unrestructured firms, then most deals would have been sold for a symbolic $1. But that was politically unfeasible,(n36) so the "first wave of wisdom" generally led to the dead end of very little foreign direct investment in existing enterprises. Restructuring theories eventually became more nuanced. It was recognized that foreign buyers did not want to be saddled with the politically sensitive tasks of cleaning up for past mistakes. It was recognized that investors wanted a "pure play" consisting of the assets that fit into their business plans, not the whole bundle of assets that seem to have been randomly agglomerated in the past. It was recognized that the very uncertainty created by the mixed bag of assets and liabilities decreased the price that buyers were willing to pay. If they could not get a pure play, they might just as well go to greenfield investment, leaving the old enterprise to the junkyard (except perhaps for a few assets purchased in bankruptcy auction). It was recognized that the government had to clean up the soft loan and environmental problems on its own and could not expect foreign buyers to come in and bail everyone out. Now governments are taking more responsibility for what is called "passive" restructuring to clean up the balance sheets and reduce the labor problems (e.g., with active labor market policies and improved social safety nets)--all still without the "active" restructuring decisions about product lines and new machinery that should be left to the buyers.


There is an old saying that one never crosses the same stream twice, and while there is much to learn from history, and from trying to understand why things turned out differently from the way many expected, our focus needs to be on the future. Two facts stand out at this juncture: Incomes and standards of living have, at best, stagnated in most countries, if not fallen; popular confidence in market processes and the optimism of transition have been eroded in many if not most of these countries. And any strategy going forward must deal with these facts.

In many countries, a strict interpretation of the rules of a market economy would lead quickly to government takeover of large proportions of existing assets that have been privatized in the past decade. In some countries, there are huge tax arrears. In a typical advanced industrial economy, arrears of these magnitudes would be swiftly dealt with, with government attachment of the property, to be eventually sold, with the proceeds used to pay the arrears. Moreover, in many of the transition countries, the banks are effectively bankrupt, in part because of huge arrears on the part of corporations to the banks. Again, the standard market process would be a foreclosure by the banks and government takeover of the banks, as they recapitalize them (unless the banks can find private sources of recapitalization, doubtful in most of these countries, without huge government subsidies). In a way this is good news, because there is now a second chance for government to go about the reprivatization of these assets in ways that take greater recognition of the problems of corporate governance and the other issues I have discussed.

But as the process of democratic transition continues, due attention also will have to be paid to other issues that were, perhaps, given short shrift. As important as restructuring existing assets and enterprises is the creation of new ones; job creation and entrepreneurship need to be put at the center of discussions. So too must stress be placed on the other aspects of the institutional infrastructure for a market economy and a democratic society, from financial institutions to social and legal structures. The high and growing levels of corruption and perceived problems of governance are both cause and consequence of the failure of the economic transition.

The last decade of this century has been full of surprises. Few anticipated the precipitous collapse of the communist governments. Having seen that collapse, there was a euphoria about the transition to a market economy. A decade later, those hopes have, in most cases, been dashed, replaced by a more somber realism about the difficulty of the task ahead. Yet I remain optimistic: Hopefully the lessons that we can extract from the experiment, which already has extracted such a huge human toll, will stand the countries in good stead as they approach this new, and hopefully more successful phase in their democratic transition.

(*) Juvenal (A.D. 60-130), Satires vi, 347. "Who is to guard the guards themselves?"


(n1.) See World Bank (1997). These statistics are not quite "fair"--in principle, we should treat large provinces in other countries as independent data points. Were we to do so, however, the picture that would emerge would not be much different.

(n2.) See World Bank (1996).

(n3.) It is important to remember that, while this may not be a statistically significant regression result, it nevertheless serves to cast serious doubt on assertions to the contrary, namely that early liberalization and low inflation were the cornerstones of a successful transition policy.

