Article body

I’m delighted to have the opportunity to comment on Alan Patten’s great book, which offers the most philosophically nuanced and sophisticated treatment of minority rights of which I’m aware.[1] I’m also very sympathetic to the central claim Patten makes—namely, that liberal neutrality offers the best normative framework to adjudicate questions of cultural and religious justice.

That said, I’m going to focus on an issue on which I’m not entirely persuaded by what Patten has to say. He claims his account of minority rights is broadly continuous with Ronald Dworkin’s resourcist conception of justice, and, like Dworkin, Patten denies that his account sanctions compensation for expensive tastes. I challenge this claim. I argue that, contra Patten, Dworkin’s theory does not provide a basis to offer accommodations or minority rights, as a matter of justice, to some citizens who find themselves at a relative disadvantage in pursuing their plans of life after voluntarily changing their cultural or religious commitments.


Patten’s account of minority rights draws on some ideas at the heart of liberal egalitarian theories of justice. First, he argues that the state has an obligation to treat citizens equally—in particular, to be equally responsive to the interests of all citizens, and not to single out some groups for favourable treatment. Second, he claims that we each have an important interest in self-determination, and so the state has a pro tanto reason to ensure that each person has a fair opportunity for self-determination—that is, to develop, revise, and pursue her conception of the good. From these ideas, Patten derives a principle of neutrality:

Neutrality of Treatment: The state violates this requirement when, relative to an appropriate baseline, its policies are more accommodating of some conceptions of the good than others.[2]

When the state violates this principle, it fails to treat citizens equally because it gives some citizens a greater opportunity for self-determination than others. If, for example, the state uses resources to subsidize the promotion of your conception of the good rather than mine, the state is not providing each of us with a fair opportunity to pursue our preferred plans of life; rather, it’s helping you at my expense.

This idea of neutrality provides the moral foundations for minority rights in cases where the state is unavoidably entangled in promoting some ways of life rather than others. For example, the state cannot reasonably avoid conducting official business in some languages rather than others, or having some official holidays and not others; and perhaps also political boundaries and jurisdictions are drawn in ways that unavoidably favour some groups rather than others. When the state cannot avoid providing this sort of support to some languages or cultural groups, it must also provide prorated forms of recognition or cultural accommodation to other linguistic or cultural groups to avoid violating neutrality—to ensure that each person is given the same fair opportunity to pursue his or her own conception of the good.

Patten says that the purest way for the state to realize neutrality of treatment is via the strategy of privatization—that is, the strategy where the state entirely avoids regulating or providing cultural goods, and where the provision and pricing of these goods is left entirely to an idealized market process. As he puts it, “the only way to achieve neutrality of treatment perfectly is through privatization. Leave people with all-purpose resources to spend and let them spend those resources in the way that best reflects their conceptions of the good”.[3] Of course, he concedes that sometimes market failures or historical injustices give us reasons to depart from privatization, but it’s clear that Patten thinks that giving each individual his or her fair share of resources to spend as he or she sees fit is the optimal way to realize neutrality.

In holding this view, Patten follows Ronald Dworkin’s account of equality of resources and he explicitly presents his theory as continuous with Dworkin’s.[4] Dworkin’s view requires that each person be provided with an equal share of resources to develop and pursue her own plan of life, consistent with the moral rights of others. The aim is not to ensure that each person is equally successful in the pursuit of his or her plans. Some people may find their conceptions of the good difficult or even impossible to pursue successfully because doing so requires more than an equal share of resources. But so long as no person can complain that she has been given fewer resources or rights to pursue her preferred conception of the good, then each has been given a fair opportunity.

On Dworkin’s account, if people have equal natural talents, abilities, and equal external resources, and face an otherwise fair set of background conditions, then we can hold each person responsible for her preferences and choices. If, at t1, Anna chooses to spend her money on parties and holidays, whereas Betty chooses to save her resources for a rainy day, then Anna cannot complain at t2 when she has fewer resources. She was given the same resources as Betty. She simply chose to spend those resources in a particular manner, one that valued more immediate consumption over long-term saving. Similarly, Carl has no legitimate complaint if he cannot pursue his very expensive conception of the good as successfully as Daniel’s much less expensive conception of the good. As Patten emphasizes at several points in the book, individuals should be held responsible for their preferences and conceptions of the good when they face fair background conditions—that is, when each citizen has been provided with her fair share of resources, and the state has not violated neutrality of treatment.[5] Given a fair distribution of resources, Carl should have known that his plan of travelling the world, never having a job, and living in a mansion was doomed to be frustrated since it requires more than a fair share of resources. He should have adjusted his conception of the good to take account of what he could reasonably afford.


