مقایسه بین سیستم های اقتصادی با کاربرد انتقال

This paper develops a general equilibrium theory of endogenous firm and class formation under non-contractibility with heterogeneous individuals. A collectivist economy, a private-ownership economy, and a mixed economy are compared on the basis of identical economic fundamentals (methodological symmetry). Each economic system generates specific inefficiencies so that none dominates the others in general. The main trade-off is between the welfare loss associated with risk-taking in the private-ownership economy and the informational problems in the collectivist economy. We then use this framework to study the political economy of transition between economic systems and provide detailed welfare results

The collapse of planned economies in the late 1980s led to a surge of interest for the economics of transition. A large body of literature has developed analysing many of the problems caused by these major changes in the basic rules governing economic activity.1 Most of this literature, however, starts with the premise that capitalism is more productive than socialism. While this is certainly a valid assumption for many short-run policy issues, this does not help us assess the relative merits of capitalism and socialism. Nor does it explain the enduring interest of many for alternatives to capitalism. In this paper, we take a step towards the development of a unified theory of comparative economic systems. Our approach is guided by the principle that any analysis of comparative economic systems should proceed on the basis of methodological symmetry within a general equilibrium framework. By ‘methodological symmetry’ we understand that economic systems must be compared on the basis of the same fundamentals, i.e. their representations must rely on identical initial conditions regarding individual preferences, endowments, technologies and the information structure.2 Methodological symmetry is a pre-requisite for isolating the effects of institutional variations on the economy.3 As, under methodological symmetry, the differences between socialism and capitalism do not lie in the fundamentals, one has to specify where they lie. We take the approach that the two systems basically differ in: (i) the allocation of property rights, and (ii) the mechanism for the allocation of goods and factors of production. In essence, capitalism is organised around private ownership of the means of production and a decentralised allocation of goods and factors of production, whereas socialism is characterised by public ownership of the means of production and a centralised allocation of goods and factors. More specifically, to compare a prototype capitalist (or, as we will call it here, ‘private-ownership’) economy with a prototype socialist (or ‘collectivist’) economy, we develop a static general equilibrium model of endogenous firm creation and class formation. The fundamentals of the economy which are invariant across systems can be summarised as follows. At stage 1, individuals (of different risk-aversion and managerial talent) must choose between becoming a manager or a worker. At stage 2, each manager sinks her labour endowment to set up a plant. Each plant may be either productive or unproductive, whereby the probability of any plant being productive depends on its manager’s talent. Each manager decides whether to learn about her plant’s productive status (e.g. by undertaking a market research), or to remain uncertain. At stage 3, workers are allocated across plants. Finally output is produced and distributed. A key assumption is that managerial talent and the input/output of any plant cannot be contracted upon.4 Under private-ownership, each plant is owned by the manager (or ‘entrepreneur’) who operates it. In the (existing and unique) equilibrium under private-ownership (P) all entrepreneurs choose to become informed about the productivity status of their plant and hire an optimal number of workers accordingly (i.e. zero if the plant is not productive). The reason is that with private ownership and a free allocation of factors of production, entrepreneurs, who make their own decisions and bear the un-insurable risks regarding their plants, are rewarded with a high return when successful and a low return when unsuccessful. These low returns act as a disciplining device on entrepreneurs inducing them to become informed and run only productive plants. Despite this productive efficiency, P is characterised by three types of inefficiency related to the absence of full insurance: in general, there are too few entrepreneurs, they have to bear their entrepreneurial risk, and some of the entrepreneurs are less talented than some of the workers, implying a misallocation of talent. Moreover, we show that P is not in general second-best either. Aggregate output in P falls as society becomes more risk-averse and, eventually, falls to zero with infinitely risk-averse individuals. With sufficiently many risk-neutral individuals, by contrast, P achieves first-best. Our formalisation of P builds on the well-established literature in the tradition of Frank Knight (1921) which has developed a general equilibrium approach to the process of firm creation and the emergence of entrepreneurial and working classes.5 In effect, our model of P extends Kanbur (1979a) by introducing heterogeneity in managerial talent and risk aversion.6 Such heterogeneity plays a crucial role when comparing economic systems. Under public-ownership, all plants are owned by a