(n4.) For instance, poverty data for Russia come from an independent survey. The eighth round of the Russia Longitudinal Monitoring Survey, conducted in late 1998, showed that the number of people living below the poverty threshold had increased from 36 percent in 1996 to 39 percent in 1998, and the number of children less than six years old living below the poverty line had increased sharply from 45 percent to a staggering 56 percent.

(n5.) Estimates suggest that, within Russia, over 70 percent of transactions (by value) are now barter. There is no agreement about the reason for the growth of barter, but several of the favored explanations focus on policies striving for macrostabilization (e.g., tax policies that use the financial system for tax collection; or cash-flow constraints unmatched by expenditure constraints, which lead to arrears).

(n6.) See Akerlof and Yellen (1985).

(n7.) See Stiglitz (1999).

(n8.) See Gaddy and Ickes (1998) for a discussion of some of these issues.

(n9.) See Edlin and Stiglitz (1995) and Shleifer and Vishny (1998).

(n10.) See, for instance, Stiglitz (1982).

(n11.) In fact, the conditions under which even firm-value maximization leads to Pareto efficiency are highly restrictive. See Grossman and Stiglitz (1977).

(n12.) Note the analogy between this solution to the public goods problem and that discussed in the public goods literature concerning lighthouses: Coase (1988).

(n13.) I first began worrying about this problem in Stiglitz 1972.  (n14.) See Dyck (1999).

(n15.) This was an idea first put forward by Berle and Means (1932), and recast, in terms of modern agency theory, by Stiglitz (1985).

(n16.) For a discussion of the role of threats of termination in credit markets in addressing problems of agency, see Stiglitz and Weiss (1983). Indeed, the advantages of the greater "control" associated with short-term debt may more than offset the disadvantages associated with the greater risk it imposes (on the borrower). See, for example, Rey and Stiglitz (1993).

(n17.) Or, more accurately, the total expected repayment, which includes the size of the residual if the firm does go into bankruptcy.

(n18.) See, for instance, Stiglitz (1992).

(n19.) See Weiss and Nikitin 1998 for econometric analysis.

(n20.) Even if corruption had not played an important role in the process, the heavy indebtedness (with implicit or explicit limited liability) would have distorted behavior. See Stiglitz and Weiss (1983).

(n21.) Some advocates of rapid reform, while recognizing this, argued that the robber barons would have an incentive to sell their assets (in a competitive auction) to those who would best deploy the resources. The presumption was that designing such an auction was simple enough--and the robber barons had every incentive to maximize value. This did not occur, in part because they may have focused their attention on other issues--such as how they could strip value out of the firms they controlled rather than enhance market value--and in part because the very weaknesses that underlay the first round privatization confronted the second round: Who had the wealth to purchase the assets? Unless foreigners could be interested, there was no presumption that those who might have purchased in a second round (which didnot occur)--those who had favored access to "bank" loans--would have been much more efficient than those who got the assets in the first round.

(n22.) To be sure, these disadvantages of capital account convertibility had to be offset against other concerns: the distorted incentives associated with attempts to circumvent the restrictions on convertibility. In retrospect, it is hard to believe that the distortions arising from these more traditional concerns outweighed the adverse welfare effects arising from the distortions just discussed. Capital account convertibility may also be related to the difference between corruption in Russia and in China.

(n23.) For an analysis parallel to that presented here, with an application to the problems in the Czech Republic, see Mejstrik (1999).

(n24.) Under the older shareholder "residual theory," shareholders cannot, by definition, loot the firm: They would only be taking assets from themselves. To be sure, they might decide to decapitalize the firm, if they believe that the return to investing in the firm is sufficiently low--and that may have accounted for some of the observed behavior. (That leaves unanswered the question of why the return to capital was so low, seemingly everywhere--not every sector can have a comparative disadvantage!) But under the multiple principal agent model, there are multiple claimants; the shareholders try to grab what they can before other claimants do. Looting behavior, in that circumstance, may be individually rational, even if collectively irrational. The behavior may have made particular sense given the imperfections in domestic capital markets. The dominant shareholders were often older men a few years from retirement. Suppose they plow back their earnings, restructure the firm, and build up the stockholders' value. There is little or no tradition of family ownership so they cannot pass the firm to their children. There is little or no private secondary market for firms, and even less a market for public new issues. Anticipating no other exit, the dominant shareholder may turn to a "loot & liquefy" strategy to convert the big ownership block into funds for retirement elsewhere.