But now consider an example that Patten presents in chapter five. In the imagined society, everyone worships either on Sunday or not at all. To accommodate this fact, schools and government offices close on Sundays and Saturdays. Over time, however, some people freely convert to a new religion, and this new religion’s day of worship is Tuesday.[6] The adherents of this new religion now demand accommodation—for example, requesting that schools and government buildings be closed on Tuesdays and Sundays rather than on Saturdays and Sundays.[7]

Patten believes there is a pro tanto reason of justice for the state to accommodate the new Tuesday Worshippers. He says:

It is possible to mimic the justice of the market by favoring formatting solutions that are responsive to the cultural attachments of different citizens as well as to the costs of providing the various formats (hence the importance of prorating recognition). If these attachments change over time—as with the emergence of the Tuesday worshippers—it is reasonable to expect that the formatting solutions expected by justice would eventually have to change with them.[8]

But there’s an obvious objection to this conclusion: shouldn’t the preferences of the Tuesday Worshippers be treated like an expensive taste, like Carl’s expensive taste for world travel and a mansion? Since the Tuesday Worshippers are responsible for choosing a religion that has requirements that are “expensive” to satisfy given the existing social rules, shouldn’t they bear the costs for that choice rather than demanding society accommodate their expensive choice?[9]

Patten denies that the preferences of the Tuesday Worshippers constitute an expensive taste. He argues that tastes or preferences are only expensive in the relevant sense—that is, in the sense that we expect the individual to bear the costs rather than being able to claim some accommodation—when the preference is formed in the context of fair background conditions. As he puts it, “my account does leave room for the rejection of preference-based demands for accommodation on responsibility or expensive taste grounds, but only when those demands are made in a context where a fair distributive scheme is in place”.[10] To illustrate, Patten offers an example where a country establishes Christianity as the official religion, and Muslim citizens complain that this violates neutrality of treatment. Patten rightly says that we cannot deny the complaint of the Muslim citizens by dismissing their religious preferences as an expensive taste since these citizens don’t face fair background conditions—the state fails to meet the neutrality of treatment requirement and thus fails to provide each citizen with a fair opportunity to pursue his or her conception of the good. Put differently, it’s implausible to say that Muslim citizens must pick up the tab for their “expensive” taste when it’s the non-neutral treatment by the state that makes their religious preferences “expensive.” For the same reason, Patten argues, it’s unfair to hold the Tuesday Worshippers substantively responsible for their “expensive” religious beliefs.

Here is a reconstruction of Patten’s argument, as I understand it:

  • P1 A person can only be held responsible for the expensiveness of her conception of the good when the conception develops in a context where a fair distributive scheme is in place.

  • P2 A necessary feature of fair distributive scheme is neutrality of treatment.

  • P3 If the state fails to offer accommodation to the Tuesday Worshippers, it violates neutrality of treatment.

  • Therefore,

    C1 The Tuesday Worshippers cannot be held responsible for the expensiveness of their religious beliefs.

It might seem that we can easily reject P3 by pointing out that the case of the Tuesday Worshippers is disanalogous to the case where the state makes Christianity the official religion despite the existence of a Muslim population. In the latter case, the state violates neutrality of treatment by privileging some citizens’ conception of the good over a conception of the good endorsed by other citizens. This is the legitimate basis of the Muslim citizens’ complaint. In the case of the Tuesday Worshippers, by contrast, the state does not violate neutrality of treatment in initially designating Saturdays and Sundays as the official holidays since this doesn’t disadvantage anyone. And if the initial decision is consistent with neutrality of treatment, then the subsequent decision of some people to become Tuesday Worshippers cannot render the existing framework a violation of neutrality of treatment. Shouldn’t a resourcist hold people responsible for the choices they make in the context of a fair distributive scheme, rather than calling for the scheme to be altered when people make new plans? If so, then P3 is clearly false.

