Rawls, Bentham and the Laffer curve (cross-post at Crooked Timber)

The 1970s saw two important and influential publications in the long debate over justice, equality and public policy. In 1971, there was Rawls Theory of Justice, commonly described in terms like “magisterial”. Then in 1974, at lunch with Jude Wanniski, Dick Cheney and Donald Rumsfeld, Arthur Laffer drew his now-eponymous curve on a napkin. Of course there was nothing new about the curve: it’s pretty obvious that an income tax levied at rates of either zero or 100 per cent isn’t going to raise any money, and interpolation does the rest. What was new was the Laffer hypothesis, that the US at the time was on the descending side of the curve, where a reduction in tax rates would raise tax revenue.

I’ve always understood Rawls in terms of the Laffer curve, as arguing in essence that we should be at the very top of the curve, maximizing the resources available for transfer to the poor, but not (as, say, Jerry Cohen might have advocated) going further than this to promote equality.

A couple of interesting Facebook discussions have led me to think that I might be wrong in my understanding of Rawls and that the position I’ve imputed to him is actually far closer to that of classical utilitarianism in the tradition of Bentham (which is, broadly speaking, my own view).

Facebook has its merits, but promoting open public discussion isn’t one of them, so I thought I’d throw this out to the slightly larger world of blog readers.

It started out with a post by Ingrid, mentioning a paper she’d presented on limitarianism (a new term to me, though not a new idea), the proposition that as a matter of public policy or of individual ethics, there ought to be a maximum limit on individual incomes. I’m hoping Ingrid might write something more about this here, but she shouldn’t be blamed for any errors I make.

On my interpretation of Rawls, it seems as his difference principle implies a kind of limitarianism. If high income earners are taxed to yield the maximum possible revenue, then their incomes will be bounded above. This won’t be an absolute bound, but will depend on their earning capacity which in turn will reflect both the general level of technology and the extent of “pre-distribution”, that is whether the economic organization of society yields highly unequal market incomes.

That got me thinking about utilitarianism. One way of looking at Rawls (not one he accepted, but still defensible) is as putting forward an extremely egalitarian rank-dependent version of utilitarianism, with all the weight on the bottom of the distribution. But it turns out that you don’t need this to advocate a tax system very close to the Laffer maximum. On the default utilitarian assumption of log utility, additional income to someone on $1 million a year yields 1 per cent of the utility of additional income to someone on $10 000. So, a costly tax-transfer system taking from money the $1 million group and giving it the $10 000 is beneficial as long as the proportion lost through incentive effects, collection costs and so on is less than 99 per cent. Diamond and Saez did the sums properly a while back and concluded that

the social marginal utility at the $1,364,000 average income of
the top 1 percent in 2007 (Piketty and Saez, 2003) is only 3.9 percent of the social
marginal utility of the median family,

The implied policy is a progressive tax system with a top marginal rate of 73 per cent.

It seemed obvious that a Rawlsian policy should be very similar. But after looking around a bit, it turned out that Rawls was, at best, ambivalent about the whole idea of progressive income taxation, let alone the kind of revenue-maximizing extraction implied by a utilitarian analysis. Another interesting discussion followed on Facebook with contributions from Jacob Levy, Jon Mandle, Mike Otsuka and others. I’ve long since lost track of FB privacy policy, but you may be able to read it here.

I got two things out of the discussion. The first is that Rawls had some mistaken ideas about economics, logically unrelated to his difference principle, which contributed to his scepticism about progressive taxes. The intellectual atmosphere of the 1970s, which gave rise to the Laffer curve and the “tax revolt” may have had something to do with this.

The second is his focus on “ideal theory” which I don’t fully understand, but which seems to me to be unhelpful, partly for the reasons put forward by Jacob Levy. The general claim made by Rawls is that, whatever the case for progressive income taxes in society as it is, a properly organized society with limits on inheritance, open access to education and no unfair market power wouldn’t need them.

I don’t buy this. In any society sufficiently close to reality to need a taxation system at all, there are going to be inequalities of pre-tax income large enough that an egalitarian would want to redistribute that income. Given that nearly all taxes are regressive, a progressive income tax is an essential part of such a policy.

