Colloquium 1996 Panel I : Balance Sheet of Economic Performance and Reforms in Cooperation Partner Countries

Between Spontaneity and Economic Engineering: Path Dependence in the Process of Economic Transition Bruno Dallago (a) Professor, Department of Economics, University of Trento, Italy Summary and Introduction The factors that determine the success or failure of transition are both general (play a role in all countries), and country-specific. Although many of them fall outside the scope of this paper, the most important among them are included in a general explanation of economic transition that must be tested with hard evidence. This cannot be done here in a fully satisfactory way: however, a first step is to present briefly our explanation and substantiate it with the most relevant development in Central and East European (CEE) economies. The Soviet-type economic system was inefficient - at least since the Sixties, and the exhaustion of extensive resources of growth (1). The system originated lack of motivation to better use existing resources and economic opportunities, and prevented adaptation to rapidly changing domestic and external environments. It also hindered and distorted technical and scientific learning and discovery and the diffusion of knowledge and information. Coordination of decision-making processes between the macro- and micro-level was inadequate. The consequences were high transaction and switching costs (2), barriers of various types and low competitiveness. Economic actors responded passively to incentives (like plan orders) and the effort to improve their individual situation was consequently addressed more to redistributive (rent seeking and possibly also destructive) activities than to productive ones, even in reformed economies. Systemic inefficiency was probably the most important factor that induced CEE countries to change the economic system. I consider here the most important features of the old (Soviet-type) system, because these are the cause of path dependence of the process of transition (3). My principal aim is to explain why path dependence plays such an important role during transition. My explanation of the differentiated path and likely outcome of transition in different countries is based on three building blocks: investment and capital, costs and the role of the state. In any system economic actors invest in system-specific assets. This investment originates individual systemic capital and produces a return that explains the motivations of actors to perform in the economy. Systemic capital creates an interest in the conservation of the system and originates path dependence when that system is changed. Individual differences in systemic capital and returns on it explain asymmetries among individual actors and therefore different individual opportunities and outcomes of systemic change (section 2). In order to overcome the conservative effect of systemic capital, an investment in systemic change must be implemented by those interested in changing the old system. These are those who invested in general assets or who get lower returns on their systemic capital. Investment in systemic change is analogous to an investment in a public good and creates externalities and opens the opportunity to invest in transitional redistribution as a precondition for capturing the great amount of contestable value during systemic change (4). Systemic capital influences the choice of goals, strategies and instruments of transition. These are examined in general (Section 4) and in the case of some of the most relevant countries (Section 6). Any change of the economic system involves great costs (transition costs) that are asymmetrically distributed throughout actors (Section 3). Transition costs, systemic capital and the opportunity for transitional redistribution require a strategic role of the state to assure a productive outcome of transition (Section 5). The paper is concluded by a mid-term assessment of transition in Central and East European countries.

Systemic Investment and the Basic Features of Soviet-Type Economies The position of individuals and organisations in an economy varies according to their accumulation of general and systemic capital. Assets and capital can be purely general, as can investment, being independent of any particular system and related to existing technology and determining its performance. Being purely general, they are not lost when the system is changed. Their value may even increase with property rights changes and the diminution of the cost of control. This is a powerful incentive to systemic change. The purest type of general investments in an economy are those related to physical assets. They consist of resources utilized to achieve specific technical knowledge and skills, to enter a productive organization, or to procure machines and raw materials for production. The Soviet-type system strongly promoted this type of investment, for education, vocational training, machines and building. For instance, in CEE countries there was a higher number of engineers per capita than in comparable market economies. However, technology (e.g. division of labor, specialization and organization) was peculiar to those economies due to economic, organizational, traditional, political and ideological reasons. As a consequence, even most of the general investment and accumulated general capital had a strong system-specific flavour (5), causing technical problems and costs during transition. Restructuring involves a particularly important and difficult task to recover general investment in the new system. On the other extreme, some assets, investment and capital are purely system-specific. These determine governance, systemic efficiency and the degree of system-specificity of any investment and capital. They are intimately bound up with the existing property rights structure and