Gregory Mankiw has made another foray into philosophy, this time by letting us in on a ‘dirty little secret’: when economists give policy advice they’re often speaking as political philosophers rather than economic scientists.
Sorry, but if Mankiw is a representative example, they can’t be speaking as either.
To speak as a scientist, you need to be a scientist, which means you need to know a science. And economics, especially Mankiw’s brand of it, really doesn’t qualify.
Meanwhile, to speak as a political philosopher you need to know something about political philosophy. Some economists qualify. Like Amartya Sen. But not, by any stretch of charitable imagination, Gregory Mankiw.
As ever, Mankiw’s not attempting philosophy for its own sake. He’s got an agenda: to show that some policies he opposes – the Affordable Care Act and the raising of the minimum wage in the US – are based on flawed political philosophy.
The philosophy that justifies these bad policies is, it turns out, Utilitarianism. Utilitarians believe that a policy is justified to the extent that it maximizes aggregate welfare. Thus the above policies are justified as making things better for people in the aggregate, even though they will make things worse for some.
But it turns out, according to Mankiw, that utilitarianism is so flawed that any policy justified on utilitarian grounds must be rejected as a matter of caution. It also turns out that all proposed interferences with the free market are justified on the basis of utilitarianism. It follows that we must never interfere with the free market. Not a surprising conclusion for Mankiw to come to.
So here are the flaws with utilitarianism he mentions:
1. Utilitarianism is known by philosophers to be an inadequate moral theory. Utilitarianism would make it morally justified, for example, to kill one person and harvest her organs to save many lives, assuming this led to an aggregate increase in welfare. And we know that that isn’t morally justified.
2. Utilitarianism commits to the theory that interpersonal comparisons of utility are possible. If we are to know that A’s loss of welfare from some policy is less than B’s gain, A’s and B’s changes in welfare must be mathematically comparable. This requires a theory of cardinal utility: the theory that welfare-changes to different agents can be quantified and compared amongst each other. But how could you ever decisively confirm that one change is more or less than another? This scepticism must lead us away from the theory of cardinal utility and towards the theory of ordinal utility. On this theory, you can’t quantify welfare-changesat all: all you can say is which options each single agent prefers to which others. You can say, in other words, that A prefers x to y and that B prefers y to x, but not that A gets more benefit from x than B gets from y.
3. Utilitarian justifications depend upon the expected consequences of policies. But policies can have unintended consequences, for which the utilitarians justifying them are unable to account.
None of these considerations, nor any combination of them, supports anything like Mankiw’s conclusions. It’s so annoying to have to explain this, since it must be so obvious to every remotely intelligent person. But the New York Times published Mankiw’s piece, so the point must have failed to get across somewhere. So I will force myself through the tedious business of debunking this nonsense.
1. As every first-year ethics student who can manage a Pass knows, the criticisms Mankiw raises against utilitarianism are meant to show that it leaves out some important feature of moral judgment. Think about the example above. Why would it be wrong to kill a person and harvest her organs, even if it brought about an increase in aggregate utility? Something has failed to be accounted for. Some say that what is missed is the fact that killing a person for this purpose violates her basic rights, and considerations of rights trump considerations of aggregate utility. Others say that the missing element is the fact that there are some things that one ought not to do regardless of the consequences. At any rate, if we’re going to criticize utilitarianism in this way, we owe some explanation of which moral principles it fails to recognize.
The problem for Mankiw is that it’s very likely that such principles will be capable of justifying the policies he wants to reject. We might, for instance, decide that all people, even Americans, have a right to affordable health care, regardless of the losses in welfare that providing it might cause to some. Or we might decide that one simply ought not to deprive people of a fair wage and there’s an end on it. Mankiw could argue these points, of course, but in rejecting utilitarianism he merely raises them.
