MAAcif 14t5 COMMFNer ON "EsmrusrG PRopnr RIGHTS," BMR ShFER Olivier Jean BIlanchard E arty privatization plans (circa 1990) were profoundly naive. They implicitly assumed that state firms belonged to the state and so could be disposed of as the state wishedL Had the wisdom of Shleifer's article been absorbed earlier, less time would have been wasted before reality sank in. Any privatization plan-actually any substantial reform-will do well to follow the steps articulated in Shleifer's artide. Assess the initial distribution of control rights. Assess the implications of the reform for the new distribution of rights. And last, but hardly least, assess the political feasibility of the reform to determine how the initial holders of control rights can be induced to accept the redistribution of these rights. As the example of privatization in Russia shows, there may be more room for politically feasible reform and thus for redistribution than is commonly believed, even in countries with a weak central authority. The reason, emphasized in the arti- cle, is precisely why reform is desirable in the first place: the initial structure of con- trol rights might be very inefficient. And the scope for Pareto improvement is exactly what generates the scope for buying off some of the initial stakeholders. To take an example developed by Shleifer, the politician who cannot credibly commit not to expropriate later has control rights that are worth little under the initial struc- ture of rights. He can thus be bought off cheaply. My comments cover three issues. First, I apply the logic of Shleifer's article to privatization in Poland. Even though the initial structure of control rights is dif- ferent from that in Russia, Shleifer's approach proves serviceable for Poland as welL Second, I look at control rights within firms, an issue that strikes me as potentially more important in many Eastern European countries than the removal of control rights from politicians. Third, I argue that a frontal assault on polid- cians, on which the artide focuses, is but one of several reforms (not all of them explicitly about control rights) that can achieve a similar goal. As an example I explore the interactions of the safety net, the inflation tax, and control rights in Russia. Olivier Jean Blanchard is professor of economic at the Massachusctts Institute of Technology. Procdngs of the World Bank Annual Confnence on Devopnme Economics 1994 118 01995 The International Bank for Reconstruction and Development / TWE WORM BNK Blanchard 119 Control Rights and Privatization in Poland In Poland ministries and local authorities had few control rights from the star. Those were held by insiders, managers, and especially workers. While the menu of privatization options is long and complex, it can, with litde loss of realism, be reduced to two main options. One is that insiders buy their firms at a discount (known as privatization through liquidation). This method legalizes their control rights, extending their explicit stake further into the future, reducing the inefficiency of the structure of control rights. But insider control also makes it harder to get outside finance and thus limits the scope for restructuring. Another option is to sell the firm to outsiders, known as privatization. While the insiders get shares at a discount, they in effect give up their control rights to out- siders, who are in a position to get outside finance and thus to restructure. Four years into the privatization process in Poland, the results are very much what would be predicted from applying the logic of Shleifer's article. A simpie for- malization mav help. Let X be the value of the firm if sold to insiders, and let Y be the value if sold to outsiders; Y is higher than X because outsiders, having access to outside finance and expertse, can more easily restructure. Let B be the benefits of control for insiders. If Y > X + B, restructuring is worth enough to insiders to induce them to relinquish control rights in exchange for a share in the success of the firm. If Y < X + B, insiders choose the insider privatization route and forgo restructuring. The evidence in Poland fits this model welL Firms for which restructuring holds the key to better prospects have chosen outsider privatization. But so far only 100 of the 8,000 state firms have chosen this route. Some 1,000 others have chosen insider privatization and are doing litde restructuring beyond shedding enough labor to stay in business. The remaining firms have not moved yet; even if many of them move to insider privatization, not much restructuring is likely to result On Control Rights within Firms Shleifer focuses on the renoval of control rights from politicans. That may indeed be the major issue in Russia. In the countries of Eastern and Central Europe, a big- ger issue is the structure of control rights within firms. The implications of workers' control and ownership are old topics of economic research. Researchers point out the importance of the transferability of control rights and the problems implied by lack of access to outside capital and by the short horizon of workers. From my perception of the problems of firms in Eastern and Central Europe, whether formally state owned or privatized through insider priva- tization, I find that this research has largely missed the main issues. Here again, the focus on physical control rights and on the efficiencr of the con- trol rights structure helps considerably. The structure is so complex that, to use Coasian terminology, transaction costs lead to enormous inefficiencies and often to paralysis. 