Bryan Caplan and Robin Hanson, two libertarian economists, had a debate on liberty versus efficiency (links to their opening posts are available here). A little backstory helps. Historically libertarianism has been about liberty. That is probably still true for the run of the mill libertarian. But it is no longer true for economists and sophisticated libertarians. They put efficiency first.
It all started with Ronald Coase. A common market failure is the case where a factory pollutes a river. The people who live downriver pay the "costs" of pollution even though they didn't buy any widgets. This is called an externality or an external cost. The standard economic response is to place a special tax on the factory called a Pigovian tax. The money can be redirected towards paying for the cleanup costs. That's the reasoning behind road pricing which charges a toll on roads during high traffic times. It is also the reasoning behind a carbon tax to "stop" global warming.
On this view external costs are a matter of rights. The factory does not have the right to damage someone else's property. Coase overturned this view. You could just as easily say that the people who lived downstream want to force the factory to stop making widgets. The people who live downstream are infringing on the rights of the factory. Coase realized that there is an unavoidable tradeoff between clean rivers and more widgets. It is wrong to simply choose one side of the tradeoff. Afterall, the whole point of tradeoffs is that you have to consider the costs on both sides. Suppose the factory produced life-saving drugs?
Coase's insight is that you need to look at the costs of getting each to change their behavior. Suppose the homeowners could clean the river at a cost of one dollar per widget and the factory could stop polluting at the cost of two dollars per widget. Then the home owners are the least cost avoiders. The efficient outcome happens when the least cost avoiders change their demand to accommodate the other side.
This insight launched the school of Law and Economics and inspired the modern libertarian movement. See Law's Order by David Friedman, son of Milton Friedman, for a defense of this school of libertarianism. I think Friedman is correct that Justice has a family resemblance to efficiency. But it clearly breaks down in some cases. Consider slavery. There is an unavoidable tradeoff between letting slaves keep the fruits of their own labor and having that go to the slave owners. How does the Law and Economics crowd respond?
I've always assumed, wrongly it appears, that they would argue that the slaves could never be the least cost avoider. But I stand corrected. I learned in the debate that they would bite the bullet and accept slavery and genocide. I also like Brad DeLong's point that putting efficiency first gives each person a weight inversely proportional to their marginal utility of wealth. In other words, Bill Gates counts massively more than others.
Monday, November 2, 2009
Subscribe to:
Post Comments (Atom)
13 comments:
The point you attribute to De Long--that the economic definition of efficiency gives more weight to the utility of those with a low marginal utility of income--is hardly new. It's discussed, among other places, in the book of mine you cite in your post, as one of the problems with efficiency as a criterion.
On the subject of slavery, I don't think "least cost avoider" is a useful way of thinking of the problem. It's actually another example of the previous point, with one added twist.
The marginal utility of income depends in part on wealth, which depends in part on the initial allocation of rights. Starting with the slaves belonging to the slaveowners, the slaves are very poor, so their utility gets very low weight, so if slavery hurts them but helps the slave owners it might be efficient. Starting with the slaves belonging to themselves, on the other hand, they are wealthier, so their utility gets greater weight, and slavery is inefficient--as shown by the fact that, after abolition, plantation owners were unable to hire the ex-slaves for gang labor agriculture at any price they were willing to pay.
This is an example of the same point I made in the book in the context of an imaginary life extension pill.
If you think of efficiency as a proxy for utility, however, it's the second answer that is right, since even after abolition the marginal utility of income was surely higher for the ex-slaves than for the plantation owners, and probably higher than for the customers the plantations served.
It appears that David Friedman is accepting diminishing marginal utility in his comment. That is consistent with criticizing the idea that, "that the economic definition of efficiency gives more weight to the utility of those with a low marginal utility of income" and the idea that rich people (like Gates or slave owners) have a lower marginal utility of income than poor people.
This is surprising because diminishing marginal utility could revolutionize libertarian economic philosophy and parts of economics like cost-benefit analysis. Although diminishing marginal utility is a common assumption in other social sciences, it was abolished from polite discussion in welfare economics back in the 1930s and 40s. It is surprising to see a libertarian discussing bringing it back. The next thing you know he will be justifying progressive taxation!
Hanson (the one Caplan is charging with accepting slavery and genocide) denies being a libertarian economist. The modern libertarian movement predates Coase's paper.
Caplan actually agrees with Delong's take on utility (at least I take that to be implicit):
http://econlog.econlib.org/archives/2009/06/how_markets_val.html
Diminishing marginal utility is in econ 101. The first ice-cream cone is great. After the sixth, you start to feel sick. Utility functions are revealed by actions, so if rich people have lower utility from the marginal dollar they will be willing to exchange them for the services provided by others with a higher marginal utility. That is what should happen in the world of Pareto optimality.
David wrote "slavery is inefficient--as shown by the fact that, after [presumably, US] abolition, plantation owners were unable to hire the ex-slaves for gang labor agriculture at any price they were willing to pay".
The former does not follow, because the latter is a mis-statement; it should read "...willing and able to pay", effective demand. But one of the consequences of abruptly ending US slavery, even before the continuing effects of war, was to move the work force into a wider cash economy faster and in larger numbers than new liquidity could flow in (and/or prices adjust downwards) to pay new cash wages. That's why sharecropping developed rather than cash rent tenancies, too.