But Patten disagrees. He argues that a fair distributive scheme, even for a resourcist like Dworkin, must be sensitive to people’s existing plans and preferences. To support this claim, Patten draws our attention to a particular feature of Dworkin’s model for realizing an equal distribution of external resources amongst a group of immigrants who wash ashore on a previously uninhabited island.[11] As we know, the distribution of the island’s resources must pass the envy test. That is, when the distribution is concluded, it must be the case that no immigrant would prefer some other immigrant’s bundle of resources over her own. Of course one way to guarantee that the envy test is met would be to magically converts all of the island’s various resources into identical bundles of some particular good or goods. In Dworkin’s example, all of the island’s resources could be magically converted into identical bundles of claret and plover’s eggs. Although this would satisfy the envy test, Dworkin argues that doing this would be unfair, profoundly frustrating some people’s preferences while perhaps perfectly satisfying those of others.[12] More specifically, he argues that this conversion would violate what he calls the principle of abstraction. This principle “recognizes that the true opportunity cost for any transferrable resource is the price others would pay for it in an auction whose resources were offered in as abstract a form as possible, that is, in the form that permits the greatest possible flexibility in fine-tuning bids to plans and preferences.” Magically converting all the island’s resources into claret and plover’s eggs is unfair because it makes the array of available resources “much less sensitive—indeed as insensitive as possible—to the plans and preferences of the parties.”[13]

Instead, Dworkin famously proposes an auction, where each immigrant is given an equal number of (otherwise worthless) clamshells to bid on each distinct item on the island.[14] The principle of abstraction also mandates that the resources available in the auction be sold with as few legal limits on their permissible use as possible, consistent with the security and property rights of others. So, for example, an auction where clay is the available resource but the winner of the bid is legally prohibited from using the clay to make satirical sculptures violates the principle of abstraction by effectively precluding bids from those who might have preferred to use clay for this purpose, thus hiding the true cost of purchasing the clay.[15] Dworkin’s auction thus aims to make resources available in as flexible a format as possible, thus presumably ensuring that whatever price an immigrant ultimately pays for some resource reflects the true cost of excluding others from owning the resource.

A crucial part of Dworkin’s argument is thus that an auction of the island’s resources—unlike the magic conversion into identical bundles of some specific good—does a better job of being neutral with regard to the actual plans and preferences of the immigrants. As Patten says: “In Dworkin’s view, then, a fair distributive scheme does have to be responsive to the conceptions of the good that people actually hold. If it is not appropriately responsive—if the auctioneer puts up a less varied mix of goods for auction than she needs to—then people with conceptions of the good that are disfavored by the auctioneer’s decision can legitimately complain about the burden they are facing”.[16]

But how exactly does this appeal to Dworkin’s principle of abstraction support Patten’s view that accommodating the Tuesday Worshippers is not equivalent to subsidizing an expensive taste? I think there are two potential arguments to which Patten might appeal. I consider each one in turn.[17]


First, imagine that the two days of the week that will be official state holidays are determined by a modified Dworkinian auction. Each citizen is allocated two clamshells to “bid” or “vote” for her preferred days: each citizen can pick two different days, or use both of the clamshells to vote for the same day. The two days with the highest aggregate number of clamshells are selected as the official holidays. We can assume that at the outset, t1, everyone bids for the same two days—Saturday and Sunday—and so there’s no problem when those days are selected as the official state holidays. But now, several years later, some people have developed new religious beliefs, and request the official holiday auction be rerun to reflect their new conception of the good. Perhaps we should accept the following claim: it is a violation of the principle of abstraction if the auction to determine state holidays is fixed at a single moment in time, t1, and cannot be rerun at later points in time. Allowing the auction to be rerun ensures a greater degree of flexibility in the way resources are made available, and so ensures greater sensitivity to the plans and preferences of citizens.

Should we accept this view? I think the answer is no. At no point does Dworkin suggest that his principle of abstraction mandates an auction that is sensitive to people’s changing plans and preferences over time,[18] and indeed other central elements of his theory militate against this conclusion. In particular, Dworkin distinguishes between a person’s personality (including character, convictions, preferences, motives, tastes, and ambitions) on the one hand, and that person’s share of personal and external resources on the other.[19] Equality of resources aims to ensure that people are afforded equal resources with which to pursue their distinct plans of life, but individuals are assumed to bear consequential responsibility for the costs of their ambitions or plans.