120 Comnvrt on .stablishing Propeny Rights" We know that even if a firm is owned by iLs workers and if transfers between workers are costless, the firm should still choose the efficient level of emoloyment- the level at which the marginal product of labor is equal to the reservation wage. Once this level is chosen, the total product of labor is then distributed between those who keep their jobs and those who do not. When transfers have a cost, however, the outcome will be a larger, inefficient level of employment, possibly to the point where the average product of labor is equal to the reservation wage. In Eastern and Central Europe transfers-think of them as severance payments to those who lose their jobs-are indeed difficult to implement. They have the dimension of present discounted values, and few firms have the cash flow to pay them or the ability to borrow from banks for that purpose. Tiu's employment is still substantially too high in most firms. This presentation only scratcher the surface, however. The more serious issues involve the time spent building and overthrowing coalitions within firms. Consider a firm that has to close one of its two plants to remain in business. In a Western firm we can simply assume that the manager makes the decision. In a worker-controlled firm the outcome is the result of bargaining between workers in the two plants. Their relative influence depends on their threat points, the damage they can inflict, the side supported by those who control services common to the two plants, and so on. The eventual outcome may turn out to be efficient, but the bargaining takes place in real time, so it is likely to involve delays, changes in decisions, and output losses along the way. These are far from just theoretical possibilities. We all know of firms with plants that should be shut down, of plants where changing coalitions have led to rapid turnover of managers, of cases where a deal made by the manager with an outside investor is then rejected by another coalition within the firm. Dabrowski and others (1993) present an array of such stories. If we are to make progress in understanding the behavior of firms, whether formally state owned or privatized through insider privatization, and the dynamics of restructuring, these are the issues that we need to think about. Frontal Attadcks In thinking about how to change control rights, Shleifer considers only frontal attacks-policies explictly designed to change the structure of control rights. He fails to mention that many macroeconomic and microeconomic policies also can be used. By changing the value of control rights, such measures can change coalitions and achieve many of the same results. Consider a simple and topical example: the effects of introducing a safety net in Russia. The disclaimer that the assumptions are just short of outrageous and that this is no more than a conceptual toy obviously applies. Suppose that firms in Russia are composed of managers and workers. For sim- plicity, assume that they produce nothing and should be dosed. By keeping the firms alive, however, managers and workers can extract X in subsidies from the state. Of Blanchard 121 this, aX goes to managers, who are needed because they know how to extract the subsidies, and the rest, (1 - a) X, goes to workers. If a firm is closed, the managers find themselves with zero, and workers receive unemployment benefits B. Both X and B are financed by the inflation tax, levied on the popuiation as a whole. For given X and B the equilibrium is clear. If B c (1 - a)X workers stay in coali- tions with managers to extract the state subsidies. The cost to the budget is X If B > (1 - a)X workers prefer to receive uncrmployment benefits and so drop out of the coalitions, and managers are left in the cold. Firms close, which is good. The cost to the budget drops from X to B, which can imply a substantial drop if a was large enough. Thus, increasing the safety net is good for efficiency (apart from any distribu- tional effect it may have, which is not captured here). In terms of the framework of Shleifer's artide, a "divide and conquer" strategy replaces the frontal attack. Workers are bribed to drop out of the coalitions with managers. This policy reduces the cost to the budget and the inflation tax. The logic of this model, and I believe what it captures of reality, implies that man- agers should oppose the safety net. Can they? I am not sure how to make the argu- ment tight, but I sense that it might be more difficult for them to launch an all-out attack on the safety net than, say, on the modalities of the privatization program. This model can be extended in many ways. B, for example, depends on more than unemployment benefits. Prospects for reemployment are also important. Thus, the better are the reemployment and wage prospects in the economy, the easier it may be to destroy coalitions in unproductive firms. Growth in part of the economy makes it easier to reduce subsidies in other parts. It mnust be clear that I had fun reading the article. I hope that my comments show how the approach can be extended to analyze the issues of transition. Reference Dabrowski, Jr., M. Federowicz, T. Kaminski, and J. Szomburg. 