influence cognitive processes and hence attitudes, choices and goals of actors, in a given system. They also produce the highest outcome for the investor (greater bargaining power, income and better systemic knowledge and information) (6). Therefore, they create relevant asymmetries in bargaining power, information, and ability to influence the outcome of systemic operation and also wealth, income and privileges. This type of investment and capital was very important in Soviet-type economies. Typical investments involved setting-up and making the central planning system work (from central allocation of resources to the administered price system), joining the ruling nomenklatura, acquiring system-specific knowledge or skills, and adapting to and taking advantage of shortage. These investments and capital are lost with system changes, or can even acquire a negative value. They have been an obstacle to transition to a market economy because of bureaucratic, authoritarian and passive institutions they helped create. The Cost of Transition and Path Dependence A precondition for transition is that systemic entrepreneurs invest in change. Their investment must be the higher, the greater is old systemic capital. However, investment in change creates positive externalities, substantially reducing the investment that other actors must make in order to adapt to the new system. Systemic entrepreneurs invest in order to capture the existing opportunities and to obtain a greater return on either their general capital or their systemic capital (diminished costs of control, more productive individual use of capital) or to have better opportunities to invest in system-specific assets. In CEE these opportunities stemmed from: the emergence of "technocrats" whose return on investment in general assets was reduced by existing property rights structure; the inability to substitute exhausted extensive resources of growth; the policy of openness that revealed the backwardness and inefficiency of the Soviet-type system and introduced imitation of western life style; nationalistic policies that were implemented to elude the consequences of political and economic difficulties; opportunities to gain extraordinary advantages in the second and the criminal economy. Investment in systemic change was open and conscious, including attempts by political opponents to overthrow the system, and implementation of reform by leaders (like Gorbachev's perestroika). In other cases it was the by-product of investment in other assets (e.g. criminal activities). Investment in systemic change is only a first step. Those that implement the transformation and adaptation of the economy are also costly. First, actors have to learn to operate, collect information and measure the advantages and disadvantages of exchanges in the new system. Additional learning and measurement costs constitute a substantial part of investment in new system-specific assets. Until this investment matures, opportunities are missed (e.g. missed contracts and production). Second, additional enforcement costs derive from overcoming defence of the old system (7), replacing old structures with new ones, enforcing agreements, controlling free riders and correcting adverse selection, and preventing monopolies. Third, transition implies the redistribution of rights, costs and benefits from social exchange (e.g. through privatization), which gives rise to adaptation costs in line with the new market system. Until these processes are over, losses are incurred in the form of missed contracts, organizational overload and low demand, from which a fall in output and employment may follow. The new system, if more efficient than the old one, should cover transition costs through a higher return in the long run. In the short-medium run these costs are a major cause of transitional depression (8). The distribution of costs among actors is uneven and this has major distributive consequences. If systemic change takes place through expropriation of the old investors and the value of transition costs is not higher than that of old systemic capital, the cost of systemic change is entirely paid by the losers (9). However, in the case of radical systemic change - as in Central and East Europe - old systemic capital becomes valueless. In that case, investment goes into destruction of the old systemic capital and creation of new system-specific assets. However, systemic change can be based on agreement between the parties to share costs and benefits and to compensate losers. This situation arose in CEE at transition, when "round tables" between the government/party in power and the opposition and international agreements among the parts of disrupted countries (fSU, Yugoslavia and Czechoslovakia) were organised. In neither case can transition costs upon those controlled be avoided. The most spread, least specialized among general capital owners are worse off, at least until the new system compensates transition costs. Old systemic capital and transition costs cause path dependence which renders some specific solution less difficult and less costly and therefore more likely to prevail than others. (10) First, actors are afraid of losing their old systemic capital and of undertaking new costs in exchange for a future and uncertain higher return on their general capital. Systemic change encounters opposition, affords opportunistic behaviour and uncertainty. Therefore, changes take time, old institutions survive informally even when formal structures are abolished, and a consistent revolutionary collective action is difficult to persist through time. Path dependence derives from efforts to reduce enforcement costs. Second, the amount, structure, and type of old systemic capital form (non convergent) subjective models that influence choices available to actors, distribute costs and advantages deriving from social exchange