The idea of market socialism has had a strong influence over economists: it seemed to hold open the possibility that one could attain the virtues of the market system—economic efficiency (Pareto optimality)—without the seeming vices that were seen to arise from private property. In the wake of the fall of the socialist economies and the repudiation of market socialism by countries such as Hungary and Poland, which had tried it, there were two reactions: defenders of market socialism argued … the idea of market socialism had not really been given a fair trial; and critics of market socialism said, in so many words, “I told you so: it was clear that socialism, in any form, simply could not work.”
In this essay I argue that the idea of market socialism is fundamentally flawed—and for many of the same reasons that the Arrow—Debreu model on which it is based is flawed as a description of the market economy. I contend that if that model (or its precursors) had provided a correct description of the economy, then market socialism would indeed have had a running chance of success. Thus the failure of market socialism serves as much as a refutation of the Arrow—Debreu model of market economy as it does of the market-socialist ideal.
The fundamental problem with both models is that they fail to take into account … the absence of perfect information—and the costs of information—as well as the absence of certain key risk markets; …These information-theoretic concerns have changed both the questions economics asks …. To the classical three questions of economics—What should be produced? How should it be produced? For whom should it be produced?—we now add a fourth: How’ should these decisions be made, and who should make them?
In the economy of Joan Robinson, or Arrow and Debreu, decision makers and the structure of decision-making play no role. Robinson described the job of the manager of a firm as simply looking up in the book of blueprints the appropriate page corresponding to current (and future) factor prices. …. Further, the lack of concern of Lange, Lerner, and Taylor for managerial incentives would be of little moment: managers could essentially be replaced by automata.
The view of economics encapsulated in the Arrow—Debreu framework … is what I call “engineering economics,” …. economics consisted of solving maximization problems. Accordingly…the advantages of decentralization amount to little more than a convenient computing algorithm, …The central point is that in that model there is not a flow of new information into the economy, so that the question of the efficiency with which the new information is processed—or the incentives that individuals have for acquiring information—is never assessed…. the fundamental theorems of welfare economics have absolutely nothing to say about … whether the expenditures on information acquisition and dissemination— is, in any sense, efficient.
Given the implicit assumptions that went into the Arrow—Debreu model, it is no wonder that standard economic theory paid so little attention to the processes by which resources were allocated. …Arrow and Debreu had the insight that… it made little difference whether there was a static, one-period resource allocation problem or a multiperiod resource allocation problem….
Thus the standard paradigm was concerned with the rules for allocating resources (e.g. setting marginal rates of substitution to marginal rates of transformation), not … how decisions about resource allocations get made, who makes those decisions, and how those who make those decisions are selected...….An enormous amount of managerial time is spent in choosing who will fill various positions in the firm… Indeed, we now have general theorems establishing that, in the absence of a complete set of Arrow—Debreu markets, there will not be unanimity among shareholders about the appropriate course of action which the firm should take. …
Does Modern Theory Suggest a Greater Plausibility for Market Socialism?
At least two of the central results of modern economic theory should, if anything, have reinforced the belief in market socialism.
Absence of Futures Markets and the Role of Government in Allocating Investment
One of the underlying assumptions in the now-standard model of (competitive) market economies…is that there is a complete set of futures markets. These futures markets are essential for making the correct investment allocations. Indeed without a complete set of futures markets extending infinitely far into the future, the economy can set off on a path which is locally intertemporally efficient, looking exactly like an ordinary rational expectations path, and only in the distant future does it become evident that the economy is inefficient. There appear to be no private incentives to correct this seeming long-run inefficiency.
The Principal-Agent Problem and the Separation of Ownership and Control
Moreover, the early work of Knight (1951), Berle (1926), and Berle and Means (1932) on the separation of ownership and control reached fruition in the principal—agent literature (growing out of the papers by Ross  and Stiglitz ) with costly information, shareholders could exercise only limited control over managers. … subsequent theoretical literature on takeovers (Stiglitz, 1972a, 1975b; Grossman and Hart, 1980) …. further reinforced the conclusion concerning .. managerial autonomy. …. For large firms there is no “single owner” maximizing the expected present discounted value of profits… Does ownership really matter?….
Is Market Socialism Less Necessary than Previously Thought?
….. Two further results of modern theory have more ambiguous implications.
At least some advocates of market socialism believed that the relevant choice was not between competitive markets and market socialism but between monopoly capitalism and market socialism (see Persky, 1989). They believed that in large sectors of the economy competition was not viable. …The growth of large enterprises in the early part of the twentieth century led many economists to extrapolate the trend and to envisage a market economy in which each of the major sectors—steel, oil, automobiles, aluminum, and so on—was dominated by one firm, or at most a few. with larger scale enterprises arising from lack of organizational control, could be limited.....