1993. Privadzadon of Polish Stare- Owned Enterprises: Progress, Barriers, Initial Effects." Working Paper 33. Gdansk Institute for Market Economics. COMMEwr ON "ESrABLISHING PRoPERTY RiGirrs," BY SHLEFER Roman Frydman Shleifer's arside is an important contribution to the rapidly growing literature reexamining the functions and determinants of property rights in the transi- don economies. Because his analysis is predicated on the belief that establish- ing property rights in the postcommunist economies primarily means reducing the detrimental effects of bureaucratic control, it incorporates strategies for reducing the inefficiencies of political control of economic assets. The first set of reform strategies would leave control rights with bureaucrats and rely on changes in the costs and benefits of political control to reduce the incentives for politicians to allocate resources inefficiently. Institutionalizing corruption, encouraging nomenkdatura privatization, reforming the state bureaucracy, and lim- iting the abuse of political control rights by strengthening democratic institutions are ways of altering incentives. But as Shleifer shows, none of these approaches is likely to be effective in reducing the inefficiency of pervasive political control over assets in transition economies. Shleifer turns next to the more radical strategy of legally reassigning control ights from state bureaucrats to other parties. Mass pnvatization programs are the leading example of this strategy. Shleifer recognizes that transferring control rights from politicians to private agents does not fully establish property rights. The trans- fer is just the first step; it does not ensure the security of those rights. Even after the transfer politicians can use the resources of the state treasury to persuade rather than to order firms to pursue political objectives. Shleifer's artide presents only a preliminary analysis of these postprivatization obstades to establishing private prop- erty rights. Here I will briefly discuss two points that require more attention in the analysis of the establishment of property rights. Shleifer points out that governments in most transition economies are too weak to provide reliable enforcement of contracts and that their political processes do not allocate adequate resources to effective contract enforcement mechanisms. Shleifer further argues that a mass privatization program is likely to create a powerful con- Roman Frydman is professor of economics at New York Unierrsity. The author rbanks AndricJ Rapaczynski for insightful comments on an earlier version of this piece and for many stimulating discs- sions in connection with their work on a relatcd set of issues Pwceaedgs of the Wortd Bank Annai Confeence on Developmet Economics 1994 122 01995 The International Bank for Reconstrnction and Development J THE WORlD MANIC Fjdmaan 123 stituency among beneficiaries of the program, who will demand that politicians allo- cate substantially more resources to the protection of their newly acquired property rights. But even with much larger expenditures committed to the task, the protec- tion of property rights by the weak governments in the region may turn out to be more complicated than Shleifer's article suggests.1 And complex forms of property, which are ubiquitous in a market economy, require even more complex forms of protection than simple contracts. They require a panoply of complicated corporate and securities laws. And such legal protection is just a beginning. In the market economies important property rights are often fur- ther protected by a self-enforcing structure of incentives embedded in the market system. As an illustration consider that the most important aspects of minority own- ership, such as the level of dividend income, are only minimally protected by law. Much more effective enforcement comes from the discipline imposed by capital markets and the potential future capital needs of the corporation (Frydman and Rapaczynski 1994). Moreover, even if the legal protection of control rights were sufficient to establish property rights after privatization, politics might stand in the way of a genuine market economy driven by price sig,nals rather than by political decisions about the allocation of resources. And that leads to my second main comment. Shleifer focuses almost exclusively on preprivatization politics as a key determinant of the political feasibility of priva- tization. Since reformers are often politically weak, they have to buy support for the program by induding preferences for politically powerful groups. This belief has underpinned the Russian mass privatization program, which in effect has transferred control rights to insiders in the majority of Russian enterprises (Boycko, Shleifer, and Vishny 1993; Pistor 1993). It is too early to tell whether