and therefore determine the opportunity for actors to invest in systemic change and in transitional redistribution. In fact, in CEE countries members of the middle class ("technocrats" and intellectuals) were the ones who invested most in systemic change. When old formal institutions are disrupted because of systemic change, and until new ones have been internalized, actors do not have the relevant information and computational ability required to take decisions. Therefore, rational actors adopt strategies of simplification, usually conservative. Path dependence reduces learning and measurement costs. Third, actors may be able to convert old-type systemic investment to reduce adaptation costs. They may exploit still valuable assets (e.g. personal connections and networks, bargaining power from strategic knowledge and skills accumulated in the old system). This was particularly the case of managers of enterprises about to be privatised, and of many intellectuals (even previous teachers of Marxism-Leninism). Goals, Strategies, and Instruments The implementation of transition requires goals outlining systemic features (a market economy) desired, strategies that define the path of transition, and the concrete instruments utilized (like the specific types of privatization). However, all these, chosen or pursued, vary at both individual and general level. In fact, the choice of goals depends on various factors (ideas, ideologies, cultural and historical factors, beliefs, illusions, propaganda, social needs, etc.), strategies are influenced by political factors and instruments are limited by technical knowledge and organizational capabilities. However, the choice and outcome are also influenced by systemic factors. First are the different types of investment and capital discussed before. Because these are asymmetrically distributed, the choice is the result of processes not perfectly competitive, in which path dependence plays an important role. In particular, in transition economies the choice was determined by technocrats and the old nomenklatura that possessed valuable assets and bargaining power accumulated in the old system. Hence, the greater the importance of technocrats in the old system and the more technically relevant the still valuable assets of the nomenklatura, the greater are chances that path dependence plays a productive character (11). In this case chances are that goals consist primarily in the attempt to increase the return from general (technical) capital, and this can be done only if the new system is more efficient and productive than the old one. Therefore strategies are chosen to increase the role of technical components of the economy and instruments are implemented which pay a low price to conservative and populistic policies. This apparently is the case of CEE countries: Poland, the Czech Republic, Hungary, Slovenia, and possibly the Baltic Republics. In the converse case, (lesser the importance of technocrats, lesser the technical significance of the still valuable assets of the nomenklatura), the greater are chances that path dependence leads to the prevalence of redistributive processes. In this case goals probably consist primarily in conserving the value of old systemic capital and converting it into new systemic capital to keep returns for individual actors. This can be done only if the new system is quite close to the old one (maybe except ideology and political colour). Strategies are chosen to keep a strong role for the non-technical components of the economy and instruments are implemented which pay a high price to conservative nationalistic and populistic policies. Changes of ruling persons and organizations, institutional framework, the structure of enterprises and their behaviour are modest and sometimes only apparent (12). This apparently is the case of East European countries and the Balkans (13). A second conclusion concerns the diverging outcome of the same strategies and instruments in countries different from the point of view of old capital and consequently of the specific character of path dependence. This complements the role of differences in the technical components of policies or instruments. An interesting case in point is the strategy to build a market economy via privatization of former state owned enterprises by voucher privatization in the Czech Republic and Russia (14). Other interesting cases are given by the quite different outcomes of gradualist policies in Hungary and, e.g., Ukraine, and shock therapy for macroeconomic stabilization in Poland and Russia or Croatia and new Yugoslavia. The Difficulties of Productive Transition and the Role of the State Transition creates a great amount of contestable income stream from, and value of, most assets. It is impossible to establish the actual flow of income that assets can produce, and it is difficult to enforce rights until the new system has been stabilised. Therefore, actors will try to capture part of that contestable value and income stream. However, with a changing system and unclear and mutable selection criteria, there is no guarantee that wealth will be captured by the most productive actors. Income and value are also affected by actors who influence the variability of attributes and do not bear the full cost of their action (15). These factors create a great need to monitor and enforce as a fundamental precondition for the productive outcome of transition. Monitoring and enforcement are all the more difficult and costly in transition and their outcome is uncertain. The costs are met by those controlled, those who invest in general assets. They are paid the market return on their capital, minus the resource costs of monitoring and policing. However, costs are reduced by internally enforced codes of conduct, like moral and ideological values. Monitoring and enforcement are implemented by investors in systemic change