Editor’s note: Christopher Freiman’s new book, Unequivocal Justice, is available for purchase for those wishing to explore the arguments below on a deeper level.
Classical liberals like Adam Smith, F.A. Hayek, and Milton Friedman and modern egalitarian liberals like John Rawls and Ronald Dworkin disagree about the size and scope of government. Classical liberals want to keep it to a minimum; egalitarian liberals prefer that the state take a more active role in redistributing income and regulating the market.
There’s a growing sense that classical and egalitarian liberals largely agree about the ends of a good political system; they simply disagree about the best means to those ends. For instance, both Smith and Rawls think that a just society will prioritize alleviating poverty, but they disagree about how to alleviate poverty.
Ideal and Nonideal Theory
Why, then, do we see such pervasive disagreement about the best institutional means to shared liberal ends?
One explanation, proposed in various forms by theorists like Thomas Sowell, Steven Pinker, and John Tomasi, is that egalitarian and classical liberals have different views about feasibility. They disagree not only about what sorts of policies are feasible, but also the extent to which political theories should be constrained by feasibility at all.
More specifically, egalitarian liberals tend to do ideal theory, whereas classical liberals tend to do nonideal theory.
Although philosophers interpret ideal theory in different ways, one of the versions put forth by Rawls involves evaluating political regimes on the assumption that everyone in society acts justly. Later, Rawls developed the notion of a regime’s “ideal institutional description,” that is, “the description of how it works when it is working well, that is in accordance with its public aims and principles of design” (Justice as Fairness, 137). He recognized that many of his critics were concerned with nonideal theory and thus “the ineffectiveness of the welfare state and its tendencies toward waste and corruption” (Justice as Fairness, 137).
However, Rawls himself was engaged in a different kind of project. His question about political institutions was this: “What kind of regime and basic structure would be right and just, could it be effectively and workably maintained?” (Justice as Fairness, 137).
The split between ideal and nonideal theory partly explains why egalitarian liberals and classical liberals disagree about policy.
- If we assume that people are fully just and that states always work as designed, it’s natural to conclude that we should endow the state with extensive power to do good.
- Alternatively, if we allow that people will sometimes fall short of justice and that government sometimes fail, we’ll be more skeptical about centralized state power. We might prefer a smaller state with less potential to do damage.
As Hayek says, the concern of Adam Smith and other early classical liberals “was not so much with what man might occasionally achieve when he was at his best but that he should have as little opportunity as possible to do harm when he was at his worst” (“Individualism and Economic Order,” 12).
The Problem with Ideal Theory
Given that that distinction between ideal and nonideal theory informs substantive policy debates, it would be helpful if we could arbitrate in favor of one approach. That’s the aim of my new book, Unequivocal Justice. I argue against ideal theories of the state, not because they are infeasible, but because they are inconsistent. That is, they don’t follow through on their own assumptions about people’s motivations and behavior. A number of theorists have similar concerns, including Will Wilkinson, G.A. Cohen, Jason Brennan, and Jacob Levy. My book develops this worry in detail and suggests that it jeopardizes egalitarian liberal theories of the just state.
I argue that a perfectly just society doesn’t need an active state with a lot of power to do good; rather, it doesn’t need a state at all. As a first pass, consider that perfectly just people wouldn’t rob banks, they wouldn’t drive gas guzzlers, and they’d donate plenty to charity. So they wouldn’t need a state to force them to not steal, to stop polluting excessively, or to contribute to a social safety net. In Cohen’s words, if moral principles “are practiced in everyday life . . . then the state can wither away” (Rescuing Justice and Equality, 2).
It’s only when people are exploitative, short-sighted, selfish, and so on that we need a state. But now we’ve opened up the possibility that the state itself will be run by exploitative, short-sighted and selfish people. Crucially, we’ve lost the right to assume that the state itself will act justly. As I say in the book:
“Ideal theory, then, contains an internal inconsistency. On the one hand, we must assume that people are not fully just to generate a need for the state in the first place. On the other hand, if people are not fully just, the state itself won’t be fully just either (it’s an institution run by people, after all). In short, the assumption that generates a need for the state—that people aren’t fully just—at the same time undermines the assumption that the state is fully just. What this means is that the only coherent theory of the state is a nonideal one.”
The problem for ideal theory can be summarized as a dilemma. If we consistently apply the assumption that everyone is fully just, we cannot explain why the state is needed at all; if we consistently apply that assumption that people are not fully just, then we may not stipulate, as Rawls does, that the state itself is fully just. Rawls and others illicitly resolve this inconsistency by flip-flopping between ideal and nonideal assumptions: injustice is ruled in to generate the need for state intervention and then ruled out to stipulate that the intervention works well.
We can get a better sense of how this dilemma plays out by looking at a couple of cases (the book discusses a number of others).
Rawls argues that one of the state’s roles is to supply public goods. The market won’t provide public goods because of the free-rider problem:
“Where the public is large and includes many individuals, there is a temptation for each person to try to avoid doing his share. This is because whatever one man does his action will not significantly affect the amount produced” (A Theory of Justice, 236).
For example, your individual contribution to clean air is insignificant and you can benefit from the contributions of others even if you don’t contribute yourself. You breathe easier when everyone else drives a Tesla even if you drive an SUV, so you have little incentive to buy the Tesla. But since everyone else has the same incentives, no one will drive a Tesla, and we won’t get clean air. So, Rawls’s argument goes, “The provision of public goods must be arranged for through the political process and not through the market” (A Theory of Justice, 236). The state could implement a carbon tax, for instance.
But how, exactly, will that tax be implemented? By Rawls’s own lights, people won’t contribute to a good when their individual contribution is insignificant and they can free ride off of the contributions of others. The problem is, an individual vote—say for the candidate that supports a carbon tax—is insignificant in most large-scale elections. It won’t make a difference to the outcome. What’s more, you’ll still enjoy the benefits made possible by a carbon tax even if you didn’t vote for it.
Given Rawls’s assumptions, then, you’ll free ride and not vote for the carbon tax. Since everyone else has the same incentive, no one will vote for the carbon tax and it won’t be implemented. Thus, Rawls’s proposal for supplying public goods is undermined by the very problem it was intended to solve. He assumes free-riding to ensure that the state is needed but then must (implicitly) assume away free-riding to ensure that the state works.
This problem reappears throughout Rawls’s work. For instance, he worries “that those with greater wealth and position usually control political life and enact legislation and social policies that advance their interests” (Justice as Fairness, 148). In Rawls’s view, electoral regulations like spending restrictions are needed to correct this problem. Since classical liberal and libertarian regimes don’t include these regulations, they aren’t just.
Rawls’s assumption that the rich control the political process is what generates the need for electoral regulation in the first place. (If the rich restricted their political spending out of a moral commitment to political equality, then there’d be no need for mandatory spending limits.) But if the rich control the political process, how can we expect the political process to produce rules that will control the rich? Shouldn’t we instead expect the rich to control the electoral regulation itself? At a minimum, Rawls is not entitled to his stipulation that the regulation will work as intended. Whether Rawls’s favored policies do a good job of protecting political liberties is an open question.
Maybe empirical inquiry shows that Rawls’s favored policies do indeed work well. However, once we stop stipulating how institutions will work and start conducting empirical inquiry into the actual, unjust world, we enter the realm of nonideal theory.
And I argue that’s exactly what we should do.