Does it matter whether rich people really like money?

Responding to the ongoing debate among economists about Congresswoman Alexandria Ocasio-Cortez’s proposal of a 70% tax rate on the very richest Americans, Chris Dillow  says: “I’ve never been comfortable with the marginal utility argument for high top tax rates. One reason why the rich are rich is that they value an additional dollar highly”. I disagree. I think the marginal utility argument is the strongest argument against inequality, and that it is consistent with higher tax rates on the rich (also, to the extent that marginal utility weakens the case for higher taxes, I think it is correct to do so).

The ‘marginal utility’ argument is that inequality is bad because poor people get so much more benefit from extra money than the rich. If £10,000 went missing from a billionaire’s bank account, the chances are that they would barely notice the loss, whereas for many of the poorest people that sum of money would be life changing. A world in which the billionaire’s taxes were raised to pay for lower taxes, income support or better services for the poor is therefore clearly better.

Yet while the argument seems pretty clear cut at the extremes of billionaire and pauper, Dillow points out that the generalisation that poorer people have more desire or use for money doesn’t always hold. As he says, at least one of the reasons why rich people are rich is because they care a lot about having money, and have prioritised getting rich over other things. It isn’t so obvious that a comfortably-off middle class person would get more enjoyment from an extra £10,000 than a millionaire. If I got that much money, it would probably top up my pension. Perhaps they would be able to buy a particularly fine wine or something and get more pleasure from that, I don’t know.

I have to say, I’m sceptical Dillow’s objection holds in most cases – I think the empirical evidence suggests that if you took any random pair of people, the one who would benefit more from a bit more money would typically be the poorer one. But I do expect that there would be a non-negligible number of instances where the richer person would benefit more (most likely where the difference in affluence is small, or where one of the pair is particularly materialistic or non-materialistic).

Dillow thinks that such cases undermine the marginal utility argument. I’m inclined to disagree. In my view, to a certain extent, we are all naturally inclined towards different tastes and interests, and some of these are more difficult to satisfy than others (what in the philosophical literature is referred to as ‘expensive tastes’). Some people love travelling to places, others are happier reading about them. Some people are connoisseurs of the fine dining, others genuinely can’t tell the difference from ordinary fare.  If everybody had the same amount of money, those with more expensive tastes would be less happy (see G.A. Cohen for a more detailed discussion of the idea of expensive tastes, arguing for a somewhat different conclusion). So I think it is fair enough for them to have more.

Does that imply I should be against high taxes on the very rich? I think not, for a number of reasons. First, expensive tastes may mean that some degree of inequality is optimal, but it doesn’t justify any level of inequality. I can easily accept that some people need two or three times more resources than others to reach similar levels of wellbeing, but it seems implausible that the super rich really need incomes hundreds of times larger than the rest of us. Second, while I accept that the rich are more likely to have expensive tastes, the correlation is unlikely to be perfect. I’m not convinced that enough of the rich have expensive tastes and enough of the poor have modest ones for redistribution not be a good idea. Third, a person’s expensive tastes may reflect a mistaken understanding of their own wellbeing. For example, there is evidence to suggest that people overestimate how happy material goods make them. This would imply that high taxes on the rich might have some paternalistic value – encouraging them, for example, to take more leisure for their own good. Fourth,  preferences are not necessarily fixed, and can perhaps be influenced by policy (both directly and through social norms). An economic system that encourages material competition among elites is one that is liable to produce more expensive tastes. By contrast, higher taxes may encourage more modest lifestyles. I don’t want to over-emphasise this, as I’m sceptical as to how sensitive such preferences are likely to be to policy, but it is at least possible.

Finally, and crucially, inequality has all sorts of spillover effects – it is not just about how well off each individual is, but also shapes their interactions. High levels of inequality plausibly lead to all manner of undesirable social relations: the development of ghettoes and social segregation,  domination of the political process by the rich, the stigmatisation of the poor and the undermining of their self-esteem (see Martin O’Neill for a fuller discussion). (Against this, of course, we should at least consider the ‘trickle down’ argument that permitting inequality encourages economic innovation and development in a way that benefits the poor).

There’s clearly a lot of supposition, speculation and open questions here. Nevertheless, I think I’ve said enough here to show that the marginal utility argument is unlikely to push strongly against higher taxes on the rich.