2. Despite what economists have been saying for a long time (see, for example Lionel Robbins’ 1932 Essay on the Nature and Significance of Economic Science), a move from cardinal to ordinal utility theory does not on its own provide any reasons against interfering with the free market. Suppose you’re an ordinal utility theorist, but you believe that everyone happens to have a preference for some distribution of resources that can only be brought about by coercion rather than voluntary exchange. Perhaps you even believe that all or most people have a strong preference for coercive distribution itself. Then you can make a utilitarian argument – an argument based on the mere consideration of preferences – for socialist intervention in the free market. Such a belief would be implausible, of course, but then so is the belief that most Americans today are happy with the wages they receive.
What really makes economists believe that the very nature of preference in some way shows the advantages of free exchange is not their commitment to ordinal preference theory. Rather, it’s their commitment to revealed preference theory. Paul Samuelson and others came up with this theory when they realized that it’s possible to be just as sceptical about ordinal utility as one can be about cardinal utility. We can question cardinal theory by asking: ‘How can we know that A’s benefit outweighs B’s loss?’ But we can just as easily question ordinal theory by asking: ‘How can know that A prefers x to y?’ The economist will answer: ‘Because A voluntarily exchanges y for x.’ But then how do we know that A is not systematically irrational, or simply confused about the identities of x and y, or exchanging for some purpose other than personal gain? To this new sceptical challenge, Samuelson (whose economic textbook is still taught to undergraduates) proposed the following solution: simply identify each agent’s preferences with what he or she displays through exchange. In other words, the sentences ‘A prefers x to y’ and ‘A will exchange y for x’ are identical in meaning.
With that identity in place, it’s easy to argue that voluntary exchange can best satisfy people’s preferences. Preferences are now, by definition, those things typically satisfied in voluntary exchange. But if moral philosophy can be done by the mere giving of definitions, I can morally justify punching Gregory Mankiw in the head merely by defining ‘good’ as ‘whatever hurts Gregory Mankiw.’ Quite obviously that is not the reason such an act would be justified.
3. Mankiw argues that, since utilitarian justifications are based on expected consequences, and since current economic science (heterodox economics far less so, though Mankiw doesn’t mention this) is insufficient to predict consequences accurately, the wisest course is to adopt a ‘do no harm’ policy. Mankiw’s analogy is with nineteenth-century doctors: given the rudimentary state of medical knowledge, such doctors were well-advised to take a do no harm policy. Economic policymakers should feel the same. But neither the analogy nor the notion of ‘do no harm policy’ tells us anything. Should the nineteenth-century doctor refuse to set a broken bone, or give water to a clearly dehydrated man, for fear that some freak of medicine might lead to terrible unintended consequences? Should nobody ever do anything for anyone, given how complex and unpredictable the world is?
Perhaps a risky act is justified when the risk is proportional to the urgency of the situation. But of course it’s hard to judge risk. Still, the principle is plausible enough. Let’s apply it to the case in hand. Millions of Americans live in abject poverty because some relative of theirs got one of the terminal illnesses that the world’s most high-profit insurance industry doesn’t feel like covering. Hundreds of thousands of Americans work sixty hour weeks, only to find they’ve still failed to provide for the basic needs of their children. In 2012, 46.5 million citizens of the world’s richest country were living below the poverty line, more than half of them children. Obama is one of the most right-wing presidents the US has ever had. The chances of any policy he supports being some kind of incipient socialist revolution are pretty much zero. But even his incredibly conservative proposals are far too much for Mankiw’s ‘do no harm’ policy. All children drowning in ponds had better hope that Gregory Mankiw isn’t nearby. He might stop anyone from jumping in to rescue them for fear that this might lead to minor bruising, or, worse, that it might somehow in some indirect way cost some rich person money.
I’m all for economists engaging in a bit of philosophy. After all I’m all for engagement running in the opposite direction. But switching from one discipline to another doesn’t license the immediate discontinuation of all critical thought. It’s not really a ‘dirty little secret’ that economists use what they think is ‘philosophy’ to inform their political advice; it’s scandalous, massive, and not secret at all that on that score there are far more qualified people to listen to. I’m not claiming to be one of them, but I know a terrible choice when I see one.