(n25.) See Commander, Dutz, and Stem (1999).

(n26.) In a sense, in many other countries, including Germany and Japan, the law "adopted" the model of the firm that we have called the multiple-principle agent or multiple stakeholder model.

(n27.) See Farrell (1987). These problems were exacerbated by the exit option offered by capital account convertibility, discussed earlier.

(n28.) For a review of the extensive literature on social capital, see Woolcock (1999 forthcoming).

(n29.) See Blanchard and Kremer (1997) and Gaddy and Ickes (1998).

(n30.) See the township-village enterprises in China as in Weitzman and Xu (1994), Lin et al. (1996), and Qian (1999 forthcoming). Other social organizations that might incubate and support entrepreneurial efforts include local township governments, unions, schools, colleges, cooperatives (housing, consumer, credit, and producer coops), mutual aid associations, guilds, professional associations, churches, veterans' associations, clubs, and extended family groups.

(n31.) Given the magnitude of the (alleged and seeming) inefficiencies in the Soviet system, one might have thought that there was sufficient surplus to be divided by the movement to new organizational forms that everyone could have easily been made better off. In that sense, one might have thought that the political obstacles to reform would have been smaller than in situations where the gains were smaller. There are several possible answers to this conundrum. The change in political climate may have made it difficult to compensate the old guard; but then, how was it that so many of them remained in positions of power, to continue to block reforms? Perhaps the problem was that there was no way that binding commitments could be made. The reforms would set in motion a set of changes that would eventually deprive the old guard of its power. They knew it, and thus they had to block the reforms at the onset.

(n32.) See Arnott and Stiglitz (1991).

(n33.) Surely our answer to that question should not be based on the government-controlled funds contracting out management to private Western investment banks-since that was done before the share flotation.

(n34.) Often the "privatized" companies were substantially owned by widely held voucher investment funds that were completely controlled by bank-owned fund management companies. Thus the banks were bailing out the companies indirectly under their own control--not exactly arm's-length due diligence.

(n35.) One anecdote is illustrative. A foreign furniture company bought a plant in a southeast European country and told the privatization agency as well as the workers that one part of the enterprise was dubious. If it could not fix it within a year's time to make a profit, it would have to shut it down. After the year, that unit had not shown much hope, so the foreign managers announced that it would be shut down. Then the police in the local town where the employment was critical occupied the plant and informed the foreign managers that they had "broken the law" and would have to leave the country or face arrest. The foreign company then had to shut down the whole operation and demand that the national government intervene to remedy the situation. That foreign buyer probably does not support the "leave restructuring to the buyer" theory.

(n36.) In fact, the political opposition in many postsocialist countries has implicitly adopted a valuation theory that says, "Any price that a foreign buyer had actually paid was obviously too low, so the government may always be accused of selling out the national assets cheaply to foreigners."

By Joseph Stiglitz

Copyright of Challenge is the property of M.E. Sharpe Inc. and its content may not be copied without the copyright holder's express written permission except for the print or download capabilities of the retrieval software used for access. This content is intended solely for the use of the individual user.

Source: Challenge, Nov/Dec99, Vol. 42 Issue 6, p26, 42p.

Item Number: 2554350


Esmu rasktījis "Challenge" žurnāla izdevējam M.E.Sharpe pēc atļaujas šo rakstu šeit darīt pieejamu. Vēl neesmu atbildi saņēmis, bet ceru, ka, tā kā to šeit lieku tikai izglītības veicināšanas nolūkos, neviens par to neļaunosies.

Juris Žagariņš

Kas jauns Latvijā?