  1. Kieran Latty · · Reply

    Here is Lerner on this question:

    ‘Another problem arises if we relax assumption (6). The experience of having a larger income may develop a person’s tastes and capacities for enjoying income so that the marginal utility curves of rich people would be higher than those of poor people. An ideal distribution of income would then give more to people who have been richer. The converse of this too may be argued with about the same plausibility. A person with high income gets used to luxuries so that he hardly notices items that would give a great thrill to a poor man unused to them. This would reverse the first argument and strengthen the case for equalizing incomes. However even if the first consideration were known to outweigh the second, and even if of the two arguments only the first were found to have validity, there would still be a probable gain from an egalitarian division of income. But this would now be true only in a longer run. If experience of higher income really had the effect of raising the curve of marginal utility of income, it would be best temporarily to leave more than the average of income with those whose levels of income has been higher. This still does not alter our conclusion. It only addresses itself to how rapid should be the transition from a previous unequal to an ideal egalitarian distribution of income, and to the importance of avoiding the establishment of inequalities of income in the first place.’ [1]

    And here is Lange:

    ‘Our argument holds strictly if the marginal utility curve of income is the same for all individuals.’ Of course, this does not correspond to reality, and one might think of taking into account the differences between the marginal utility curves of income of different individuals by granting higher incomes to the more ” sensitive ” persons. But as such differences as to ” sensitiveness ” cannot be measured the scheme would be impracticable. Besides, the differ- ences in ” sensitiveness ” existing in present society are chiefly due to the social barriers between classes, e.g. a Hungarian count being more ” sensitive ” than a Hungarian peasant. Such differences would disappear in the relatively homogeneous social stratification of a socialist society and all differences as to ” sensitiveness” would be of purely individual character. Such individual differences may be assumed to be distributed according to the normal law of error.2 Thus, basing the distribution of incomes on the assumption that all individuals have the same marginal utility curve of income, a socialist society would strike the right average in estimating the relative urgency of the needs of different persons, leaving only random errors, while the distribution of income in capitalist society introduces a constant error-a class bias in favour of the rich.’ [2]

    [1] Abba P. Lerner, “Utilitarian Marginalism (Nozick, Rawls, Justice, and Welfare),” Eastern Economic Journal 4, no. 1 (1978): 59.

    [2] Oskar Lange, “On the Economic Theory of Socialism: Part Two,” The Review of Economic Studies 4, no. 2 (February 1, 1937): 125.

    1. Thanks for this. I know Lerner and Lange by name, but don’t know their work much at all, so this is interesting. I think I agree with all of it.

      Lerner’s conclusion about rapidity of transition is potentially a big concession, no? That we should have a slower transition to full equality to make it easier on those that start rich?

  2. Kieran Latty · · Reply

    Lerner continues:

    ‘A qualification might appear to be necessary for differences in income due to greater effort. When an individual applies himself more assiduously to his work this might be held to indicate that he has greater use or need for the extra income derived from the extra effort. Some might contend that such an individual should receive a larger income. This does not follow. A person works hard only when the wage plus the attractiveness of the work itself (or minus the irksomeness of the work), in relation to the opportunity cost of the alterna tive leisure forgone, is greater than for others. But there is no more reason for believing that a person works harder because an additional dollar is worth more to him than that he works harder because the marginal disutility of work or the marginal utility of leisure is less for him. We know only that these are less in relation to the marginal utility of consumption goods. But it is all of these marginal utilities combined that constitute the marginal utility of income and there is no reason here for believing that that is higher or lower than for others.’

    p 59.

  3. You are too nice about this. I do not think anyone seriously believes that comparing utilities is impossible, *at least for this reason.* I suppose I can accept that someone might reject all interpersonal utility comparisons, but the idea of objecting to redistribution because maybe the rich really, really, really like money is so absurd that I don’t think anyone who has seriously reflected on the view honestly holds it, and I strongly suspect some of the advocates of this objection (e.g. Posner) of intellectual dishonesty on this particular question.

    No one really believes that Bill Gates enjoys his marginal billion dollars more than 10,000 people would enjoy receiving a hundred thousand dollars

    1. So it’s obviously silly at the extreme, but it seems plausible to me that it could legitimate quite wide inequalities ie some people could need two or three times the resources of other to achieve similar utility.

      Whether that has any relevance to real-world levels of inequality is, of course, a separate question.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: