Created 5/17/1996
Go to Brad DeLong's Home
Page
The Sunday Telegraph
A relatively conservative newspaper in Britain, The Sunday Telegraph,
has taken to publishing critiques of my work with Larry Summers on the sources
of comparative economic growth across nations.
While I would not say that our case is airtight, I think it is much stronger
than The Sunday Telegraph allows--and that their main lines or criticism
are dangerously flawed.
So I am working on a letter to the editor in reply, a draft of which is
below:
Letters to the Editor
The Sunday Telegraph
Dear Mesdames and Sirs:
On April 14, 1996, Bill Jamieson wote, under the title of "A Cool Look
at New Labour's Summers", a criticism of research articles Larry Summers
and I have published the Quarterly Journal of Economics and elsewhere,
and in which we presented evidence that faster GDP growth came from boosting
particular kinds of investment.
I am writing, for the record, to correct an important error made by Mr.
Jamieson. He believes that our analysis is relevant only for developing
and not for developed economies because "[w]hen the non-OECD high growth
countries were stripped out, the... analysis fell apart."
But stripping out the non-OECD countries introduces a serious bias, and
makes any analysis conducted untrustworthy, as I pointed out back in 1988
in the American Economic Review. Restricting analysis to the OECD means
that you ignore the very cases that are the most informative.
Suppose you are a Latin American country near the top of the world's GDP
per capita scale just after World War II, but low investment in equipment
has slowed your growth. Were you invited to join the OECD when Japan, Australia,
and New Zealand were admitted? No--you had become too poor. Suppose, on
the other hand, you are an East Asian country that is now nearing the top
of the world's GDP per capita scale. Do you belong to the OECD? Again, no--you
haven't been in the club of rich nations for long enough.
The procedures by which the OECD was formed and its membership enlarged
have had the systematic effect of excluding (a) once-rich non-European nations
where low investment has led to low growth, and (b) once-poor non-European
nations where high investment has led to high growth.
Is it surprising, then, that analyses focusing on the OECD do not find a
high contribution of investment to growth? I can prove that all swans are
white-if you let me throw the black ones out of the sample.
It makes as little sense to analyze growth by looking only at the OECD as
to analyze unemployment by looking only at people with jobs. In the first
case you conclude that not much affects growth (because look: all these
countries are relatively rich). In the second case you conclude that unemployment
is low (because look: very few people with jobs are unemployed). Both sets
of conclusions are wrong.
Sincerely yours,
Brad DeLong
Created 5/17/1996
Go to Brad DeLong's Home
Page
Associate Professor of Economics Brad De
Long, 601 Evans
University of California at Berkeley; Berkeley, CA 94720-3880
(510) 643-4027 phone (510) 642-6615 fax
delong@econ.berkeley.edu
http://www.j-bradford-delong.net/
Politics