08:18 PM 5/26/2003 +0100


Dear Professor Chomsky

...

Professor Ravi Kanbur of Cornell has now published a paper on differential
mortality and poverty measurement.  The paper says essentially the same as I
had written prior to a conversation with him in 2001.   He had not
previously given any structured thought to the subject.   His paper relates
to the fact that all economists' measures of poverty look better if poor
people die.

In the paper he did not say that all measures of wealth do the same,
including average income.  ...

...economists' practice of misrepresenting income statistics as
representing gains in purchasing power, without any specific information on
price inflation for poor people's goods.   This is another huge flaw which
development economists do not want to think about.    My point is
essentially the same as Pogge and Reddy's  -  but with far wider scope.
The same basic flaw applies to all large-scale studies of ""income
inequality"".   If you do not know the inflation rate for poor people, then
you do not know whether they gained or lost in terms of consumption
adequacy.   This applies to the World Bank document claiming that ""growth is
good for the poor"".    What Reddy and Pogge wrote about  -  international
purchasing-power parity rates -  just add additional sources of uncertainty.
The inflation problem is always there.    In almost all their studies,
economists use income statistics deflated, for rich and poor, by the overall
inflation rate, and therefore do not know the value of poor people's money
to those poor people from year to year.   Staggeringly, again, it really is
as simple as that.

...I have been questioning common assumptions and the
meaning behind the jargon.    I wonder why I did not spot the inflation flaw
before.   I realised it was a general flaw after reading Reddy and Pogge's
paper on the international dollar, and then comments by economists that they
needed more price data on poor people's goods.    It seems to be the few
sentences at the end of academics' papers which give clues to the most
serious theoretical issues.

I thought you might be interested in a short version, which I attach here.


In another piece of writing I make a distinction between the economist's
definition of ""utility"" as a welfare level and the utilitarian's definition
in terms of consequences.    A macroeconomist aiming at a higher average in
a later population is not a utilitarian.    The change in income per capita
cannot tell us the average gain or loss.   A change in average ""utility""
level, in the economist's sense, cannot tell us anything about the greatest
good to the greatest number.    This confusion is, I think, why someone like
the head of the WDI (see above) could confuse the issue of whether on
average people gain or there has been just been a rise in the average.

What say about poverty measurement is this:  there are so many technical
problems in measuring economic welfare in each key aspect that not only has
it never been done, but also it cannot be done at any reasonable cost.   The
calls from economists to take into account assets, age structure, prices to
poor people and so on simply serve to highlight the fact that existing
methods of ""welfare economics"" are defective.     The cost is not justified
in relation to hungry people.   Even measuring daily food adequacy from food
surveys is in practice very complex.   The cheapest, simplest and least
corruptible measure of consumption adequacy for any government serious about
helping poor people is life length.     It is the only outcome measure
practical enough for governments of poor countries to use.    The best
outcome measure for hungry people is how long they live.


Best wishes

Matt Berkley
Oxford, England


Noam Chomsky	chomsky@MIT.EDU