investment by outsiders (individuals not previously employed by the privatized enterprises) will follow under the Russian conditions. As important as preprivatization politics may be, it is clear that the dangers of the con- tinuous politicization of resource allocation decisions and the need for continuous prorection of control rights do not necessarly end when a mass privatization pro- gram is completed (Frydman and Rapaczynski 1994). Slileifer's artide does not really deal with the politics created by the allocation of control rights following pri- vatization. This reallocation modifies the demand for state intervention by the newly enfranchised private actors and therefore leaves open the possibility that the demand for state intervention might not diminish, at least in the short run. Note 1. Shlcifer occasionally suggests that the control rights transferred throngh a mass privatization pro- gram arc physical rights. Thcse rights arc implicitly considered as less complex than contracts and thus casicr to protect. It should be noted, howvcer, that the concept of physical rights is at best applicable to owner-nanaged firms in which the owners are physically present to cxercis their control rights over the assets lcgally designated as theirs. Thc notion of physicau rights is certainly not sufficient for more com- plac forms of propcrty, in partidcular thosc pertaining to ownership of corporations and other securitizcd, intangible forms of property that usualy result from the privatization of industrial firms in the transition economies. 124 Comment on "Establishing Propeny Rights" References Boycko, Maxim, Andrei Shlcifer, and Robert W. Vishny. 1993. 'Privatizing Russia." Erookings Papeo on Economic Activity. Washington, D.C.: The Brookings Institution. Frydman, Roman, and Andrzej Rapaczynski. 1994. Privatiwtion in Eastern Europe: Is the State lVithering Away? Prague: Central European Univcrsity Press, in association with Oxford University Press. Pistor, Katharina. 1993. "Privatization and Cornorate Governance in Russia: An Empirical Study." Paper prepared for the Workshop on Economic Reform, Centcr for International Security and Arms Control, Stanford University, Stanford, CaliE FLOOR DISCUSSION OF "ESTABLISHING PROPERTY RiGHTS," BY SHLEIFER In an extension of the question Roman Frydman (discussant) posed in his com- ments, a participant from Harvard University asked why the de jure reassign- ment of property rights would affect the de facto distribution of control rights in the way Shleifer assumed. Even Western systems with well-defined private prop- erty rights still see high levels of bureaucratic intervention, he said. Perhaps the political varizbles that control bureaucracies are more important than the complete definition of property rights. Assuming the concrete example of a firm that is under the control of some min- istry, Shleifer said that he was talking about mechanisms for depriving the ministry of the opportunity or ability to tel the firm's manager what to do, because that would result in more efficient resource allocation. Does that mean we have arrived at an efficient structure for control rights? Definitely not, Shleifer said. Full man- agerial control is not ideal, either, and shouid probably be complemented by the rights of majority and minority shareholders, for example. In other words, said Shleifer, by refusing to protect the rights of ministries- which is essentially what privatization and corporatization accomplishes in these countries-you begin to reduce political influence on these firms and increase oper- ating efficiency. But laws allowving shareholders and other participants in a market economy to exercise their rights do not come about because Western advisers tell transition governments to write them down. Such laws come about because of inter- nal lobbies and political pressures to create them. The people pushing for securities laws and corporate laws in Russia are the ones who have a financial interest in mak- ing sure such laws exist-the new shareholders in privatized companies. The impe- tus for enforcement mechanisms is created precisely through this reallocation of control rights. So what Olivier Blanchard (discussant) calls the frontal assault is the right first step; enforcement strategies and other mechanisms will follow. Frydman amplified one point he had only touched on in his con.nent: yes, one must make compromises at the outset to push a policy through, but it may also be necessary to institute new policies later if the original policiec result in inefficient This session was chaired by S. Shahid Husain, vice presiden; Managenent and Personnel Services, at the World Bank. Proceedngs of the World Bank Annual Conference an Development Economics 1994 01995 The International Bank for Reconstmction and Development / llE WaORuM MMic 125 126 Floor Discussion of fEsablishing Propertyr Rbgb-s control structures. One should not assume that once a privatization program is in place it will do only gond; examples to the contrary abound. Romania's so-called privatization program, for example, was actually the biggest nationalization pro- gram in history: in the end 70 percent of industry was in state hands. A joint