In his book Unequivocal Justice, Christopher Freiman contends that liberal egalitarians in general, and John Rawls in particular, are trapped in a dilemma. On the one hand, if society is fully just, possessing both fully just institutions and fully just citizens, then the state is unnecessary: the citizens of such an ideal society would do all that justice required of them without a state. So the kind of ideal theorizing practiced by Rawls is anarchic. On the other hand, if society is not fully just, because its institutions and citizens are not fully just, then the state is not in a position to make it just: the state is staffed by these same imperfectly just citizens and therefore cannot be assumed to be just itself, and unjust states will fail to construct just societies. So the kind of ideal theorizing practiced by Rawls is otiose.

Freiman then accuses Rawls of trying to escape this dilemma “by implicitly violating what Geoffrey Brennan and James Buchanan label ‘behavioral symmetry’—i.e., applying one's model of behavior consistently across institutional types. More specifically, Rawls posits unjust behavior in market and civil society (giving the state a job to do) and fully just behavior in politics (ensuring that the state does its job correctly)” (3; emphasis added). Such behavioral asymmetry makes the standard liberal-egalitarian policy prescriptions—redistribution and regulation to equalize political power and social status and to ensure a decent social minimum and equality of opportunity—look artificially attractive. I say “artificially,” because Freiman argues there is no compelling evidence that people behave differently in politics than they do in market and civil society. If so—if rulers and ruled alike are corrupt—then the kinds of capitalist regimes that Rawls rules out as inconsistent with justice as fairness (namely, laissez-faire capitalism and welfare-state capitalism) may come back into consideration.

Freiman's claim that justice as fairness contains a behavioral asymmetry, however, hinges upon an overdemanding conception of justice that Rawls explicitly, and rightly, rejects. Individuals in Rawls's ideal society merely possess a duty to “support and comply with just institutions … [and] further just arrangements not yet established” (A Theory of Justice [Belknap, 1999], 99). Within these bounds, they are free to pursue their own purposes. In order to claim that “Rawls posits unjust behavior in market and civil society,” Freiman ascribes a much more demanding conception of justice to Rawls, one defended by G. A. Cohen, in which justice requires “an ethos, a structure of response lodged in the motivations that inform everyday life. … There is not, then, on the one hand, an economic structure that is organized to achieve a certain form of justice and, on the other, a set of individual economic choices that need show no respect for that justice” (quoted at 25 and 57).

I would suggest that this more demanding conception fails to coincide with our considered convictions. Imagine, for the moment, a baseball player who neither bends nor breaks the rules of baseball and who, moreover, labors conscientiously to improve these rules over time to maintain the fairness of the game. Within these bounds, however, he plays as competitively as he can, both for his own sake and for the sake of his teammates. Would we describe such a player as “unjust”? Cohen (and presumably Freiman) would, because such a player, when he plays ball, does not play in order to advance fairness; rather, he plays to win, consistent with the rules. Most people would, I think, find such an assessment counterintuitive: we may not admire him for his cold but fair drive for victory, but we surely would not call him unjust. The same judgment should apply to the highly talented software engineer who bargains hard for a higher salary and gives nothing to charity, but who also obeys the laws, pays his redistributive taxes, and votes for redistributive policies. Would we find him admirable? Surely not. But unjust? Again, surely not. And the same judgment should apply to the highly talented antitrust lawyer for the Justice Department who bargains hard for pay and advancement and refuses to give up weekends to fight monopoly, but who also obeys the laws, discharges his contractual and civic duties, and votes accordingly. There is just no asymmetry here. Cohen would call these people schizophrenic. I would call them human—but Homo sapiens rather than Homo socialismus. To recognize that a fully just human being—one who obeys, supports, and even works to reform rules of justice—can still vigorously pursue his own purposes without blame is to recognize our divided nature, our moral-cum-prudential selves. To fail to do this is to dabble in utopianism.

The rules that Rawls's fully just agent follows concern all and affect all; consequently, they should be endorsed by all. This egalitarian insight explains one function of the state, even in ideal theory: democratic deliberation about, and endorsement of, the rules of justice. But there is another function of such a state. Return to the baseball example. Suppose that a team, swayed by Cohen's arguments, collectively decided to throw a game to an opponent who had had a bad season through no fault of their own and therefore (let us suppose) inconsistently with the spirit if not the letter of fair baseball. How could they know that this would advance fairness? Would this not depend upon, among other things, the precise reasons for their opponent's bad season, what other teams did, and how (badly) they threw the game? Teams, and individuals on teams, are generally not in a position to make such synoptic judgments, which is why justice cannot reasonably demand that they do so. Rather, justice only requires that they assess the evidence together, deliberate, and reform the rules over time to make the game fairer. This colegislative role is separate from, if superior to, their role as players and teammates; it superintends, not subsumes, their self-seeking competition. This, then, is another function of the state: a democratic coordination of individuals and groups within society in the interest of justice via the rules of justice—an epistemic division of labor, if you will, between state and society. (Interestingly, Freiman appears to recognize this at one point in the text [56] but does not see its full implications.) There is just no anarchy here.

If, as I have argued, Rawls's ideal theory is neither behaviorally asymmetric nor anarchic, does this imply that Freiman's conclusions about capitalism are also wrong? Perhaps surprisingly, no. The same charge of utopianism that I leveled at Cohen might be leveled against Rawls: if there is no realistic prospect of moving from the partial-compliance world we inhabit, where rulers and ruled are symmetrically corrupt, to the full-compliance world delineated by Rawls, then nonideal theory is indeed the only game in town. If so, then as Freiman repeatedly points out (e.g., 35, 153), we are thrown back on comparative institutional analysis, which might very well cast a better light on laissez-faire and welfare-state capitalisms. Unfortunately, Freiman also points out that this vital task is not his book's. That leaves me hoping for a sequel.