Game theory has proved a useful tool in the study of simple economic models. However, numerous foundational issues remain unresolved. The situation is particularly confusing in respect of the non-cooperative analysis of games with some dynamic structure in which the choice of one move or another during the play of the game may convey valuable information to the other players. Without pausing for breath, it is easy to name at least 10 rival equilibrium notions for which a serious case can be made that here is the “right” solution concept for such games.

In 1970 Amartya Sen exposed an apparent antinomy that has come to be known as the Paradox of the Paretian Libertarian (Sen, 1970b, pp. 152–57). Sen introduced his paradox by establishing a simple but startling theorem. Roughly put, what he proved was that if a mechanism for selecting social choice functions satisfies two standard adequacy conditions, there are possible situations in which it will violate either the very weak libertarian precept that every individual has at least some rights or the seemingly innocuous Paretian principle that an option should be judged unacceptable if there is an available alternative that everyone prefers to it. Many economists and philosophers have proposed solutions to Sen's problem, but there is no general consensus on what solution (if any) is correct. In the present paper I argue that Sen's original theorem fails to establish the existence of any conflict between libertarianism and Paretianism. Furthermore, I contend that Sen has misinterpreted certain other theorems which he has used to defend the existence of a paradoxical conflict between these two doctrines. In general, I try to show that whenever Sen posits a Paretian-libertarian conflict to explain an apparently troubling result in social choice theory, the difficulty can be better dealt with either by claiming that the theorem in question imposes overly strong background constraints on the form of social choice functions or by claiming that it relies on an unacceptable construal of individual rights.

Radical and liberal theories of egalitarianism are distinguished, in large part, by the differing degrees to which they hold people responsible for their own well-being. The most liberal or individualistic theory calls for equality of opportunity. Once such “starting gate equality,” as Dworkin (1981a) calls it, is guaranteed, then any final outcome is justified, provided certain rules, such as voluntary trading, are observed. At the other pole, the most radical egalitarianism calls for equality of welfare (assuming that interpersonal welfare comparisons can be made, so that such equality makes sense). In between these two extremes are egalitarian proposals that equalize more than conventional opportunities, yet less than full welfare. Sen (1980) speaks of equality of basic capabilities as a goal; implementing that requires more than starting gate equality, because some will require more resources than others to attain the same capabilities. Meeting basic needs is another objective. Equality of needs fulfillment is perhaps less radical than equality of basic capabilities and more radical than equality of opportunity. Rawls (1971) takes equality of primary goods as a benchmark; he distinguishes primary goods from welfare, but includes among them goods that are more complicated than conventional resources and opportunities, all of which are supposed inputs into any conception of welfare. One could imagine proposing an egalitarianism that equalized some quite measurable outcome across populations, such as infant mortality. That would be an outcome-equalizing theory where the rate of infant mortality is a proxy, presumably, for some more complicated maximand, such as the degree of wellbeing of a population.

One feature of socialism that has been little discussed in the recent revival of interest in Marx is the basic form of economic organization that will characterize such a society. Marx's view, to be documented in what follows, is that socialism would not have a market economy. This prediction should be a matter of some embarrassment or consternation to twentieth-century socialists outside of the Soviet bloc who claim a Marxist heritage. Despite the fact (or perhaps because of the fact) that some socialist regimes in the first half of this century did carry out a program of (largely) abolishing the market, the abolition of markets is no longer on the agenda for developed and developing countries. On the Left, anyway, some form of market socialism seems to be the preferred alternative; Yugoslavia is often pointed to as an imperfect real world exemplar of what market socialism might be like.

I want in this paper to do two things. First, I want to respond to some studies that argue that people are often not rational: that people regularly and systematically depart from rationality. The conclusion itself does not worry me. I pressed for the same in a recent book (Schick, 1984). But the arguments seem to me wrong, and wrong in an interesting way. There may be something to be learned from seeing how and why they fail.