pro- ject between Bulgaria and the World Bank will basically ensure that bureaucrats control the so-called government-sponsored funds. These were reassignments of control rights, but whether they were efficient is clearly open to question. A participant from the United Nations asked Shleifer to define the bureaucrat his paper was targeting. Is he talking about the civil service, the ministries, the party cells, or even the workers and managers of state-owned enterprises? It seemed to this participant that Shleifer was saying that property rights originally belonged to the state (or society or the Commnunist party) but over time accrued, not legally but de facto, to various elements of society, from worker councils up to civil servants or ministerial bureaucrats who had the say over what was sold, at what price, and so on. So the bureaucat in this view was someone who possessed control rights or negotiated contracts that were not enforceable because the rights were never trans- ferred legally. Which brings us, the participant said, to what Frydman was talking about: how do you make this enforceable when you do transfer the rights in some form? Blanchard had said that the French civil service might be a model, the partic- ipant continued, and Shleifer had responded that you cannot reform the bureau- cracy, which might well be the case in Russia. But if you cannot reform the bureaucracy, how are you ever going to enforce property rights in law? How are you going to defend those who have residual control rights? The bureaucrats are certainly not the workers or managers, Shleifer replied. In many of these countries control rights over industrial enterprises tend to lie in the ministries; in smaller enterprises they lie in local governments and admcinistrations. How do you enforce property rights once you transfer them from these bureaucrats? Shleifer said he was still struggling to understand why the point is so controversial, but fundamentally it is difficult to enforce the bureaucrat's rights if the state machine is not supporting the bureaucrats. Once you disengage the bureaucrats from control over an enterprise's assets, it is not nearly as difficult to enforce at least the physical rights of the enterprise managers. And having control rights in the hands of managers instead of bureaucrats leads to more efficient resource allocation because the man- agers have longer time horizons than the bureaucrats and greater access to owners and to the firms' cash flow. Who will challenge the managers? In some countries, the workers; in others, the shareholders, other managers, or thieves. Some protection is called for, but managers are probablv in a better position than bureaucrats to protect control rights from all parties. That is why, Shleifer continued, the first objective of reform should be to efininate the bureaucrats' control rights. The second question raised by the U.N. partcipant-if you cannot reform the bureaucracy, how can you hope to protect property rights?-is the central issue of eco- nomic development, said Shleiier. In some ways you need the government to protect property rights, says Shleifer, so why is he arguing against the government so strenu- ously? The problem in many Eastern European countries and in transition economies Floor Discussion of "Establishing Property Rights" 127 is that the governments do precisely the wrong things: they don't enforce contracts and protect propertv rights and most of their efforts are directed at grabbing addi- tional control rights for themselves. So what is his position on reforming bureaucracy? The East Asian strategy was a bureaucracy-led transformation. Is this a viable alterna- tive for the transition economies? Given what Shleifer knows about Eastern European bureaucracies, that is an insane idea. But the participant is absolutely right, he con- cluded. At some point civilized government is needed if you want a civilized economy. But reforming the bureaucracy is not the immediate strategy for economic reform. A participant from the World Bank said that he found Blanchard's explanation of the need for coalitions compelling and that he had never before heard articulated how workers could be involved in the restructuring of firms. But he wondered if there vwasn't some inconsistency between trying to create consensus and trying to achieve efficiency. If the problem with workers is that too many decisionmakers cre- ate inefficienqc, isn't there an inherent contradiction between building coalitions and achieving efficiency? Hasn't the Western experience been that you get efficiency gains only when a minority exercises authority and corntrols a firm's assets? You don't want to look at what happens the day after privatization, responded Blanchard; you want to think about what will happen over time. Once a firm has taken the liquidation route to privatization in Poland, for example, there is a period when workers, managers, and ownership are extremely decentralized-when nobody really knows how to play the game. For a while the exact control rights of each worker or plant are not well defined, and you see coalitions conming and going and little positive outcome in terms of restructuring. Over time, however, a gifted