and by those who obtained relevant advantages in the new property rights through transitional redistribution. It is likely that the main advantage will accrue to actors who own still valuable old systemic capital, and who invested in transitional redistribution (16), and to those more determined in the use of violence. This may lead to the prevalence of inefficient mechanisms of selection of actors and decisions. Where the dominant form of investment and capital is system-specific, transition is first addressed to dispossess the old nomenklatura. A typical nomenklatura defensive action in this situation is the adoption of strategies of simplification, like promotion of nationalism and, if necessary, secession, maybe by utilizing the state to this end. This strategy also offers major opportunities to convert old systemic capital into valuable new capital, and to seize new opportunities. Therefore, if the goal is a socially productive outcome of transition, the question of the structure and the role of the (new) state is particularly relevant. This is not an easy task, because the structure and role of the state are determined by the nature of existing systemic capital and consequently by path dependence. The role of the state during transition exceeds the functions normally performed in a well established market economy. Destruction of the (old) state must be linked to promotion of collective economic action for transformation of the economic system. In these circumstances the state must intervene first to define the rules of the game, that is formal institutions. Consequently, the state circumscribes the outcome of social action, by excluding certain results deemed socially undesirable (consequence of asymmetries unjustified in the new system) and it promotes institutional discovery and learning in specific directions. Second, during transition the duty of monitoring and enforcement is of particular importance. Third, at the beginning of transition there is a lack of market actors with sufficient expertise and weight, so the state must also assume an active (possibly indirect) role in the formation of market actors, in two senses: by promoting the development of new market actors, and by assisting the behavioural change of existing actors. International assistance is particularly important in this third line of action. Different states have different capabilities to implement the three lines of action. However, this should not prevent building a new structure and role of the state in a market economy. First, the structure and role of the state is not necessarily the same in all countries. Second, the implementation of feasible structures and actions which serve the development of the desired economic system is essential, even if the present situation may be quite different from that of most developed market economies. The state's role is to help the transformation by utilizing - (in an innovative way) - useful elements of path dependence and progressively help change the others. To fulfill this set of goals, a proper selection of state structures, agents and payoffs must be made through careful planning and trial and errors. An example is provided by the outcome of voucher privatization in different countries. In the Czech Republic, a carefully planned program through free distribution of vouchers to all adult citizens aroused initially only limited interest. The situation rapidly changed when investment funds intervened. These new economic actors were founded and controlled by state-owned banks and enterprises and foreign investors. For a long time investment funds were free of state regulation, and this encouraged various abuses: opportunist and predatory behaviour, formation of monopolies and the persistence of non-market behaviour. The situation improved with state regulation of the activity of investment funds, thus laying a sounder basis for competition. However, even after this intervention and successful completion of voucher privatization, the fundamental question of corporate governance is still to be properly solved. In countries in which initial conditions for a market economy were less favourable than in the Czech Republic and the state apparatus was dismantled without a new well-established one to replace it, state failures are much more numerous and have more serious consequences. This was particularly evident for voucher privatization in Russia and Mongolia. This solution was adopted to speed up privatization, defeat the old nomenklatura, and create strong popular support for systemic change and privatization. The outcome was also affected by development of a secondary market for vouchers: privatization was largely dominated by the old managers, who collected enough shares to control the firm, usually at low cost. This outcome was eased by inflation and very generous self-granted bonuses. The existence of so many (often fundamental) functions for the state creates a risk of various (often major) failures. Particularly important among these is creation of a new and powerful bureaucracy, difficult to control and remove when transition is over. However, first, the performance of the central role of the state requires specific policies, functions and priorities and not necessarily large staff or great financial resources. Second, in a democratic society the state must be controlled by political and administrative means based on full political, social and economic accountability. Also the rapidly growing private sector can counterbalance the power of the state. These factors induce the government to divest itself of its tasks whenever these can be better executed by other organizations (17). Path Dependence and Transition Outcomes Taking into account the factors discussed above, one can explain why path dependence has led different countries to pursue different goals, with different strategies, different