Thus the alternatives facing economies were to allow monopoly capitalism to take hold, .. that might follow; to have direct government control of these sectors; or to attempt to regulate and control the exercise of monopoly power…. Few democratic governments found the first acceptable. …more than fifty years after the passage of the landmark antitrust legislation, many of the core American industries remained highly concentrated; …These events simply reinforced belief in the second strategy—government ownership and control.
Countervailing this intellectual trend, which …. provided greater support for market socialism, is the internationalization of the world economy. …. While the American market may have been large enough to sustain only three large producers, the world economy is large enough to sustain many more.
Of all the market failures … the greatest was the Great Depression, …. Keynes … showed that the market failure could be corrected without abandoning market processes. Only limited market intervention …was required. …. Ironically, the past twenty years has seen a reduction in confidence in Keynes’s analysis. … the historical evidence is of limited value…. As socialist economies tried to decentralize, moving more toward some versions of market socialism, they exhibited greater difficulties in macroeconomic control ….
But there was another message of Keynes that was heard more clearly: the macroeconomic ills of capitalism were curable. The economic system needed no fundamental reforms but only selective government intervention. It is in this sense that Keynesian economics greatly weakened the case for market socialism.
.. I suspect that the developments in modern economic theory ….should have led to greater doubts concerning the effectiveness of market processes. Yet most economists today would express greater, not less confidence in market processes than they would have fifty years ago. They would cast doubt on the relevance of the Lange—Lerner—Taylor theorem asserting the essential equivalence of competitive markets and market socialism. ….
.. I present five central economic reasons for the failure of market socialism. … the model of market socialism
1.Underestimated the significance of the incentive problem;
2. Underestimated the difficulty of making a “full-pricing” system work and, ... underestimated the role of non—price allocation mechanisms within the economy;
3. Underestimated the difficulty of allocating capital;
4. Misjudged the role and function of decentralization and competition; and
5. Simply ignored the role of innovation in the economy.
In these errors, market socialism was not alone, ….: each of these charges could be leveled…against the standard neoclassical model of the economy, …
Market Rationing and Non—Price Allocation Mechanisms within Market Economies
… the market socialism model …. overestimated the role of prices and underestimated the difficulties of making the price system work. It is, of course, not surprising that Lange, Lerner, and Taylor, basing their analyses on the traditional paradigm, …, would similarly stress the role of prices in allocating resources. They differed from the traditional model only in their view of the processes by which prices were to be determined…
Why Prices Cannot Function in the Way Presumed by the Standard Model
… the underlying problems with the “price” model arise from the complexity of the commodity space…The complexity of the commodity space has two fundamental implications. First, it makes it virtually impossible for a central planner to set prices, …. For instance, there would have to be prices for each quality level (a continuum), and each quality level would have to be precisely specified. Moreover, since every commodity has many dimensions, even if there were a limited number of specifications in each direction, the full dimensionality of the product space is enormous. ..
Market-socialist economies … learned the hard way what happens when the product is incompletely specified. If a price is specified for “nails,” short nails made out of any cheap material will be produced. If the length is specified, but not the thickness, then excessively thin nails will be made. If length and thickness are specified, the producer may still make nails out of a cheap material, which may be excessively brittle. For more complex commodities, almost no matter how many characteristics are specified, there remains scope for discretion— and, in particular, cost cutting—which adversely affects how well the commodity performs the task for which it is intended..
The market-socialist model—and the neoclassical model—both fail to recognize the importance of the interface between producers and those who used the products …..The central message of these models, that communication between the two could be limited to price signals, is fundamentally wrong. The process of production is often more one of “negotiation” than of “price-taking.” Firms negotiate delivery times and product characteristics, as well as price. Information (about the needs of the buyers, the technological capabilities of the sellers, etc.) is transmitted in the process. …
The second important aspect of the complexity of the product space is that markets are frequently….imperfectly competitive. …
Imperfect Information and the Limited Role of Prices
… A second explanation has to do with the costs of observing differences in the commodities; that is, even if we could costlessly specify all the relevant characteristics, ascertaining whether a particular item does or does not have those characteristics is expensive. In such a situation, prices may affect the average quality of what one in fact obtains in a market transaction. … causes and consequences of the dependence of quality on price (Stiglitz, 1987a). Perhaps the most important consequence is the “repeal” of the law of supply and demand. …
Three things are important about these non-market-clearing results ..
1…. importance of … screening and providing incentives, which were almost totally omitted from the traditional paradigm.
2.When markets do not clear or… nonprice mechanisms are generally employed to help allocate resources.
3.When markets do not clear, prices do not necessarily convey the kinds of signals concerning scarcity …Information about scarcity may be conveyed in ways other than through prices; firms respond, for instance, to signals such as “orders” and “changes in inventories.”
