Given the (debatable) difficulties of interpersonal utility comparison, the Kaldor-Hicks compensation test has been proposed as a way of assessing whether or not a given policy improves social welfare.
To understand the idea of a Kaldor-Hicks improvement, we first must introduce the idea of a Pareto improvement, which a Kaldor-Hicks improvement generalises. A policy is said to lead to a Pareto improvement iff at least one person is left better off, and no one is left worse off. However most policies will leave at least someone worse off, so this criterion is difficult to apply in practice. Consequently, the idea of a Kaldor-Hicks improvement was proposed in order to generalise the Pareto improvement into something workable:
Courtesy of Wikipedia:
A re-allocation is a Kaldor–Hicks improvement if those that are made better off could hypothetically compensate those that are made worse off ensuring a Pareto-improving outcome. The compensation does not actually have to occur (there is no presumption in favor of status-quo) and thus, a Kaldor–Hicks improvement can in fact leave some people worse off.
Crucially, Kaldor-Hicks improvements are usually calculated over dollars or other units of concrete resources, not utils. This is because dollars and concrete resources are transferable, whereas utils aren’t.
It is not difficult to imagine circumstances in which the pursuit of Kaldor-Hicks improvements might be pernicious.
Imagine a political system with the following characteristics:
- There are exactly two citizens.
- The polity only makes decisions regarding the output of an oracle. One of the two citizens is a dictator who decides whether or not to accept the proposals of the oracle and, separately, whether to pay compensation for those proposals.
- The dictator is ruthlessly self interested.
- Each of the citizens has an income.
- The utility of each citizen is equal to log(income)
- As mentioned earlier, there is an oracle. The oracle regularly spits out proposals which are not Pareto improvements, but which are Kaldor-Hicks improvements with regards to the income of the two citizens. When these proposals are given, the polity (i.e., in practice, the dictator) makes a decision about whether or not to enact them. Separately the polity (again, really just the dictator) makes a decision about whether to pay compensation.
Under these circumstances, every proposal will be enacted. However, only those proposals which are initially to the detriment of the dictator will be compensated, whereas those that are to the detriment of the plebeian will be ignored. In the long run, this will generate growing inequality between the two citizens.
We could give a mathematical model, but that would be to labour the point. Because the utility function of the citizens is log concave, there is thus no guarantee that in the long run, from the point of view of aggregate utility, the society will be better off with the Kaldor-Hicks oracle than without it. In fact if we set the parameters at reasonable values, and let the simulation run for, say, ten steps, aggregate utility is quite likely to fall.
The point is that, in an unfair society, the Kaldor-Hicks framework for judging policy might well be a weapon against the poor, and if the marginal utility of income declines steeply, this will likely reduce aggregate utility(1).
We don’t live in a dictatorship. However we do live in an oligarchy. The exact extent to which our society is oligarchical is debatable, but the evidence seems to suggest that the rich have substantially more say than everyone else, at the least. Just as under a dictatorship, so under an oligarchy, Kaldor-Hicks improvements are likely to be pursued, but only compensated when the party that suffers was initially well to do. Of course our society is not a perfect oligarchy, and how this will effect things is hard to calculate.
Indeed, to the extent that the oligarchy we live under is only partial, it may well be that the popular classes and their defenders find it beneficial to block Kaldor-Hicks improvement proposals, recognising that it is sometimes easier to stop such things, than it is to ensure that compensation will be paid and will continue to be paid. Many economists will no doubt condemn their short-sightedness, not recognising that it is in fact themselves who cannot see the full circumstances.
We don’t have magical oracles which are responsible for all policy proposals. We do however have economists, who play an important role in legitimating policy and generating options.(2)
Economists often defend the idea of generating policy which is ‘best’ in some abstract sense, without reference to prevailing political conditions. Paul Krugman quotes Harry Johnson as saying:
Second best policies are always devised by third-best economists working for fourth-best politicians.
The point being that there’s something seedy and -heaven forbid- unscientific about viewing policy generation through the prism of how it might favour existing political interests. If however we do live in an oligarchy, identifying good policy through the lens of Kaldor-Hicks improvements might simply give pre-generated ideas and a ready made justificatory framework to the powers that be, perhaps ultimately reducing aggregate utility.
The nature of Kaldor-Hicks as a decision criterion is that it provides a fully general argument for redistributing resources to oligarchs under almost any circumstances- simply carry out ‘improvements’ and only provide the compensation if that is in the interests of the oligarchical class.
The point that economic inequality makes KH a poor decision rule has been ably made before so what is novel here? What I am trying to do in the above analysis is to show that If KH is a bad decision rule when there is economic inequality, it is a terrible decision rule when political inequality accompanies economic inequality. And of course economic and political inequality will always accompany each other.
The public-policy theorist or economist is not just watching the game, they are entangled in it. Providing advice that can be used strategically by one side without being aware of that -of how that advice will deployed and used as a justifcatory framework- is irresponsible. The nature of Kaldor-Hicks as a decision criterion is that it provides a fully general argument for redistributing resources to oligarchs- simply carry out the ‘improvement’ and only provide the compensation if that is in the interests of the oligarchical class. If you care about human welfare, you should be very careful about promoting Kaldor-Hicks efficiency as a decision framework.
— — — — — — — — — — — — — — — — — — — — — — — — — — — — — — —
(1) “But!” you protest “I am not a utilitarian!”. Neither am I really, but it hardly matters. I mention aggregate utility as a gesture to utilitarianism, but honestly the results will come out the same whatever your view on distributive justice and social welfare is. From the point of view of basically all normative theories of social welfare- utilitarianism, priortarianism and Rawlsianism for example, the operation of Kaldor-Hicks under an oligarchy or dictatorship is likely to be A Bad Thing. From the point of view of the major rivals to social welfare theories- deontic theories like the historical theory of justice- Kaldor-Hicks style criteria are irrelevant. So who exactly is meant to be persuaded by them?
(2) “But” you protest “If society is an oligarchy- what does it matter what ideas economists promulgate- the elites will simply do as they will and justifications like Kaldor-Hicks efficiency will be irrelevant.” It is not so simple. It seems likely we live in only a partial oligarchy, elites do not have uncontested control, and they must act and speak so as to maintain legitimacy. Ideas can be deployed strategically, and the struggle to justify one’s actions through ideas is part of the struggle to maintain hegemony. We have argued that “Kaldor-Hicks improvement” is a rather useful strategic idea- allowing elites to justify profiting at the expense of non elites, if overall output rises, even if aggregate utility falls. It’s exactly the kind of an idea an imperfect oligarchy might find appealing.