instruments and different outcomes. In this section I present the major differences among transition countries that I consider most significant for their transition specificities. I shall then try to connect these differences to the building blocks of my explanation of transition. Hungary had the longest record among former Soviet-type economies of a rather consistent and successful economic reform towards decentralization. It was natural that it used a market approach to privatization and institutional reform. Neither a strongly centralized and effective central administration of the economy nor a strong and closed nomenklatura existed in Hungary. Consequently, investment in system-specific assets was limited (compared to investment in general assets), as was systemic capital. Therefore, the basic instrument for a centralized approach to transition was missing or ineffective. Enterprise managers enjoyed substantial autonomy (further strengthened and converted into ownership through spontaneous privatization) while workers' councils were powerless. This reduced the possibility of voucher privatization or other types of preferential schemes. A dynamic, fairly strong non-socialized sector and underground economy existed, which produced in strict contact with state-owned and cooperative enterprises. Hence, a limited amount of capital also existed in private hands. Therefore, on the supply side entrepreneurial capacities and structures of corporate management were better developed than in other CEE countries. The use of market methods in economic transformation were also required by the difficult macroeconomic situation, with high domestic and external debts and inflation. The debts required new sources for financing - and privatization was too appealing in this sense. However, the consumer market was balanced and consumers were used to paying for many goods and services roughly a market price, of which market privatization could be seen as an extension. The openness and relatively fast change of the nomenklatura, although paternalistic, made generally acceptable the change of persons in leading positions. The nomenklatura could hardly react to the menace that transition brought to their privileged positions with nationalistic policies, secession or authoritarian solutions. It was easier, and more advantageous to adapt to the new situation, either by joining the new parties and competing in the electoral arena or by using their still valuable systemic capital to gain advantages in the new property rights structure (18). A rise of nationalism and populism was noticed and played a role in the victory of the center-right coalition in the first elections, but they did not become dominant factors in politics. Overall, transition was rather gradual and pragmatic. Transition costs were increased by technical and political mistakes, leading to some waste of potential. Poland, before transition, led a far less consistent path than Hungary. However, behind an apparently inconsistent and sometimes tragic situation a dynamic society and economy existed, used to relying on their own forces. Also, the unstable political situation rendered the nomenklatura precarious and therefore open and technically based. Investment in pure system-specific assets was limited and the government's control over the economy weak. Enterprises were autonomous de facto, the private sector relatively strong, and the underground economy widespread. The macroeconomic situation was the most difficult among CEE countries, with the highest foreign debt, strong inflationary pressures and shortages. In this situation a shock therapy type of macroeconomic stabilization was a must. It was also socially and politically feasible, because it rapidly ended a long period of shortages on the consumer market. However, it destroyed private monetary savings that could be used in privatization and for investment. This plus the difficult and loss-making situation of state-owned enterprises, made privatization difficult on the demand side. However, it was even more so on the supply side, because the success of macroeconomic stabilization and other factors made institutional reform slow and inconsistent. The uncertain transition favoured rapid expansion of the greenfield private sector and restructuring and corporate governance in various state-owned enterprises. Managers of state-owned enterprises chosen for their technical competence rather than for system-specific coherence, often kept their positions. Lacking any opportunities to reconvert their systemic capital by obtaining ownership of enterprises, they concentrated their efforts on improving corporate governance and productive outcome. This decreased chances of large scale privatization but rendered the economy more productive. Rapid development of the private sector and adaptation of state-owned enterprises reduced the weight of otherwise high transition costs. In the fSU, the senior nomenklatura ranks that were more concerned about the economic and military decline tried for some years to reform policy, but the economy was hardly touched. Reform policy in a country with a strong and very closed nomenklatura and with great system-specific capital proved to be too difficult. Unable to adapt and lacking alternatives, the nomenklatura defended its role, power and privileges. Finally, the entire system was destabilized. System-specific assets were separated from general assets, and the former clearly dominated the political and economic scene. Also the corrupt nomenklatura was costly and unreliable, making the cost of monitoring and enforcement high and return on general investment modest. General investment was concentrated in a few sectors which had enjoyed priority during the Soviet period. Therefore, transition is hardly smooth and successful, costs are high and popular support is passive. With investment in system-specific assets