Preference Change and Tax Policy, Again

by Will Wilkinson on August 3, 2005

Let’s go back to Layard’s attempt to justify his relative position “pollution” argument for taxation against his strawman libertarian critic:

Libertarians strongly object to this argument. They say it panders to the ignoble sentiment of envy, which ought to be disregarded. This is an extraordinarily weak argument. Public policy has to deal with human nature as it is. The desire for status is ubiquitous, and we all recognise it. Greed is also common, and libertarians do not disallow it. So why should they disallow the desire for status? Both sentiments are features of human nature. We are not perfect, and public policy should help us make the best of what we are. [Happiness, p. 153]

How many things are wrong with this argument? Feel free to count to ways in the comments.

Let me just say that it is cheap to equivocate between the desire for status and the desire for relative position in the income distribution. The latter is a contingent cultural expression of the former, as Layard himself concedes. And so public policy “dealing with human nature as it is” isn’t therefore public policy that treats our tastes for relative position in the income distribution as having normative weight. Our taste for coalitional solidarity and out-group xenophobia is natural in precisely the same way that our taste for status is. But we don’t think that this confers any significant normative weight on any old cultural expression of our tribalist impulses, such as apartheid, Jim Crow, or the Final Solution.

OK. Now try to square Layard’s “rebuttal” to the libertarian with the following. . .

First, Layard emphasizes in a number of places that many of our tastes, desires, preference, etc., are determined endogenously by institutional and cultural factors. Second, he argues convincingly that we can choose to change our own preferences.

The fact is, we can train our feelings. [p. 188]

Advocating Buddhist meditation Layard says:

Buddhism tells us to address the “poisons” that are disturbing our peace of mind: our unrealistic cravings and our tempestuous anger and resentment. [p. 189]

It is difficult not to draw the connection between this “poison” and the “pollution” that is caused by our taste for relative position. The difference is that here Layard clearly admits that the poisons are a function of our own states of mind–our unrealistic cravings–not simply the external fact that others have moved up relative to us.

Advocating the practice of cognitive psychotherapy, he writes:

If happiness depends on the gap between your perceived reality and your prior aspiration, cognitive therapy deals mainly with the perception of reality. [p. 197]

From his discussion of Buddhism, Christian mysticism, and cognitive therapy, Layard draws this lesson:

A second conclusion is that we have to control our tendency to compare ourselves with others. [p. 199]

(The first conclusion is that we ought to develop Buddhist habits of mindful attention to the present, which is also a way of controlling desire based in social comparison.)

Finally, Layard says that public policy can help by undertaking the “education of the spirit” through, naturally, mandatory state-funded spiritual education programs aimed at producing more open, compassionate, benovelent, and less comparison obsessed children.[p. 200-1]

Other than the monstrously illiberal suggestion that there ought to be a state curriculum in spiritual education, I agree entirely with Layard’s emphasis on our ability to shape our preferences through meditation, cognitive therapy, and, I would add, literary experience. But the spiritual education idea, in any case, amounts to conceding that a tax may not be the conclusion of the externalities/pollution argument, even if we insist on structuring policy around human nature as it is.

Indeed, if one of the most important lessons we should take away is that we can and should control our preferences based in social comparison, then why would we make public policy of a tax that is justified entirely in terms of those same unhealthy and controllable preferences? By choosing to treat these preferences as having normative weight in tax policy, isn’t the state sending exactly the wrong signal? Wouldn’t this be like arguing for a special tax on blacks on the grounds that this would increase total utility by pleasing the racist white majority, even though one has admitted that racist preferences are pernicious, and should be changed?

Viewing 30 Comments

    • ^
    • v
    Let's leave the Coasian social cost world and consider one variety of your preferred contractarianism:

    Wouldn't the "reasonable rejecter" probably accept the principle,

    (1) If it turns out that there are wealth inequalities that (somehow) cause those with less to feel that their status is threatend causing them sufficient unhappiness AND I happen to be one of the wealthy ones, then I should accept some transfer to them as long as its done in accordance with the rule of law and this policy doesn't screw up the economy too bad.

    but decisively reject the principle,

    (2) If it turns out that there are two races one of which has historically enslaved the other -- with current poorer members of the enslaving race compensating for their relative-poverty-caused feelings of inferiority by loathing and harming the enslaved race -- AND I am a member of the enslaved race, then I should accept a tax on me to make those poor white folks feel better.