Nonpricing Mechanisms in Resource Allocation
The importance of the nonpricing mechanisms in resource allocation …. First, a large fraction of all production occurs within firms, within a context in which there is only limited reliance on pricing… relations. … Capital is not allocated by an auction market, with those who are willing to bid the highest getting the capital. …. In allocating capital, it is important to know not only what the user “promises” but what is actually likely to be repaid. ….
The fact that the price system is limited implies that economic relations are frequently governed by both contracts and reputation. Which are totally ignored in the Arrow—Debreu model and in the models of market socialism, ….. Consumers, for instance, rely heavily on reputations in choosing products. …. The buyer (and the producer) rely on reputations. …., lenders rely heavily on the reputation of the borrower.
Contracts almost always involve nonprice terms. Credit contracts, for instance, often have provisions for collateral. The earliest principal—agent literature stressed the importance of nonprice terms, ….Insurance and employment contracts often have “exclusivity” provisions: … for example, the insured agrees to install fire extinguishers. Thus contracts are required because “markets” simply do not exist for all the possible commodities …
Rents and the Reputation Mechanism
…. reputation mechanisms require a modification of how we view the pricing system as working. …. if the maintenance of reputation is to provide an incentive, there must be a cost to losing one’s reputation. …. Economic relations must entail rents, payments in excess of the minimum necessary to induce an individual to be willing to engage in the transaction. Profits, in the conventional sense, cannot be driven to zero; price cannot equal marginal costs. The basic pricing relations underlying the “theory of value” are in that sense wrong, and the corresponding model of market socialism based on the standard theory of value, must on that account be incorrect.
Underestimation of the Difficulties of Allocating Capital
In the traditional view of market socialism, markets were used to allocate goods—given the capital stock—but capital was not allocated by a market system. The failure of markets to allocate capital efficiently, including the failure to coordinate investment decisions, provided part of the rationale for The turn to market socialism.
Market socialists were correct here in identifying a market failure. ... although they did not fully grasp the problems that arose from the absence of a complete set of prices, they thought they understood the consequences of the more obvious absence of a complete set of futures and risk markets. But while the Arrow—Debreu model simply ignored this problem, the market socialists were naive in believing that the government could easily remedy this market failure. ..
But the very reasons—largely information—that lead markets to have problems with allocating capital … all posed serious problems for alternative allocative mechanisms. The absence of futures markets means that firms have to estimate future prices of what they sell and of the inputs that they purchase; so too must the government estimate the shadow prices of goods and services. But what are the incentives provided that those estimates be accurate, that all relevant factors be taken into account? In the market, those who make a mistake are (in theory) disciplined, and they bear a large part of the costs of those mistakes. This is not so in the case of publicly owned enterprises, particularly if they face soft budget constraints….
Thus capital markets cannot be well described as auction markets. Capital is not simply given to the highest bidder. ...The fact that both markets and market socialism engage in direct allocative mechanisms does not mean, of course, that they are identical. The incentives of banks and planners may differ. …
There is another implication, … lenders employ a variety of nonprice provisions in the loan contract in an attempt both to sort among loan applicants better and to provide better incentives. Loan contracts are not described simply by the interest rate….A major failure of market socialism was the absence of both the incentive structure and the discipline mechanism.
Moreover, the introduction of nonfinancial concerns into the social objectives of choosing investments made the task of ensuring that good decisions were being made all the more difficult.
Markets, Market Socialism, and Models of the Market Economy
Asymmetries of information mean that capital markets are characterized by credit rationing and equity rationing. Equity markets—which have important risk-spreading advantages over loan markets—account for a relatively small fraction of new capital raised in almost all countries. …
The competitive market paradigm has exercised an enormous influence over our thinking about how the economy functions. It provides some valuable insights on the importance of competition, the role of prices, the interdependence of markets, and the potential for decentralization. But most of these insights are incomplete: though competition is important, it is not well described by the kind of price competition of the Arrow—Debreu model; prices are only one part of the market resource allocation mechanism; the interdependence of markets operates not only through prices but also through credit; and decentralization is limited, as much production, even under capitalism, operates within large organizations that make limited use of price systems for the internal allocation of resources.
The ideas and ideals of market socialism have held great sway for more than half a century. These ideas, in turn, were greatly influenced by ideas of how market economies function—ideas of Smith, Walras, Arrow, and Debreu. Research in the past twenty-five years has shown how badly flawed both sets of ideas are.
We need not speak of the economic miseries which may have followed from the too-ardent pursuit of the policies based on either set of models. We have seen how, by focusing on a particular set of information problems, and a particular mechanism (the price system) by which information is transmitted, they have ignored a wider range of information problems and a wider range of mechanisms by which information is exchanged. These models have diverted attention, both of economists and of governments, from attaining a more balanced view of the role of the government. it is perhaps no accident that in the major successes in economic growth during the past quarter-century – the “Asian miracle” neither neoclassical economists nor market socialist ideas have played important role.