the most relevant component of social investment, the public had invested also in those assets (though modestly) but with little control over that investment. General capital and investment in priority sectors also had a strong systemic bias. Great uncertainty in the nomenklatura and society led to failure to create a group interested in increasing return on general investment because of the implicit loss of systemic capital. A typical response in this situation is nationalism, ranging from mild nationalistic policies to violent secession, as happened in (and within) all post-Soviet states. Nationalism is the classical response of the dominant class when its power is threatened (19): it recreates consensus around that group presenting itself as defender of the most traditional 'glue' of social cohesion. Nationalism may also play a role in rebuilding the system after the break with the old one. This development plays a major role in explaining the relative stability of post-Soviet Russia, but it is also a threat to democracy and the development of a market system. In these conditions privatization cannot contribute to the creation of a market economy or to improve corporate governance. The endowment of systemic capital determines the selection of managers and general capital enjoys low returns. This opens a wide room for redistributive processes. Managers, likely to be members of the old nomenklatura, do not feel strong pressure to invest in general assets. Consequently, enterprises may continue to depend on state allocations and competition is likely to be strong only on this side. The economic system is likely to keep old fundamental features and redistributive processes are likely to be more important than productive ones. In former Czechoslovakia, the Soviet-type system was highly centralized and orderly, as was transition. The dominant form of investment was in system-specific assets, and the nomenklatura was closed and powerful. However, traditional institutions and the rule of law made costs of monitoring and enforcement low and the nomenklatura was not corrupt. The return on general investment was high, and general capital was important before transition. Also the selection of the nomenklatura was - at least partially - based on their endowment with general capital. This was quite different from the Soviet-Russian case. Managers, selected also according to their endowment with general capital, were used to rather hard rules from central planners. The central administration (the nomenklatura, except a few political leaders) were technically endowed and well disciplined, as were the law abiding enterprises. It was thus possible to implement a highly centralized approach to transition. Both privatization and the state structure took this form through secession. Both were decided from the center, subject to no democratic process. Privatization via free distribution was made possible by a macroeconomically well balanced economy. Managers passively accepted these developments and continued to run their enterprises. So did workers and citizens. Free allocation with no secondary market for vouchers eliminated opportunities for redistribution, and transition increased the returns on investment in general assets with lower costs of monitoring and enforcement and more powerful and effective incentives. Therefore, transition costs were lower than in most other countries. However, free allocation of vouchers together with the lack of new systemic knowledge and dispersed ownership rights gave control of enterprises to investment funds. These were usually owned by (mostly state-owned) banks. The question of corporate governance has remained unsolved, and the budget constraint is still soft. However, uncertainty over property rights was solved rapidly. Secession was the easiest solution in two new countries (the Czech Republic and Slovakia) that were quite different in systemic investment, economic structure and dependence on foreign markets. Only the Czech Republic implemented consistently and rapidly the above outlined strategy. However, secession was not accompanied by strong nationalistic policies, because it was chosen after fundamental political and personnel changes. Yugoslavia presented the most complex situation. The country was a decentralized federation without a strong center in control of the economy. After Tito's death it remained without a clear political guidance. Members of the nomenklatura invested more and more in peculiar system-specific assets: namely, in nationalistic assets. Investment in federal system-specific assets and in general assets declined. Nationalistic struggles and policies, with interrepublican and interethnic redistribution, left no possibility to promote production. Stagnation and high inflation were inevitable, pushing investment in nationalistic assets even further. War and breakup of the state and country resulted. Transition costs increased substantially and depressed returns from general capital further. At the end of the process hardly any federal asset existed (not even a federal nomenklatura) and general investment was modestly rewarding. In particular in Serbia and Croatia, corporate control was nationalized, as social enterprises were subjected to control of those investing in nationalistic assets. These were members of the nationalistic nomenklatura, that is the old communist nomenklatura who were ready to change the nature of their investment. The later privatization gave also ownership to the same persons, often for free, thus expropriating definitely the population at large. However, this hardly solved the question of corporate governance and investment in general assets, because selection of investments and people is still based on nationalism. Consequently, returns from general capital remained depressed. Conclusions: A Mid-term Assessment In the previous sections we have