Last set of comments on Tomasi. Today: ideal theory, and the argument for social justice.
Justice in Fairyland
When Tomasi was writing Free Market Fairness, I predicted it would have the effect of making left-liberals more skeptical of the value of “ideal theory”. They would see that it’s relatively easy for institutions to satisfy the demands of justice under ideal conditions, and so begin to think that ideal theory doesn’t tell us much about institutions. So far, based on conversations I’ve had with others, plus reactions I witnessed as Liberty Fund event on John’s book, my prediction seems accurate.
Rawls says there are four basic questions we might ask when theorizing about constitutions:
- Are the regime’s institutions right and just?
- Could a regime’s institutions be effectively designed to realize its declared goals?
- Would citizens comply with the regime’s institutions and with whatever rules apply to them?
- Would citizens be competent to play whatever role they hold?[i]
In Justice as Fairness: A Restatement, Rawls evaluates five types of regimes. He concludes that that only two of these five regimes could, in principle, realize justice. However, when making this argument, Rawls explicitly says he puts aside questions B through D, and focuses exclusively on question A. For instance, Rawls directs us to ignore so-called “conservative” criticisms of the welfare state that are based on the idea that the welfare state might be abused or might not work as intended. Ideal theory, he says, involves asking, “what kind of regime…would be right and just, could it be effectively and workably maintained?”[ii] In ideal theory, we get to pretend there is no corruption.
The best way to characterize ideal theory is disputed. Still, on one prominent interpretation—an interpreted suggested by the fact that Rawls explicitly tells us to put questions B through D aside—ideal theory concerns asking which constitution would best realize justice under these four conditions:
- People have as strong a sense of justice as humanly possible. They might not be perfect, and they might experience strains of commitment if social institutions demand too much of them. However, everyone is as good as real human being could be, if not as good as an angel.
- Everyone is competent to play his or her role.
- The basic institutions of society achieve their announced public aims and ideals.[iii]
- There are favorable background conditions.[iv]
The first three conditions are highly idealized. Notice that ideal theory involves not only idealizing human motivation, but also idealizing certain human abilities. (After all, we are to put aside questions of competence.) In ideal theory, we are to imagine that people are competent to play their role in society, and that if a constitution is intended to produce some end, then it will in fact produce that end.[v] This means, for example, that any economic planners and bureaucracies have all the information they need, always act on that information, always produce their intended consequences, and never produce negative unintended consequences. There are some constraints, though: In Rawlsian ideal theory, we are not supposed to imagine that human beings have superpowers, or are stronger, longer-lived, or innately healthier than they are.
In ideal theory, constitutions are assumed to achieve their announced public aim. Since Rawls sets aside question B, this seems to suggest that in ideal theory, we are allowed to imagine that any institutions could produce any announced aim, regardless of how unrealistic it would be for them to do so.
However, at least in some passages, it seems Rawls intends for ideal theory to include some restrictions on whether regimes may be imagined to achieve their goals. For instance, (following David Estlund and Tomasi) imagine a regime-type called Chicken Bonedom. Chicken Bonedoms attempt to realize justice as fairness by burying chicken bones in the backyard of their capitol buildings. They have no other institutions intended to realize justice. Despite putting aside question B, Rawls means to exclude this kind of regime as excessively unrealistic. Instead, at the level of ideal theory, for a regime-type to realize justice, it must have institutional arrangements that do not “blatantly violate the laws of common-sense political sociology”.[vi] For institutional arrangements to count as just, it must be sociologically realistic that such institutions could realize justice, if only people behaved well and acted competently.
As far as I can tell, this so-called sociological realism is a weak constraint. For a system to count as sociologically realistic, it need not accord with everything in the social sciences. So, for instance, though economics tells us that real people are too incompetent to centrally plan a large-scale economy, this does not imply that centrally planned command economies are sociologically unrealistic. Sociological realism just requires that institutional arrangements could in principle realize their intended goals, if people were competent and just. Chicken Bonedom could not in principle realize justice, but a command economy could, if only people were more competent.
I’ve often been worried that when Rawls theorizes about institutions, he ends up doing a lot of cheating. For instance, he seems to me to be comparing idealized forms of his favored regimes—liberal socialism and property-owning democracy—to less than ideal versions of his disfavored regimes.