There has been a persistent tendency to identify what is called “the freerider problem” in the production of collective (or public) goods with the prisoner's dilemma. However, in this article I want to challenge that identification by presenting an analysis of what are in fact a variety of collective action problems in the production of collective goods. My strategy is not to consult any intuitions about what the free-rider problem is; rather I will be looking at the problematic game-theoretic structures of various situations associated with the production of different types of collective goods, thereby showing what sorts of difficulties a community concerned with their voluntary production would face. I call all of these dilemmas free-rider problems because in all of them certain individuals find it rational to take advantage of others' willingness to contribute to the good in a way that threatens its production. Some readers may feel that the term ‘free-rider problem’ is so identified with the prisoner's dilemma that my extension of the term in this way “jars”; if so, I invite them to coin another word for the larger phenomenon. My aim is not to engage in linguistic analysis but to attempt at least a partial analysis of the complicated structure of collective good production.

In the first half of the 1970s, two books appeared which have subsequently been regarded as major works in political philosophy: John Rawls's A Theory of Justice (1971), and Robert Nozick's Anarchy, State, and Utopia (1974). Economists have devoted a considerable amount of ink to commentary, pro and con, on A Theory of Justice; and it is getting to be a rare public finance textbook that does not, in its discussion of governmental redistribution, describe the Kantian contract made behind the veil of ignorance. On the other hand, while Nozick has not exactly been ignored, economists have not joined the debate over Anarchy, State, and Utopia with the same gusto. When economists have joined the debate, their concern has been, more often than not, with Nozick's entitlement theory of distributive justice, as is the case with Varian (1975) and Sen (1977). What is largely missing, then, is any economic analysis of the processes that give rise to Nozick's morally legitimate state, which he calls the minimal state, and the characteristics and likely activities of the minimal state within the moral boundaries set by Nozick, his assertions to the contrary notwithstanding.

Is rhetoric just a new and trendy way to épater les bourgeois? Unfortunately, I think that the newfound interest of some economists in rhetoric, and particularly Donald McCloskey in his new book and subsequent responses to critics (McCloskey, 1985a, 1985b), gives that impression. After economists have worked so hard for the past five decades to learn their sums, differential calculus, real analysis, and topology, it is a fair bet that one could easily hector them about their woeful ignorance of the conjugation of Latin verbs or Aristotle's Six Elements of Tragedy. Moreover, it has certainly become an academic cliché that economists write as gracefully and felicitously as a hundred monkeys chained to broken typewriters. The fact that economists still trot out Keynes's prose in their defense is itself an index of the inarticulate desperation of an inarticulate profession.

One of the most actively discussed aspects of Keynes's thought during the last decade has been his concern with uncertainty and probability theory. As the concerns of current macroeconomic theorists have turned increasingly to the effects of expectations and uncertainty, interest has grown in the fact that Keynes was the author of A Treatise on Probability (1921) and that uncertainty plays a prominent role in Chapter 12 of The General Theory as well as in three 1937 papers in which he summarized The General Theory's main point. Not surprisingly, though, there has been very little agreement in this recent discussion about exactly what the significance of Keynes's early work in probability was to his later work as an economist, or about what the roles of uncertainty and expectations are in The General Theory.

Why do people expend resources to vote in large-number situations where the probability of their affecting the outcome is close to zero? In a recent article, Geoffrey Brennan and Loren Lomasky (1985, p. 198) argue provocatively that Adam Smith's The Theory of Moral Sentiments (TMS) not only predicts such behavior, but further predicts that people “frequently” vote for outcomes that cost them more than they would individually be willing to pay. In other words, in the relevant environment, they claim that individuals will systematically express false preferences for costly outcomes they “really” do not want. I submit that a fair reading of TMS in fact provides no basis for their view; if anything, TMS suggests the opposite.

Scott Arnold's recent paper, “Marx and Market Socialism,” advances a provocative thesis: market socialists are advocating an economic system that has a strong, internally generated tendency to revert to capitalism. They are, in short, “capitalist roaders” (Arnold, 1987).