seen that actors invest in general and system-specific assets and consequently own general and systemic capital. This, together with subjective factors, explain diverging interests and choices and the pursuit of different goals. Transition from one economic system to another is determined by the above mentioned factors, by the possibility to convert old systemic capital into new, together with the costs of transition. This creates path dependence of systemic change. Three major conclusions concerning path dependence: Transition cannot be implemented in one stroke, because of the time and cost required by investment in systemic change and by the reconversion of still valuable systemic capital, and because of the resistance of owners of old systemic capital. This concerns institutional transformation, while other components (particularly macroeconomic stabilization) may require shock therapy. Institutional changes can, and sometimes should, be implemented rapidly. Voucher privatization in the Czech Republic, Russia and Mongolia, are cases in point. However, implementation of the technical operation does not mean that the desired economic system is in place. Much more is needed, involving a long period of adaptation and adjustment. Different countries present different outcomes, deriving from dissimilar starting conditions and transition strategies, based on differences in the fundamental types of investment and capital. Hence, development of substantially diverging economic (and political and social) systems can be anticipated, raising the fundamental question of "what type of capitalism?" (20). One view is that two basic types of capitalism are being developed in the Czech Republic and Slovakia, and in Poland: the latter (and possibly Hungary) is approaching the Italian model of political control (or, better, of political bargaining), while the former two are moving towards the German bank-centric model. From this perspective, the Anglo-American model of indirect capital market control is of secondary importance. One should add that in other countries - notably a large part of the fSU and Yugoslavia - the most likely outcome is some form of corporatist capitalism. Therefore, transition is developing into a crucial issue in debate on economic development. Similar policies may have quite dissimilar consequences, with different outcomes, in distinct contexts. Because of the factors that generate path dependence, chances are that formally similar policies are implemented differently in various countries. Macroeconomic stabilization and privatization in CEE offer many examples. However, determinism as envisaged by path dependence does not fully explain transition. As mentioned above, subjective factors may also be relevant. Even more so are policies that the government and collective bodies put into effect. Policies implement specific frameworks and influence the structure of payoffs. Thus they may have great consequences for the outcome of transition. This also refers to international assistance and cooperation. For this reason, the role of the state and collective bodies pursuing public functions has the most influence to help a country depart from its "natural" development path and approach the socially desired goal. Certainly, public intervention is bounded in its choices and strategies and must consider the reasons and consequences of path dependence. But this is another story, which cannot be developed in this paper. Footnotes The research on which this paper is based has been financially sponsored by Euroest (the Centre on the Evolution in Eastern Europe at the University of Trento). The author is grateful to Silvio Goglio (University of Trento) and participants at the 1996 NATO Economics Colloquium for helpful comments on an earlier version of this paper. However, any responsibility for errors and weakness of analysis remains solely with the author. An economic system is inefficient when it is the cause of economic performance lower than could be achieved using the same amount of actual or potential resources. Systemic inefficiency was easily testable by comparing the performance of Soviet-type economies with comparable market economies. By switching costs is meant the value of the resources needed to replace in certain circumstances the solution of a particular set of problems or the use of a specific technology employed to solve problems with a different one. Particularly relevant was the incapability of these economies to solve the problem of the transformation from "extensive to intensive growth". By transition is meant a situation when economic agents are not coordinated anymore by old institutions and structures (e.g. central planning), while new ones (e.g. market) are not yet in place. Transitional redistribution refers to property rights over existing general assets (like the ownership of enterprises and real estate) or system-specific assets (like positions in the new government, political parties or privatization agencies). It is enough to think of the oversized workforce, a high concentration of production in enormous enterprises, the unplanned although essential role of tolkachi within state-owned enterprises, the supply of social services by enterprises, the state monopoly of foreign trade, the production for Comecon markets, the mono-bank system and the system-specific features of financial and accounting services to have a clear picture of the system-specific character also of general investment and capital. Although ideology played an important egalitarian role particularly in the early years of the new system, the greater investors in Soviet-type system-specific assets enjoyed the highest returns in terms of bargaining power and prestige, but certainly also wealth, income and privileges. Consequently, and if we exclude agents operating in the underground economy, they were also among the richest persons