    ??
    • ^
    • v
    And, for that matter, isn't the "reasonably rejecter" more likely to reject,

    (3) If their are wealth inequalities that make the less well-off feel that their status is threatened AND I happen to be less well-off, then I should get over my feelings of inadequacy and lesser respect rather than making demands on the very wealthy IF the costs of taxation to their happiness even slightly exceeds the costs of my getting over it.

    than (1) above??
    • ^
    • v
    Bill,

    No.

    It's the "unreasonable rejecter" who prefers (1) to (3).

    Now you know.

    But, even if 99% of people accept (1) for themselves, I prefer a world that respects the choice of the 1% to reject the transfer (and of the others who don't want to force them) over one that forcibly imposes these transfers upon them.

    If you think wanting these transfers is so reasonable, why wouldn't you expect them to take place through a voluntary process? Surely it wouldn't take much more than someone like you explaining to thim how reasonable it is for them to agree, right?

    And organizing lots of people voluntarily, and making their transfers contingent on some threshold being reached, has become relativley cheap.

    Is there a good reason to prefer the coercive regime over the voluntary one other than to satisfy some irrational, destructive, impulse to punish those who have achieved more success than others?

    Oh and if we're counting hedons, I think we should factor in the unhappiness that the coercive regime causes people like me, who would prefer that the autonomy of others is respected.
    • ^
    • v
    Bill, You're not forming the principles at a sufficient level of generality. The question is whether anyone in a position of impartiality and adequate information could reasonably reject a principle that justifies coercive transfer in terms of contingent preferences for relative income. The answer is, yes, such a principle could be reasonably rejected. Because (1) those preferences should not be granted normative status (their contingency & malleability just reinforces this); (2) the principle doesn't recognize the separateness of persons (i.e., if my entitlements are too wrapped up in your preferences, then I'm not being conceived of as a distinct being with separate legimitate projects, and thus not being extended the respect persons deserve); and more, I'm sure. I think this is a no-brainer in Scanlonian terms. Preferences are evaluable. And just because you have a preference for something doesn't give your preference weight. It more obvious that the racist tax should be rejected. But the grounds for rejecting them are pretty similar.
    • ^
    • v
    Will: You originally said "it is cheap to equivocate between the desire for status and the desire for relative position in the income distribution." But then you turn around and re-write my principles in terms of "contingent preferences for relative income." You surely agree that that is a different question. It may be your question, but it's not mine and I can't see how yours is more useful except to make your point. To say that the preferences are "contingent" begs an important questions and to describe them as for "relative income" avoids seeing through to the real preferences at stake including (a)not to be seen as less worthy because of lesser buying power, social staus that comes from wealth, etc., (b) to be able to confer benefits on others and perform social functions that require sufficient relative income, and a host of other reasons. (Layard may be a convenient foil for you here, but as you keep noting, he's hardly making the best available argument.)

    When you say that I formulated the principles in too specific terms, are you just saying that they are more specific then the kind Scanlon uses? If that's true (I'm not convinced), I disagree with Scanlon and would defend my version of his views over his.

    I don't know about Scanlon's politics. Is he a libertarian or is it just that you think he should be given his views as you understand them?

    Its very dubious and highly disputed that respecting autonomy requires not taxing high incomes more than low incomes (for whatever reason). The thesis that the motive of allieviating suffering from relative income deprivation makes the taxation especially anti-autonomy is way out. Also I have argued all over this blog that an individual's [relative] autonomy should be evaluated by looking at the whole range of choices he/she faces and how attractive those are compared to others' (in his/her view). I suggest that this is a much more meaningful view of autonomy than one that sees it as having ones rights as side-constraints respected.
    • ^
    • v
    Here's an improved version of my last post. Please excuse the reposting which, I admit, can be annoying:

    Will: You originally said "it is cheap to equivocate between the desire for status and the desire for relative position in the income distribution." But then you turn around and re-write my principles in terms of "contingent preferences for relative income." You surely agree that that is a different question! It may be your question, but it's not mine (or Layard’s, I think) and I can't see how yours is more useful except to make your point. To say that the preferences are "contingent" (upon the individual’s choice I think you mean) itself begs an important questions. To describe them as for "relative income" avoids taking stock of the real preferences at stake including (a) not to be see oneself as less worthy because of lesser buying power, (b) not to suffer from lesser social status that comes from being known to have less wealth, (c) to be able to confer benefits on others and perform other social functions that require sufficient relative income, (d) to have more choices about what sort of employment to accept and perhaps to be an employer instead of an employee, and a host of other reasons. No one has a preference for higher relative income PER SE! If that is really what Layard thinks, then that makes him a strawman. (So why waste all this time debunking him.) All these ills are consequences of having a lower relative income. This does raise the interesting issue of what we are trying to do when we describe agents in terms of their “preferences”, itself a word that makes them sound relatively trivial. (“I ‘prefer’ that there not be genocide in the Sudan.”)