Rawls rejects state socialism. Now, he defines state socialism as necessarily infringing political liberty and freedom of occupational choice. But if we’re allowed to imagine away immorality and incompetence, then surely there’s a form of state socialism that could work and be consistent with Rawls’s theory. Imagine the state socialist regime has elections every five years to determine which party gets to do the central planning. Then imagine the central planners issues job recommendations to each citizen. (Dear Jason Brennan: The Central Committee has determined that we need you to farm. Please?) Each citizen, having a perfect sense of justice, then goes along with the plan voluntarily, because she wants to realize justice. Freedom of occupation is preserved, because everyone can say no, though no one says no.
Now consider one of Rawls’s arguments for equalizing wealth more than the DP would otherwise require. According to Rawls and others, if we allow great concentrations of wealth, then the rich might capture the democratic process for their own ends. Thus, we need more equality, as well as public financing of campaigns, and so on. One problem with this is that it’s anti-empirical. In fact, there’s a great literature on this issue in political science, and it turns out that money doesn’t matter in politics that way Rawls assumes it does.[vii] But the other problem is that Rawls isn’t allowed to make this argument, given his ground rules. We are supposed to assume that people are immune to public choice-type problems, such as corruption and moral hazard. Just as Rawls tosses aside “conservative” critiques of the welfare state based on corruption and moral hazard, so we must toss aside “left-liberal” critiques of campaign finance. What happens if people acquire so much wealth that they could use it to abuse the democratic process? Under ideal conditions, by hypothesis, nothing bad! Only good stuff happens. By hypothesis, they have nearly a perfect sense of justice. Thus, I’m not sure why Tomasi is at great pains to deal with this issue in FMF. If we’re playing by Rawls’s rules, it’s easy.
Similarly, Tomasi deals at length with questions of how his favored capitalist regimes will provide sufficient welfare for the poor, if they need it. But isn’t that an easy problem? Just as Rawls can just decide to imagine that government programs work exactly as intended, without problems of incompetence or moral hazard, so Tomasi can imagine that the market, private charity, (and/or whatever more minimal government social insurance he favors) will work exactly as intended. Are the poor getting too poor? The wealthy, having a nearly perfect sense of justice, will make sure they get what they need. Does justice require the poor to have moral capital, not just more money? The rich, having a nearly perfect sense of justice, will just give the poor the capital they need. No problem. It’s easy.
I often wonder why, at the level of ideal theory, we don’t just advocate some form of anarchism. Curiously, Rawls cites public choice theorists James Buchanan and Mancur Olson in support of his public goods argument for government.[viii] Yet Buchanan and Olson do not do ideal theory. Instead, Buchanan and Olson’s arguments for government rely on what they believe to be a realistic description of human motivation: citizens are predominantly self-interested, rather than overwhelmingly motivated by a sense of justice. Yet when Rawls defends property-owning democracy under ideal conditions, he explicitly puts aside Buchanan’s and Olson’s assumptions, as well as their corresponding worries that property-owning democracy would encourage rent-seeking, corruption, and moral hazard.
Rawls appears to be inconsistent here. He relies upon public choice theory and its take on human motivation and competence to generate a need for government, but then rejects this theory when discussing how government should operate. That is cheating. He cannot have it both ways.
Perhaps Rawls intends the sense of justice in ideal conditions to be strong enough that citizens are immune to moral hazard and the temptations to seek rents and abuse power, but not so strong that citizens would voluntarily contribute money to the roads or defense. But this seems implausible. It takes a weaker sense of justice to want to contribute to public goods than to be immune to the desire to abuse power.
Perhaps Rawls recognizes that the public goods argument for government fails under ideal conditions. He says,
The sense of justice leads us to promote just schemes and to do our share in them when we believe that others, or sufficiently many of them, will do theirs. But in normal circumstances a reasonable assurance in this regard can only be given if there is a binding rule effectively enforced…Many of the traditional activities of government…can be accounted for in this way.[ix]
Perhaps Rawls intends the phrase “normal circumstances” to invoke non-ideal conditions. After all, what better assurance could there be that other citizens will do their part than 1) that they are perfectly competent, 2) that institutions will achieve their intended goals, and 3) that they all have a nearly perfect sense of justice, a sense of justice so strong that they are never tempted to seek special favors from government or abuse power, and would not dare take advantage of others’ charity? Yet, if Rawls recognizes that in ideal conditions, the public goods argument for government fails, then it us unclear how he could advocate democratic government instead of anarchist alternatives.