Imagine that Dworkin’s auction has been successfully conducted on the island amongst all the immigrants. Albert spends the bulk of his clamshells on a plot of land that will allow him to earn a decent living as an apple farmer. Two years pass, and Albert’s career as an apple farmer is moderately successful: he is around the median in terms of wealth on the island. But then Albert, previously an atheist, experiences a religious revelation, and as a result he becomes devout. A feature of Albert’s new religious beliefs is the importance of constructing a place of worship in a specific location, at the base of the island’s only mountain. Betty, who bought it at the auction, owns the land where Albert wishes to build his place of worship. At the time of the auction, this land wasn’t particularly expensive, since it wasn’t obvious to anyone that it had any special features. Betty, however, has used the land to construct a lucrative vineyard (the shade from the mountain has proved ideal for this purpose), and the rest of the immigrants love Betty’s wine so much that she’s able to charge steep prices for her wine.[20] Given how valuable Betty’s land has become, Albert doesn’t have enough wealth to buy the land from her, and thus his ability to pursue his new conception of the good is significantly hampered. But had Albert had his religious revelation prior to the auction, he would have outbid Betty for the land, since he would have highly valued the land, but Betty was uncertain about the land’s commercial value. Albert therefore requests that the auction be rerun, since the auction’s allocation of property is no longer responsive to his plan of life.

A Dworkinian will deny that Albert has a claim of justice to a rerun of the auction or to any compensation for the fact his new conception of the good is expensive. Albert chose to use his clamshells in a way that does not now seem optimal to him, but he did so to pursue a plan of life, and so he must bear consequential responsibility for this choice in the same way that Dworkin would expect someone who develops expensive tastes for champagne and caviar to bear responsibility for her preferences. Albert has no egalitarian complaint that his new conception of the good is expensive, given his past choices. If he did, then it’s not clear what it would mean to be responsible for the costs of your choices against a background of equal resources. An imprudent person could spend most of her resources on parties and holidays, then sincerely repudiate this imprudent lifestyle and decide to become a prudent person instead, and demand the auction be rerun with a new set of clamshells where she can purchase a large plot of land that will be a good long-term investment for the future. This is absurd from a Dworkinian perspective.

The upshot is that the principle of abstraction does not require that the auction be revisited whenever someone changes her conception of the good and as a result wishes the initial auction had been conducted differently. The flexibility mandated by the principle of abstraction requires that the auction be sufficiently sensitive to persons’ plans and preferences at the time of the auction, but individuals are expected to take responsibility for the choices they make during the auction.

If we imagine, as we did earlier, that official state holidays are selected via a modified Dworkinian auction, then Dworkin’s principle of abstraction cannot be deployed to defend the view that the auction should be re-run when the Tuesday Worshippers change their religious beliefs. Dworkinian auctions do not need to be designed to be responsive to people’s changing plans and preferences.

But there’s a second way one might appeal to the principle of abstraction to defend the accommodation of the Tuesday Worshippers. Recall that the principle of abstraction “insists that people should in principle be left free…to use resources they acquire, including the leisure they provide and protect through their bidding program, in whatever way they wish, compatibly with the principle of security.”[21] So, to take our earlier example, if the winning bidder is legally prohibited from using the clay to make satirical sculptures, this violates the principle of abstraction by effectively precluding bids from those who might have preferred to use clay for this purpose, and so hiding the true cost of purchasing the clay. More generally, the principle of abstraction is inconsistent with any restriction on the use of private resources that is not required to protect the security or property of others.

Failing to accommodate the Tuesday Worshippers might seem inconsistent with this implication of the principle of abstraction. Suppose each person is the owner of a private resource: two holidays per week. Initially, everyone is happy to use this private resource in the same way—everyone wants Saturday and Sunday as the holidays, and this is why there’s nothing objectionable about these serving as the official state holidays. But it would be a violation of the principle of abstraction to declare that once people identify Saturday and Sunday as their preferred holidays, no further use or change is permitted. One might argue this constitutes an objectionable restriction on the use of an individual’s private resources in the same way it would be objectionable to sell a plot of land at auction, but restrict the owner’s use of the land by declaring that that owner must choose one use of the land (e.g., housing, farming, or commercial), and once she makes a decision, she cannot change her mind about how to use the land she has bought.