in their country, particularly since the Brezhnev years. In the form, for instance, of setting up police control or paying transfers and other incomes or privileges to specific social groups to compensate their losses and obtain their support. This term was introduced by J. Kornai (1993) to denote the typical output fall during systemic change, which is largely independent from starting conditions or the specific transition path. In fact, the phenomenon has characterized all Central and East European economies. On the different aspects of transitional depression see R. Holzmann, J. Gács and G. Winkler (1995). Losers are those who own old systemic capital that becomes valueless. Winners are those who invested in systemic change and imitators who succeeded in converting their old capital into new systemic capital. Also owners of general capital are possibly among winners - at least in the long run - provided that the new system decreases the cost of control. Losers are concentrated in the system specific production that was stopped as a consequence of transition. This is the production of planning offices or that associated with large international investments or those serving the Comecon market, or that part of industrial production that responded to the social and political necessity of maintaining full employment and supporting system-specific social groups, like sectoral lobbies. This outcome was proved by many authors in the case of organization also on the basis of laboratory experiments. Cf. A.T. Denzau and D.C. North (1994), M. Egidi and A. Narduzzo (1996). According to László Csaba (1996, p. 53), by the late 1980s "...Hungarian corporate bosses ceased to be clerks taking and transmitting commands. They had evolved into independent-minded, autonomous and very innovative business executives...". As was recently proved by a study by Olga Krishtanovskaya, published in Izvestiya and cited in Financial Times (January 13-14, 1996), the members of the "new" political and economic elite largely coincides with the old nomenklatura. Apparently, this is also the position that Yegor Gaidar holds in his recent (1995) book as quoted by Pekka Sutela (1996, pp. 40-41) when he detects "... a universal attempt of bureaucrats to 'privatise' their power into property. Late Soviet socialism developed into a bargaining - or administrative market - economy, and perestrojka opened the gates for nomenklatura privatisation. For external and internal reasons Soviet socialism ended not with a violent revolution, but with an essentially evolutionary compromise. Therefore the choice between open market economy of the Western type and nomenklatura capitalism, another version of the Asiatic mode of production still remains to be made." This point is corroborated by a recent statement on voucher privatization by Anatoly Chubais, formerly responsible for Russian privatization, in Izvestiya on December 6, 1995: "Distribution of property rights in Russia, ..., as also in other countries, takes place in proportion to existing power elites." (quoted by P. Sutela, 1996, p. 40). This induces the author (P. Sutela) to conclude: "Russian privatization fundamentally legalised existing de facto property rights. It did not redistribute them." (p. 39) These are, for instance, irregular or criminal actors, old politicians, managers or even workers. A typical case is spontaneous privatization in transition countries. Managers, politicians and directors in the old administrative and political machine used their control over assets to convert it into ownership of specific industrial assets. Conversely, the state, workers and the population were expropriated of property rights enjoyed in the past. Cf. É. Voszka (1996). An interesting idea to solve the problem of the role of the state is that of the competitive government proposed by A. Breton (1989). Csaba (1996, p. 55) expresses this concept as follows: "The reformist wing of the outgoing rgime had rightly seen [the negotiated, gradual change in 1989-90] as a medium-term investment if it plays an initiating role in orchestrating transition." On the role of nationalism see in particular A. Breton (1964), A. Breton et al. (1995). H-J. Wagener (1996). References Breton, A. (1964), 'The Economics of Nationalism', The Journal of Political Economy, vol. 72, August, N 4, pp. 376-386. Breton, A. (1989), 'The Growth of Competitive Governments', Canadian Journal of Economics, November, N 4, pp. 717-750. Breton, A. et al. (eds.) (1995), Nationalism and Rationality, ed. by A. Breton, G. Galeotti, P. Salmon, R. Wintrobe, Cambridge University Press, Cambridge. Csaba L. (1996), 'Transformation in Hungary and (in) Hungarian Economics (1978-1996), paper presented to the conference on Economics and Systemic Change, Berlin, June 27-30, 1996 Denzau, A.T. and North, D.C. (1994), 'Shared Mental Models: Ideologies and Institutions', Kyklos, Vol. 47, N 1, pp. 3-31 Egidi, M. and Narduzzo, A. (1996), 'The Emergence of Path-Dependent Behaviors in Cooperative contexts', University of Trento, Trento (mimeo.) Holzmann, R., J. Gács and G. Winkler (1995), eds., Output Decline in Eastern Europe: Unavoidable, External Influence or Homemade?, Kluwer, Dordrecht Sutela, P. (1996), 'Economics under Socialism: The Russian Case', paper presented to the conference on Economics and Systemic Change, Berlin, June 27-30, 1996 Voszka, É. (1996), 'The Revival of Redistribution in Hungary', in. Dallago and L. Mittone (eds.), Economic Institutions, Markets and Competition. Centralisation and Decentralisation in the Transformation of Economic Systems. Cheltenham: Edward Elgar, pp. 274-291 Wagener, H-J. (1996), 'What Type of capitalism is Produced by Privatization?', in: B. Dallago and L. Mittone (eds.), Economic Institutions, Markets and Competition. Centralisation and Decentralisation in the Transformation of Economic Systems. Cheltenham: Edward Elgar, pp. 90-110