34 thoughts on “Rawls, Bentham and the Laffer curve (cross-post at Crooked Timber)

  1. @J-D

    What I mention would not be a mere tax increase but a major limitation on oligachic capitalist power and a step on the road to worker socialism. The USA would not permit a country in its power orbit to do that. Or haven’t you noticed what happens to countries that;

    (a) do something the US doesn’t like; and
    (b) don’t have their own nuclear deterrent?

  2. @J-D

    Just so we are clear. Exactly where did anyone say that:

    The USA would respond to an increase in tax rates in Australia — even a drastic increase in taxation — by attacking Australia and reducing it to rubble.

  3. Without wishing to be disrespectful, I cannot find any use of the Laffer curve. IMO, the Laffer curve belongs to the 2D economics, aptly ridiculed by Newtownian. The curve, by construction, has a maximum and this is interpreted as the maximum tax revenue. I am told, this maximum is of interest because it would provide the maximum revenue for income redistribution. Maybe. It could also provide the maximum revenue for expenditure on war machinery or expenditure on the salaries of advisers who sing the song that is in fashion. Why “the” Laffer curve? Is there only 1 income distribution and it is perfectly equal? If not, what is “the” Laffer curve supposed to represent? An economy with 1 individual? In short, all the important dimensions of an economy are left out and, for reasons given in JQ’s post, the graph has no intrinsically interesting properties.

    There is a paper by Hal R Varian, titled “Equity, Envy, and Efficiency”, JET, 9,63-91 (1974), which I came to read only recently. He makes precise the notions of equity, envy and efficiency as well as the concept of a ‘fair’ allocation in terms of preferences. He discusses his notion of ‘fair’ in relation to Rawls’ notion of just society. Varian also contrasts his approach with social choice theory.

    I do not have sufficient knowledge about Rawls’ theory to form an opinion. I’d greatly appreciate JQ’s and others opinion in this regard

    I like Varian’s approach. While it may not be directly helpful for a specific policy question, it is, IMO, very helpful in providing a framework for policy formation and public discussions. For example, results which hold for an exchange economy (no production), do not extend to an economy with production (what does a Laffer curve have to say on this point?); a welcome reminder that a reference to ‘a market economy’ says essentially nothing.

  4. @ChrisH discreetly
    Race can sensibly be regarded as extended family so it is never surprising if people’s feelings for who is outside the tribe or family are attached to some of the conventionally used msrkers or indicia of race. And religion would be just as likely in many countries. Quite a lot of rich Scandinavians didn’t wait for a lot of Middle Eastern Muslims to turn up in their midst to prove that support for the Scsndinavian type of welfare state was special and associared with the feeling that people in their country were family. They had views, not publicly expressed, about class
    and discreetly went about establishing foreign connections, living well on expense accounts and using tax loopholes that, admittedly, showed less than complaisant willingness to cough up high taxes long before the racial, religious or other ethnic factor spread their distaste for high taxes to the average Scandinavian.

  5. @Ikonoclast

    You ask me whether I have ‘noticed what happens to countries that;
    (a) do something the US doesn’t like; and
    (b) don’t have their own nuclear deterrent?’

    Lots of countries have done things that the US didn’t like. In some case the US has attacked those countries and in some cases it hasn’t.

    The US did not like it when Iran held US diplomats and others hostage in the US embassy in Teheran, but it did not reduce Iran to rubble (although one Iranian was killed during the abortive US attempt to rescue the hostages).

    The US did not like it when New Zealand banned the entry of nuclear-powered and nuclear-armed warships, effectively ending US naval visits to New Zealand pots, but it did not reduce New Zealand to rubble (although the US did announce that it was suspending treaty obligations to New Zealand).

    Did you not notice those events? But perhaps that question is unfair. Perhaps you are younger than I am and you do not have the personal recollections of those events that I do. Still, they are well documented and you can look them up in sources you trust if you don’t want to take my word for it.

  6. @Ivor

    That’s what I thought Ikonoclast meant. Did you have a different understanding of Ikonoclast’s comment? Can you explain how your interpretation was different from mine?

  7. @Jereboam

    It is not correct to say that race can sensibly be regarded as extended family. If there are people who regard race as extended family, what they are doing is not sensible.

Leave a Reply

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

WordPress.com Logo

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

Google photo

You are commenting using your Google 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