[PEN-L:8220] Re: Summers Memo

[Brad De Long <delong@xxxxxxxxxxxxxxxxx>]

>>
>>"I think the economic logic behind dumping a load of toxic waste in the
>>lowest-wage country is impeccable
>>and we should face up to that."
>
>The thing is that by the logic of neoclassical economics, Summers/Pritchett
>is absolutely right. It's an indictment of the discipline, not his/their
>personal evil. No one seems to have had the nerve to "face up to that."
>
>Doug

Actually, no. Ceci Rouse says that Larry Summers gave a very good lecture
on these points in his public finance course the year before he went to the
World Bank...

With no externalities, no increasing returns, no market power, etc., the
Walrasian equilibrium of a market economy maximizes a particular social
welfare function: a weighted sum of individuals' utilities, where each
individual's utility is weighted by the *inverse* of his or her marginal
utility of income. If indirect utilities as a function of income are
roughly logarithmic--so that each doubling of your standard of living gives
you an equal increment to utility--then these implicit weights are
proportional to your level of consumption. Thus Bill Gates--in the social
welfare function maximized by the market--has his preferences and desires
receive roughly two million times the weight that the preferences and
desires of a Bengali peasant receive in the market's implicit utilitarian
calculus.

Thus when you say that a given policy move raises real incomes (adjusting
appropriately for the disutility of work) what you are saying is that that
policy move raises the level of social welfare as calculated by the
market--with the rich having so much higher an implicit weight than the
non-rich.

Why then would anyone use average real incomes as a proxy for social
welfare? I don't think--on neoclassical utilitarian principles--that
anybody should. I prefer to take the geometric mean of consumption levels
(which is the right thing to do if indirect utility really is a logarithmic
function of income) whenever it makes a difference.

When I ask my Chicago friends why they use average real incomes as a proxy
for social welfare, I get back one of four answers:

	--that my premises are mistaken: people have a natural right to
		their property and the fruits of their labor, and that to
		carry out utilitarian social welfare calculations (whatever
		the weighting scheme you use) is a moral and political
		crime.

	--that fights over distribution tend to be negative-sum games,
		while fights to expand production are much more likely
		to be positive-sum games. Accepting the market's implicit
		weighting function and pretending that distributional
		consequences don't exist is a crude-but-effective way
		of focusing political attention on positive-sum games.

	--that the state is an executive committee for advancing the
		interests of the ruling class, and that I am a fool
		for believing that the government can carry out
		progressive redistributions. The market's distribution
		of income is as good as we are going to get, and
		must in principle be the starting point for any
		utilitarian social welfare calculations.

	--that in practice the distribution of income is very hard to
		move, and in practice the average change in relative
		distribution from economic policy will be very small.
		Given that distribution is in practice (but not in
		principle) invariant to shifts in economic policy,
		taking average real incomes as a proxy for social
		welfare points you in the right direction.

I don't find any of these four answers convincing, and think that
distribution has to be factored into any social welfare evaluation from the
very start...


Brad DeLong