Since Tomasi adopts the Rawlsian framework, he also adopts all these problems as well. What institutions realize justice as fairness in ideal conditions? Isn’t the more interesting question what institutions don’t?
Making Cold Hearts Bleed
Again, I see Tomasi’s strongest arguments as gestural rather than explicit. He wants to argue that libertarians can and should embrace social justice. (I agree with this conclusion, for what it’s worth.) Tomasi asks libertarians, “Hey, if you could embrace social justice without giving up your cherished ideals of liberty, wouldn’t you want to?” I suspect many thoughtful libertarians will answer yes, in part because I’ve asked many thoughtful libertarians this question, and many say yes. But beyond that, does Tomasi have a real argument? I’m not sure. It seems he does two things:
- Try to show that many classical liberals and libertarians—even Ayn Rand!—have what appears to be an implicit foundational commitment to something like social justice.
- Show that Hayek’s claim that social justice is incoherent is based upon a misunderstanding.
If I had to reconstruct Tomasi’s argument, or what I think it is, it’s something like this:
- The liberal principle of legitimacy: Coercive institutions are legitimate only if they (can be/must be?) accepted by all reasonable people subject to those institutions.
- Property rights, etc., are coercive.
- For some institution X to be acceptable to all reasonable people, everyone must have a sufficient stake in X.
- Social justice: Therefore, for property rights to be legitimate, everyone must have a sufficient stake in them.
Questions for Tomasi:
- Is this what you have in mind?
- How important is the issue of coercion here? Imagine what we might call “pacifist anarcho-capitalism”. Suppose a bunch of people live by a Rothbardian-style market. They possess objects and trade them on the market. Some are rich and some are poor. However, suppose they absolutely refuse to use coercion to enforce any of their claims to their stuff. Everyone regards one another as having something similar to an ownership right to their stuff, but they also reject using violence to enforce these claims, no matter what. In this society, since there’s no coercion, does that mean the issue of social justice doesn’t apply? (If social justice is a precondition of legitimacy, it seems that claims of social justice cannot arise in a pacifist society.)
[i] John Rawls, Justice as Fairness: A Restatement (Cambridge, MA: Harvard University Press), 136.
[ii] John Rawls, Justice as Fairness: A Restatement (Cambridge, MA: Harvard University Press), 137.
[iii] Rawls 2001, 136-37. See also Law of Peoples p. 12.
[iv] For example, there is not too much scarcity to make society impossible. However, conditions are not so favorable so as to make society unnecessary (as it might be if there were no scarcity at all).
[v] Rawls 2001, 137, §41.3.
[vi] John Tomasi, Market Democracy, chapter 3. Tomasi means to be summarizing John Rawls, The Law of Peoples (Cambridge, MA: Harvard University Press, 1999), 12-13; Rawls, Justice as Fairness, 4; John Rawls, Political Liberalism (New York: Columbia University Press, 1996), 281-82.
[vii] E.g., Ansolabehere, Figuieiredo, Snyder, “Why is There So Little Money in United States Politics?” JEP; Robert Hall and Allen Deardroff, “Lobby as Legislature Subsidy”, APSR; Baumgartner, Barry, et al, Lobby and Policy Change: Who Loses, Who Wins, and Why; Nathaniel Persily and Kelli Lammie, Perceptions of Corruption and Campaign Finance: When Public Opinion Determines Constitutional Law, 153 UNIVERSITY OF PENNSYLVANIA LAW REVIEW 119 (2004)
[viii] Rawls 1971, 267-8.
[ix] Rawls 1971, 267-8. Italics mine. Note that Rawls cites William Baumol in defense of this last sentence. But Baumol, like Buchanan and Olson, is not working at the level of ideal theory, and instead assumes citizens are largely selfish.