The clearest and most concise statement of John Roemer’s project in A Future for Socialism occurs at the end of his book.footnote＊ In its concluding chapter he summarizes his argument as pivoting on two ‘crucial ideas’—the idea that ‘socialism is best thought of as a kind of egalitarianism, not the implementation of a particular property relation’, and the idea that ‘modern capitalism provides us with many fertile possibilities for designing the next wave of socialist experiments’ (pp. 124–5). Associated with these ideas are two other claims, that ‘the failure of the Soviet experiment is ascribable not to the egalitarian goals of communism but to the abrogation of markets’ and that ‘Modern capitalism does not. . .owe its success specifically to the embrace of the right to unlimited accumulation of private property’ (p. 125).

Roemer’s project appears to express two, mutually supportive interests: the political interest he has in developing a reformulation of the socialist project in a form that is feasible and defensible in the historical context arising from the Soviet collapse; and his intellectual interest in the ways in which the recent history of capitalism has shown the connections between market institutions and property relations to be far less determinate, and far more variable, than is allowed for in either standard economic theory or neo-liberal ideology. The upshot of both interests is a political and theoretical displacement and subordination of property relations. Within socialism, they are to be regarded in wholly instrumental fashion, their assessment and reform being governed by their contribution to the achievement of the egalitarian objectives in terms of which the socialist project is best understood. Within capitalism, property relations are recognized to be extremely complex and variable, embodied in diverse legal practices and cultures which create environments in which market competition can occur.

Roemer’s two-pronged argument is that, since market institutions are far more significant in generating the competitive efficiencies and productivity of capitalism than are private property relations, socialism should not be conceived as the project of harnessing market institutions, of embedding them in a variety of property relations, with the aim of promoting important equalities. In Roemer’s account, well-designed, or intelligently reformed, market institutions are functionally indispensable to both successful capitalism and feasible socialism. The distributional goals which define the socialist project are compatible with a variety of property regimes, including ‘social-republican’ institutions—in which private property is constrained in its uses by requirements having to do with active participatory membership and the limitation of inequality—of the sorts found in some capitalist societies. More particularly, Roemer argues for a modified version of market socialism as most likely to achieve socialist distributional goals while maintaining, or enhancing, the efficiencies deriving from market competition. In Roemer’s market socialism, though the capitalist corporation is in its current forms no longer the dominant form of productive enterprise, firms continue to be run on a competitive, profit-maximizing basis without intervention by political authorities. Further, there is no presumption as to any particular pattern of ownership rights for firms in a market socialist economy, with Roemer canvassing a range of possible regimes that encompasses both an economy of labour-managed firms and one in which de jure private property rights are unchanged but bargaining powers have been transformed. Roemer’s pluralistic and radically revisionist model of market socialism is one in which market institutions may be conjoined with any among a large variety of property relations, provided socialist goals of equality are thereby advanced. In proposing this model Roemer sees himself as building on the results of the historic debate between Oskar Lange and Friedrich Hayek, by revising the original market socialist model in response to Hayek’s and later criticisms (such as those of Janos Kornai). Itmay well seem that in modifying the market socialist conception so comprehensively Roemer has removed from it all features that are peculiarly, or even recognizably, socialist. Certainly, in Roemer’s conception, socialism has ceased to signify any definite system of institutions; and it has no essential connection with the interests or needs of any social class. Yet his animating purpose is to give the socialist project another lease on life, in which it can survive the Soviet debacle and—as a corollary of acknowledging the failures of centralized economic planning institutions—be made compatible with the recognition of the functional indispensability of market institutions in modern economies. In this overriding project, however, Roemer signally—if perhaps also inevitably—fails.