A better insight comes from comparing outcomes in former British colonies after emancipation, where the transition was (mostly) peaceful, slower, and compensated. In some, like British Guiana, freed slaves could set up for themselves and could not be hired back - because they had realistic alternatives. Indentured labourers from India replaced them, but former slave owners still suffered financially (as was noted in the US south). In others like Jamaica the former work force could be and was hired back; only there was the question one of plantation owners' willingness to pay for quite such intensive use of labour. Lack of willingness did show up, but not that materially.
So, which case applied in the US south? Freed slaves didn't get 40 acres and a mule, and so were still forced into (different) dependent situations. There, it was a cash flow constraint that prevented rehiring for partial versions of old work patterns.
For an example of a person today who defends slavery I present Mencius Moldbug. In response (in a later thread) I provided some economic evidence that the lot of freedmen did in fact improve after emancipation. Anecdotally, a letter from a former slave to his master makes the same point.
TGGP wrote "...I provided some economic evidence that the lot of freedmen did in fact improve after emancipation. Anecdotally, a letter from a former slave to his master makes the same point."
Unfortunately, it is more complicated than that. That only applied to slaves who were in a position to benefit from being freed, something which varied greatly with age and skills (in the broadest sense). For instance, on emancipation there was some violence in the British colonies - a minor slave revolt against it, by slaves who feared being turned out to starve Mansfield case style. That was ended when it was made clear that arrangements had been made to head that off, with a transitional period of "tutelage" at the owners' (compensated) expense during which they would be made ready to fend for themselves. That was so successful that the period was cut short and full emancipation brought forward.
To see that theirs was no mere ungrounded fear, learned helplessness (though that is none the less real for being in the mind), compare and contrast that with the fate of the coolies in the same areas who were abruptly turned loose with the abrupt end of the indentured labour system that replaced slavery where free locals could not be hired cost effectively. V.S.Naipaul records their miserable fate, turned loose without the return passages or land grants that their predecessors had got at the end of their contracts. Or consider how missionaries recorded that the two worst things that could happen to convicts in French Guiana were doublage (an administrative punishment for later offences, doubling their sentences), and to become worn out libérés, no longer fed, clothed and lodged by the authorities but still prevented from leaving the colony. Also, Richard Burton records the distress among the slaves of Sind after the British conquest brought them under British law and so forced them to be free; Napier peccavit indeed (in comparable situations the French and Belgians essentially nationalised the slaves without compensation, declaring them free as part of the mission civilisatrice but then promptly conscripting the suitable ones into a native soldiery and attaching the rest to that as support services). Even in the USA, Henry George observed how at least some former slaves felt they had become worse off as a consequence of emancipation.
This was so well understood a problem that many slave codes had provisions against freeing a slave against his will, had obligations on former owners to support freedmen if they became indigent later (rather than becoming a burden on the rates), and so on; no Mansfield case could ever occur there, because the initial turning loose would have been prevented or compensated. The Janissaries' headgear even incorporated a soup spoon or ladle as a symbol and reminder in their very uniform that - unlike the common if free people - they would never starve.
Of course, that predicament of the abandoned freed would not have been theirs had they been free all along, coming as it did mostly to those particular slaves (and coolies and convicts) who had lost the best years of their lives, so it may be accounted to the debit of slavery rather than of emancipation; but it was none the less a concomitant of badly managed emancipation. However, even the enslaving in the first place was sometimes a less worse option than many common practices in the source countries, whether in Africa or "harvesting the steppe" in the borderlands of Christian Europe; often, destruction was carried out anyway for other reasons, or to get a select few captives, and when captives were of no value - why, they were killed as bycatch (this was why it was a custom in Central Asia for the wealthy to have their children taught a trade, as they never knew what fortune would bring and the Islamic world, Al Alam Islamiye, rarely needed bulk, unskilled plantation type labour as they could and did exploit the often Tadjik peasantry).
TGGP wrote"Diminishing marginal utility is in econ 101." Diminishing marginal utility may be in SOME econ 101 texts as a simplifying assumption, but it is explicitly eliminated from more advanced courses except in limited domains such as risk preference and it is particularly avoided in welfare economics. That was the whole point of the 1934 Hicks-Allen ordinal revolution. It is hard to even find it mentioned in a public finance text.
"TGGP wrote"Diminishing marginal utility is in econ 101." Diminishing marginal utility may be in SOME econ 101 texts as a simplifying assumption, but it is explicitly eliminated from more advanced courses except in limited domains such as risk preference and it is particularly avoided in welfare economics. That was the whole point of the 1934 Hicks-Allen ordinal revolution."
Von Neumann's redoing of utility theory to cover choice under uncertainty, however, made utility again cardinal (up to linear transformations), hence put content back into the idea of declining marginal utility.
After making the previous post, I noticed the "except in ... risk preference" comment. The exception, however, swallows the rule, since if you insist on making utility ordinal for all other purposes you then have two different utility functions to do the work that one can do.
Let me confirm that diminishing marginal utility of income is a completely standard part of economics today. Whoever said otherwise is clearly way way out of touch.
"I learned in the debate that they would bite the bullet and accept slavery and genocide." "He," not "they." I attacked Robin for biting these bullets.
I take back my last comment. On closer reading, "they"="the Law and Economics crowd," not "Caplan and Hanson." Sorry.
Post a Comment