collectivist institution which hires a manager (or ‘director’) to operate each one. This institution also allocates workers across plants.7 We show that the (existing, but not necessarily unique) equilibrium of the collectivist economy (C) does not achieve first-best. In C, the non-contractible output of each plant accrues to the collectivist institution. At the same time, and because directors are indispensable to their plants, the collectivist institution must get each director to agree to operate her plant. As a result of bilateral bargaining, the collectivist institution makes a sure transfer payment to each director. This transfer is based on a sure surplus when the plant is known to be productive or on an expected surplus when directors have remained uninformed. Hence, the incentives of directors to become informed and to run only viable plants are much weaker than in P, since C provides de facto insurance. On the other hand, the amount of risk-bearing is lower in C, which reduces the inefficiency relative to P. For a sufficiently risk-averse society, C Pareto-dominates P. However, with sufficiently many risk-neutral individuals, C is Pareto-dominated by P. Similarly to P, C is not in general second-best either. While C is characterised by productive inefficiency, public ownership provides non-contractual insurance to (uniformed) directors, which is not feasible in P. We believe that this trade-off between productive efficiency and insurance captures one fundamental aspect in the comparison between capitalism and socialism. The crucial importance of the allocation of property rights and of non-contractibility bridges with the literature on incomplete contracts following Grossman and Hart’s (1986) work. In our model, employment in and output of any production plant are not contractible (while, as in Grossman and Hart, transfer payments between individuals or through the state are). Due to this non-contractibility, residual incomes and residual control rights in each production plant come to play an important role.8 In P, non-contractibility prevents any form of insurance and the allocation of property rights to entrepreneurs makes them the sole decision makers in their respective plants. In C, the necessity for both parties to agree on an allocation of workers at stage 3 and the associated bargaining on this issue has crucial implications on the ex-ante incentives of the manager, as is typically the case in the incomplete contract literature. We then extend our model allowing for a mixed economy in which the two systems may co-exist. People individually choose whether to move to the P-sector (either as entrepreneurs or as workers) or to the C-sector (as directors or as workers). We show that the equilibrium of the mixed economy (M) exists and is unique. An interesting feature of M relates to a negative pecuniary externality imposed by the P-sector upon the C-sector. Because of higher returns, the P-sector attracts informed and talented managers from the C-sector, leaving the latter with only uninformed and relatively more untalented managers whose marginal product is lower. This, in turn, drives incomes in the C-sector down. If this effect is strong enough, it may even lead to a complete close-down of the C-sector, i.e. to M coinciding with P. We finally use our model to shed light on some of the forces underlying the democratic choice of economic system. Since managerial talent and risk-aversion differ across individuals, society is not in general unanimous in its choice between economic systems.9 Our model identifies the losers from a move from C to P as being the failed entrepreneurs together with possibly the new working class (if their wage in P is lower than in C). In a popular referendum under majority rule, P may be preferred to C, even when it is known that there will be a majority worse-off ex-post. This is due to the fact that a majority of people may hope to become part of the new class of successful entrepreneurs. These results can be related to three key stylised facts regarding recent transition experiences. First, and as predicted by the model, income inequalities in formerly socialist countries have risen substantially (see Flemming and Micklewright, 1999, for a summary of the evidence on this issue). This increase has been more moderate in countries where entrepreneurship has been vibrant, as also predicted by the model. Second and following Fidrmuc’s (1998) empirical analysis, the winners and losers from the transition tend to be those identified by the model. That is, the winners consist of the more educated people (accepting education as a proxy for our concept of managerial talent), successful entrepreneurs, and even workers in areas where private entrepreneurial activity is strong. Third, the model offers a possible explanation for the political backlashes observed in several countries (such as Poland, Hungary and other countries in Eastern Europe) where the initial reformist governments were not re-elected and were replaced by coalitions including former communist parties. The remainder of this paper is structured as follows. After introducing the economic fundamentals, which remain the same in all systems (in Section 2), and characterising the first-best allocations (in Section 3), we develop equilibrium in the private-ownership, the collectivist and the mixed economies (in 4, 5 and 6, respectively). Section 7 is devoted to the political economy of transition between systems. Section 8 concludes and sketches some possible extensions.