We should reject the argument in the preceding paragraph. The argument depends on an implausible assumption—namely, that each person “owns” a private resource: two official holidays per week. But this misrepresents what’s at stake in the Tuesday Worshippers example. The issue is not whether some group of people is being prevented from using a purely private resource, like land, in the manner they prefer. Which days will be designated official state holidays is not a question of how, or in what way, people’s private resources will be restricted. Instead, it’s a question about the format in which some unavoidably public resource will be made available to everyone. Each person has his or her own preferences about how the public resource will be made available, and fairness requires that each person’s preferences be taken into account in some way. But people’s preferences about this issue, or the votes or bids they make to express those preferences, are not private resources subject to the same moral principles as the resources that get purchased in a Dworkinian auction. They are, rather, like the clamshells that people use to purchase resources in Dworkin’s auction. And just as it is not a violation of the principle of abstraction to prevent people from reusing the same clamshells to make different purchases at a later date, I do not believe it’s a violation of the principle of abstraction to allow people to express their preferences about official state holidays only once during their adult life. Just as Dworkin expects people to bear consequential responsibility for the way they choose to spend their clamshells in his auction, a Dworkinian can plausibly expect people to bear consequential responsibility for the preferences they express—or their votes—regarding which days of the week should be official state holidays.

Of course, that it is not a violation of the principle of abstraction is consistent with it being permissible for citizens to choose, via a fair political process, that decisions about official state holidays should be revisited every X number of years. But if a policy of revisiting the decision is optional in the sense of being subject to the distribution of preferences in the political community—what Dworkin would call a choice-sensitive political issue—then, contra Patten, the Tuesday Worshippers do not have a claim of justice, since claims of justice should not be ignored simply because a sufficient number of people prefer to do so.[22]

To some, this conclusion will seem clearly mistaken. A liberal theory of justice permits people to change their religious convictions whenever they like, and so shouldn’t a liberal state also be responsive to the changes in people’s religious convictions, with that responsiveness including changes in official holidays as needed?

I share the intuition that the Tuesday Worshippers have a pro tanto claim of justice for accommodation—I simply don’t think that this conclusion can be derived from Dworkinian assumptions. The core difficulty is this: Dworkin’s theory depends on the view that an idealized market process is the correct way to measure whether people have been given equal shares of resources with which to pursue their distinct plans of life. Dworkin is thus committed to two ideas. The first is that other people’s preferences are parameters of justice, and so we have no justice-based complaint if other people’s preferences, as expressed via the market, make our preferred plan of life more difficult or costly to pursue.[23] The second is that we must accept consequential responsibility for the decisions we make within a fair market, and, in particular, we must accept the costs associated with option luck—that is, “whether someone gains or loses through accepting an isolated risk he or she should have anticipated and might have declined.”[24] If the state’s policies with regard to culture and religion are presented as Dworkinian resources, to be allocated by a fair market mechanism that respects these ideas, it’s difficult to see why a Dworkinian should accommodate the Tuesday Worshippers. If one already has a fair share of resources (as, by hypothesis, everyone initially does in the Tuesday Worshippers case), then choosing to change one’s conception of the good looks like a paradigmatic instance of something for which you are responsible, in Dworkin’s sense. In the marketplace, when some item goes up for sale and you choose to buy something else instead, you have to live with that choice: you cannot demand a second chance to buy the item when you change your plans a year or two later.

You might think the larger lesson to be drawn from all this is that minority rights or multicultural accommodations cannot be robustly justified from within a resourcist view of distributive justice. G. A. Cohen argues for this conclusion. He suggests that various cultural and religious preferences are, within the Dworkinian framework, expensive tastes, but he regards that conclusion as just one more reason to abandon resourcism and opt for something closer to his own equal access to advantage account.[25]

But that’s not the lesson I think we should draw. Instead, I think we should conclude that the emphasis that both Dworkin and Cohen place on personal responsibility—on bearing the costs of certain choices for which one can be said to be responsible in the relevant sense—is unhelpful when tackling certain questions of cultural and religious justice.[26] In cases where some accommodation for a cultural or religious minority is a matter of justice, I suspect it’s usually irrelevant whether the members of the group made a voluntary choice to be in a minority, or found themselves involuntarily in this position.[27]

I’ve tried to sketch part of my view on this elsewhere, and I don’t have space to defend it here.[28] My main point is simply to raise a worry about Patten’s aim to provide an account of minority rights that is congruent with the Dworkinian view of distributive justice.