David Schweickart has challenged a number of claims that are central to my argument that market socialism would probably degenerate into something only nominally distinguishable from capitalism. Chief among these is the claim that competitive pressures would force the workers in a worker-controlled firm to create pay and authority differentials that would make such firms structurally homologous to capitalist firms. Schweickart challenges this on two fronts: He argues that there is no good reason to believe that market forces under market socialism would create the pay and authority differentials characteristic of capitalism. He further argues that certain structural features of market socialism would insure that competition would not be as intense as it is under capitalism. Consequently, even if capitalistically structured firms were more efficient, it would not make much difference, since no sword of Damocles would hang over the heads of those firms whose workers prefer more collectivist methods of control. Let us consider each of these points in turn.

In their article, Geoffrey Brennan and Loren Lomasky (1985, p. 199) present an alternative to market theories of voting behavior. Contrary to market theories which view the voter as acting to maximize the expected self-interest, the alternative view sees voting as fundamentally an act of self-expression: “Voting, like speech, is an expressive activity providing an outlet for one's moral sentiments. We suggest that it is the expressive return to a vote that frequently determines the behavior of individuals in large-number electorates.”

How may we best understand the motivational structure that stands behind individuals' acts of voting? In “The Impartial Spectator Goes to Washington” we suggested that expressive concerns swamp narrowly consequential motivations, in contradistinction to normal market transactions in which the priority is reversed. A striking consequence of this fact is that individuals will be led to vote for outcomes that they would reject were they in a position to act decisively. In this regard we found the moral psychology Adam Smith develops in The Theory of Moral Sentiments (TMS) (and, to a lesser extent, in The Wealth of Nations) remarkably fecund in suggesting alternatives to what we call the standard theory of electoral behavior.

Professor Arnold's reply to my reply seems not to have touched the substance of my argument. Perhaps I have been unclear. Arnold contends that any form of market socialism, if unchecked by central authorities, would revert to a system essentially undistinguishable from capitalism. Against this contention I have argued that a democratic, worker-controlled, market socialism that generates its investment fund by taxation exhibits no such tendency. Specifically, I argued that in such a society

1. there exists no tendency for socialized property to revert to private property (since workers have no incentive to sell off their firms to private individuals);

2. the ability of private individuals to accumulate vast wealth is sharply curtailed (since interest and stock income are absent);

3. the authority structures within enterprises are much more democratic than under capitalism;

4. intra-firm income differentials will be far more egalitarian than under capitalism.

Rosenberg (1986) argues that economists have embraced the methodology of scientific research programs, and the writings of Imre Lakatos (1978), at the same time that philosophers have been abandoning that approach. According to Rosenberg, the methodology of scientific research programs (MSRP) appears to allow some work in economics, which is neither tested nor testable, to be “scientific” nonetheless. That is, MSRP justifies some current practices which look hard to justify on strict falsificationist, or dogmatic positivist, grounds.

Since Schweickart asserts that I have not addressed his main argument, let me consider briefly the four claims he advances at the beginning of his second reply.

Regarding 1: To argue, as I have, that there would be a strong tendency for market socialism to degenerate into capitalism, it is necessary to spell out carefully what capitalism is. Following Marx, I defined capitalism as a system in which the workers do not control the means of production and the workers sell their labor power as a commodity. As G.A. Cohen has pointed out (Cohen, 1978, pp. 219–23), this definition can be given a rechtsfrei characterization in terms of effective powers over means of production and labor power. Actual legal arrangements are not the issue. In my original paper, I never said that the workers would sell the means of production. Thus 1 is irrelevant. The real questions are: (a) Who effectively controls the means of production? (b) Who effectively gets the profits? If the answer to both of these questions is, “Not the workers,” it follows that the workers are effectively proletarians and the system is a form of capitalism.

Weintraub is not really interested in whether economics is “science” or not. “Economists are not so unsophisticated as to think that calling economics a ‘science’ says anything about what economists do or should do” (1987, p. 140). But can it really be a matter of indifference to him whether the subject has the character of chemistry as opposed to literary criticism?