The theory developed in this paper satisfies methodological symmetry and offers a general equilibrium perspective on comparative economic systems in an economy with endogenous firm creation and class formation. Due to non-contractibility, managerial incentives to discover the productivity of their firm and to hire workers accordingly depend on whether managers own their firms or not. When firms are owned by a collectivist institution, managers can limit their risk exposure by remaining uninformed. The marginal product of labour is then not equalised across plants. In turn, no individual needs to bear risk. In P, with owner-managers who have an incentive to become informed, production is efficient. Entrepreneurs, however, inevitably have to bear risk so that the very risk-averse individuals refrain from becoming entrepreneurs. The resulting shortage of entrepreneurs reduces aggregate output so that either economic system may Pareto dominate the other, depending on the distribution of risk-aversion in the population. Our model of M illustrates the fundamental difficulties associated with the co-existence of socialist and capitalist economies when the people are free to move from one to the other, i.e. ‘to vote by their feet’. The systems exert a negative externality on each other. P attracts the talented people and leaves C with a high proportion of untalented individuals. C, in turn, increases the shortage of entrepreneurs by exerting an upward pressure on the wages in P. The model also displays the conflict of interests between workers and entrepreneurs. If anyone prefers C to P, it will be the risk-averse and less talented individuals who chose to be workers. If anyone prefers P to C, it will be those individuals who want to become entrepreneurs, i.e. the less risk-averse and more talented. This theoretical framework can help us understand various stylised facts of transition economies such as increased inequality after reforms, this inequality being negatively correlated with the amount of privatisation, the identification of winners and losers, and political backlash. Despite the possibility of a vote in favour of reforms, some countries are still governed according to the old rules, often even by the old rulers. This could be simply due to the fact that a majority of the population is extremely risk-averse and, therefore, better-off under a collectivist system than with a market economy, at least in the short-run. Another observation which is consistent with our theory refers to the fact that the political debate in many countries has become polarised on the pro-reform/anti-reform issue. Moderate options are not popular in many countries. Our model proposes an explanation for this since partial reform is the majority’s preferred choice only under very special conditions. Instead, partial reform is more likely to be found in countries where reforms are not democratically decided. China and Vietnam offer examples of countries, where a still significant collectivist sector survives despite the presence of some alternative highly competitive institutions. Of course, our model remains extremely rudimentary and only puts the emphasis on certain aspects of the comparison between economic systems and the transition between them. In particular, our simplifying view of the labour market ignores unemployment. It would also be desirable to relax our extreme assumption of non-contractibility of a plant’s input.27 Among possible extensions, one could also think of the introduction of aggregate risk, since it is difficult to argue that macro-economic risk should not be taken into account during transition. Based on the Knightian theory of entrepreneurship, we have presented a static comparison of economic systems. Although it is often claimed that capitalism has defeated socialism through its superior ability to produce innovations, we are not aware of many formal dynamic models of comparative economic systems in a general equilibrium framework.28 It would be relatively easy, though, to develop a dynamic version of our model. One possible way to capture the Schumpeterian role of entrepreneurs as innovators is at stage 2 of our model. If we assume that managers who decide to become informed can use this information to increase the productivity of their plant, the technology available to an informed successful manager becomes instead of Eq. (2). This gives rise to a dynamic version of the model where the production function of a plant in period t depends on the number n of periods it was run by an informed manager, i.e. With this modification, our model is transformed into a simple endogenous growth model that preserves the main features of our comparison of economic systems. Without the need to specify more details of this dynamic model, it is apparent that the growth rate of the economy in period t will depend on the number of informed successful managers. Together with the fact that in P entrepreneurs always become informed while in C directors may chose to remain uniformed, it follows that P’s performance, relative to C’s, is better in the long run. This would support the view that private ownership is favourable for growth, because it provides better incentives for innovations, while directors of collectivist plants may prefer not to innovate in order to obtain insurance. But even in this dynamic version of the model, extreme risk-aversion still has adverse effects on the outcomes of P and, to a minor extent, of M and C. Specifically, with an infinitely risk-averse society, output and growth in P both tend to zero forever, while in C the growth rate tends to zero but not the output. This indicates that the qualitative nature of all our results would be preserved in a dynamic version of the model that incorporates the Schumpeterian view of entrepreneurship and innovation.