leader emerges and puts a coalition together, workers basically give up their control rights in exchange for good decisions, and the result is a governance structure involving a small number of players. l's a strange process, and one that is likely to take a long time and to vary from firm to firm. A crucial point in Blanchard's model, said a participant from the International Finance Corporation, was comparing unemployment benefits with what workers could extract from the state. Did Shleifer agree or disagree that credible cash con- straints might be more important dtn any assignment of property rights? And does the nature of the firm make a difference? Poland has mainly manufacturing firms, for example, whereas many of the state firms in Russia andT Ukraine are involved with oil, gas, and other mineral resources. It may be more difficult to enforce cred- ible cash constraints on the oil and gas firms, and the returns on corruption are much higher in those economies than in economies like Poland. Shleifer said that he had addressed precisely that question. Obviously, if mone- tary policy is loose, even if you transfer control rights from the bureaucrats to pri- vate agents, bureaucrats could use subsidies and other mechanisms to convince private agents not to restructure, or to pursue policies that the bureaucrats want, which is what's happening in Russia. Monetary stabilization is essential to enterprise restructuring because loose monetary policy can undo all the benefits from trans- forming property rights. Extending the example of the Polish firm, one might ask if anytding really need be done about property rights if you have hard budget con- 128 Floor Dcssion of "Establishing Property Right?' straints. Shleifer is skeptical that such a strategy would be viable in the long run because so long as you retain bureaucratic control over enterprises, either you will have a problem sustaining the hard budget constraint or there will be incessant pres- sure from within for further subsidies. In the long run you need both hard budget constraints and the efficient allocation of property rights. Was there a special problem in the oil industry? Shleifer believes that in some ways the oil industry was in the best shape. If you gave management some incentives and cash flow rights, these firms could become profitable in the long run. In terms of property rights it might be easier to establish efficient control structures in these firms than in others. The problem in these firms was theft by managers, even if they owned an equity stake. Eventually, some kind of Pnforcement must follow the real- location of control rights. Property rights are not really enforced by law, said Frydman; they are self-enforced by a decentralized structure of incentives. Even in advanced economies law enforce- ment works only at the margin. That is why it is important for macropolicy to create a decentralized structure of incentives that support the economic, rather than the political, use of resources. That was the ultimate goal of transition, said Frydman. In a sense macropolicy-including import and other commercial policies-plays two roles: it denies bureaucrats the ability to use subsidies, and it creates incentives by forcing firms to return to the capital market. If a manager knows that he doesn't have to worry about outside investors, the firm is unlikely to attract outside investors. A participant from the World Bank noted that in an earlier presentation someone had praised the coupon method of privatization, and he wondered if this vehide was in danger of making ownership even more concentrated than it had been under the comrnunist system. He understood that in the Czech Republic one fund has de facto ownership of 600 industrial units, everything is managed from Prague by absentee landlords, and if, for example, the fund decides to dose one textile mill, a unit in Prague is more likely to survive than one elsewhere. Could the method could be modified to bring management doser to the firm's actual location and problems? Blanchard responded that one problem with Czech privatization had been this concentration of ownership, which is why the commercial banks that adjust man- agement of the funds should not be allowed to purchase shares in them; this policy would disperse ownership of the funds. Whether that policy was politically possible, Blanchard didn't know, but it would certainly make bank ownership weaker and the funds more independent. Shleifer's recollection of the numbers was that the refer- enced fund, which had tried to concentrate its holdings, had large equity stakes in only 50 enterprises, not 600, and that the funds that had stakes in hundreds of enterprises owned only trivial equity stakes, so things were not so bad as the partc- ipant might think. He did not see overconcentration of assets as much of a problem. These funds will develop expertise in providing capital and governance for the enterprises, said Shleifer, and until he saw evidence that they were up to no good, there was no reason to fear them unduly. After all, Fiat controls between 20 and 40 percent of industrial assets in Italy. If the Czech economy were to do as well as Italy... we should all be so lucly.