    When you say that I formulated the principles in too specific terms, are you just saying that they are more specific then the kind Scanlon uses? If that's true (I'm not convinced), I disagree with Scanlon and would defend my version of his views over his.

    I don't know about Scanlon's politics. Is he a libertarian or is it just that you think he should be given his views as you understand them? I’m pretty sure he’s not a libertarian.

    It’s very dubious and highly disputed that respecting autonomy requires not taxing high incomes more than low incomes (for whatever reason). The thesis that the motive of alleviating suffering from relative income deprivation makes the taxation especially anti-autonomy is way out. (You seem to think Scanlon would be onboard here. I strongly doubt it.)

    Also, as I have argued all over this blog, an individual's [relative] autonomy should be evaluated by looking at the whole range of choices he/she faces and how attractive those are compared to others' choices (in his/her view). I suggest that this is a much more meaningful view of autonomy than one that sees it as having ones rights as side-constraints respected. Are you going for a side-constraints view of autonomy, or do you accept that it has to do with the quality of choice faced?
    • ^
    • v
    Gil: The point of Scanlon's contractarianism is to ask what principles people we would accept if we didn't know, for example, whether we were going to be rich or poor. The main difference between his view and Rawls' is that we're not just considering what society we'd be willing to take the risk of occupying an unknown position in. Rather, for Scanlon the right principles are the ones that give us a society in which we'd be willing to accept any position. (That's a rough characterization and I'm no expert.)

    So, given the current distribution, it may be impossible to organize a "redistributive movement" among the wealthy (without democratic politics involving the institutions of government). But even so, what I say about which principles would more likely be rejected by Scanlon's contractors could be true and morally relevant (on his theory as I understand it).
    • ^
    • v
    Thanks for the explanation, Bill. I was more familiar with Rawls than Scanlon, but I think they both "game" the system with their rules, and I don't accept that either of their bargains capture a good conception of justice.

    In any case, I'm not at all inclined to accept 1). And, I'm a reasonable man. So, you and Scanlon must accept that it's unjust.

    It seems reasonable to me to consider things other than my relative place in some social or economic hierarchy (like whether people's autonomy is respected to a degree that makes a slide into tyranny unlikely). If Scanlon defines this as outside of reasonable consideration, then I think he's cheating and the bargain is intrinsically unfair.
    • ^
    • v
    Gil: I'm not claiming to speak for Scanlon. I don't know what he'd think about (1).

    I don't know what you mean by autonomy, but you can find a sketch of my thoughts in this thread at the end of my response to Will. Also, I had this extended discussion with a libertarian reader of this blog under a thread about the "Political Class". That had lots of talk about autonomy in it.
    • ^
    • v
    Bill,

    By autonomy I mean people's ability to persue their own peaceful projects with the minimum coercive interference from other people.

    This leaves a huge range of wonderful, creative, kind voluntary cooperation that lets people help each other without treating each other as means rather than ends.

    So, I endorse the Nozickean rights as side-constraints view, rather than forcing other people to improve the choices of the worse-off.
    • ^
    • v
    Gil: The definitive critique of Nozick's political philosophy is G.A. Cohen "Self-ownership, Freedom, and Equality" and (more tersely) in a 1997 issue of Critical Review titled "Libertarianism" (I think). These works have convinced me and many others.
    • ^
    • v
    Well, Tom G. Palmer's critique seems pretty strong to me.

    Perhaps you and many others should reconsider.
    • ^
    • v
    Palmer's article totally misses the point, because Cohen's not arguing that the world should be (have been) jointly owned. He's just saying that nothing about self-ownership tells you whether it is (was) and Nozickian argument assume that it wasn't. The main arguments Palmer makes against joint ownership are consequentialist (and, if I remember, Palmer claims to be a consequentialist which is surely incompatible with being a Nozickian). So this whole route dismisses the Nozickian side-constraints, self-ownership world and accepts consequentialist reasoning which is what I was advocating. Sure, you could say that consequentialist reasoning just establishes that the world should have not initially been jointly owned and after that its Nozick all the way. But that would be totally arbitrary.
    • ^
    • v
    And, to be fair, Palmer's article also adduces all sorts of reasons why self-owners should/would/might agree to divide joint holdings or their product. I forgot about that part. Those are the most interesting sections of the paper, but they miss the point too because Nozickian self-owners under joint ownership would have the RIGHT to NOT reason the way Palmer argues they shoud. Also, absent ignorance, they wouldn't agree to anything that would ultimately make them worse off than continued joint-ownership. So we're back to considering the (of course terrible) consequences of joint-ownership and, ultimately, to rejecting self-ownership + side-constraints in favor of, if not some form of consequentialism, then at least a contratarianism that makes stipulations with regard to possible future outcomes.
    • ^
    • v
    Bill,

    I don't know what you're talking about.

    It looks to me like there's absolutely nothing left of Cohen's "argument".

    "Also, absent ignorance, they wouldn't agree to anything that would make them worse off than continued joint-ownership"

    But, almost anything would make them better off than continued joint-ownership. Continued joint-ownership is suicide.

    Everyone, absent insanity, would welcome the improvements that come from private ownership.

    Requiring unanimity to allow this (or even to agree to the disposition of jointly-owned property) is a death sentence to all.

    So, remind me why anything from Cohen should lead someone away from self-ownership+side-constraints?

    I mean, other than the fact that liberty leads to possible circumstances that some people don't like.

    I think all systems have that feature.

    How does Cohen get us somewhere better?
    • ^
    • v
    Cohen is not arguing for joint-ownership. That's precisely what Palmer (and you apparently) don't see/accept. I'm also not arguing for it.

    Your statement,

    "...almost anything would make them better off than continued joint-ownership."

    seems to illustrate your acceptance of my point that deciding between joint-ownership and private-ownership requires considering the consequences of these regimes. It can't be done on self-ownership grounds alone. (That's Cohen's point.)

    Nozick's side-constraints view of rights, on the other hand, says that these rights must be respected no matter what the consequences. But, says Cohen, doing so won't necessarily lead to inequality if the world is jointly owned.

    Ergo, if you want to defend non-joint property rights as side-constraints, it will have to be on the basis of the good consequences that flow from such a regime. Once these rights are not absolute but depend on the good consequences, then respecting them absolutely (as side-constraints) can be attacked by arguing that better consequcnes (e.g. for individual autonomy) would ensue if they were respected less than absolutely.
    • ^
    • v
    Bill,

    We're talking about people, right?

    Individual people who have individual goals, and respond to incentives?

    I think Nozick's model approximates my sense of our moral obligation to each other to respect each other's autonomy and limit our interactions to the voluntary (other than defense).

    And, his notion of the distributive justice where justly acquired and transferred holdings lead to a just distribution, regardless of how that distribution looks, also squares with my thoughts.

    Why should anyone care whether you or Cohen think that in the unworkable fictional state of joint-ownership of everything some idiot might not agree to selling part of it?

    This doesn't seem to go anywhere towards demonstrating that I must change my notions of morality and justice because people in that situation would only sell their property by considering consequences (as opposed to what?).
    • ^
    • v
    Gil: Oh, on the contrary, you MUST change them!

    :)
    • ^
    • v
    Oh.

    Okay.
    • ^
    • v
    I'll wade into this. The problem with your consequentialist leap, Bill, is that it smuggles an egalitarian premise into the discussion without justifying it: the premise that we can define a notion of better or worse overall consequences through interpersonal utility comparisons, i.e. that we can speak of "better consequences" without answering the question "consequences to whom?".

    But this won't work, because different people's utilities are incomparable, and this incomparability *is* a consequence of self-ownership: it is, in fact, pretty much what self-ownership means. I own myself because I am an end in myself, not a means to anyone else's ends or to a greater social end of maximization of good overall consequences. And my utility is neither lesser, greater, *nor equal* to anyone else's. That's why you need rights to be side-constraints and not balancing acts.

    So consequentialist egalitarianism cannot be used to reconcile attacks on property rights with self-ownership.
    • ^
    • v