Sumo cum laude

R. A. Hettinga rah at
Fri Jan 9 16:26:48 PST 2004


The Economist
Economics focus

Sumo cum laude

Jan 8th 2004
>From The Economist print edition

You have to respect a research programme that extends from Japanese
wrestling to "The Weakest Link"

EVERY two years the American Economic Association (AEA) awards the John
Bates Clark medal to the economist under 40 who has made the greatest
contribution to the discipline. Paul Krugman, the Princeton professor, New
York Times columnist, and Clark medallist in 1991, has rightly, if
immodestly, pointed out that this award says more about the winner's
excellence as an economist than does the better known, but rather less
exclusive, Nobel prize. That honour, after all, is sprayed out every single
year, often to more than one winner, and decades after the laureates have
done their best work. Look to the Clark medal to see who is really at the
top of the field, here and now.

At the AEA's annual meeting in San Diego this week, the award was given to
Steven Levitt, of the University of Chicago. Mr Levitt is a worthy and
unusually interesting winner. He has distinguished himself not with fancy
theory or one or two particularly important breakthroughs, but with a
series of remarkably wide-ranging and ingenious empirical investigations.

It is clear from Mr Levitt's research that he practises his discipline not
as an end in itself but because he is interested in the world. Also, if he
does not actually relish a politically incorrect finding, he certainly
never shies away from one. Sadly, both traits set him apart.

Publish and be damned

If you browse through the working papers circulated by the National Bureau
of Economic Research (at you will find that in 2003 alone Mr
Levitt wrote or co-wrote seven. His topics included the effect of school
choice on educational results; the causes and consequences of distinctively
black names; the effect of legalised abortion on crime; how to test
theories of discrimination using evidence from the television programme,
"The Weakest Link"; the gap in test results between blacks and whites in
the first two years of schooling; gambling and the National Football
League; and teachers who cheat in appraisals of their students'
performance. Among the work he has published in prestigious peer-reviewed
journals are a series of papers on crime and punishment, drug-gang finance,
penalty kicks in soccer, money and elections, drunken driving, and the
effect of ideology as opposed to voter preferences on the policies
supported by politicians. In 2002 the impeccably sober American Economic
Review published a paper co-written by Mr Levitt on corruption and sumo
wrestling. You get the idea.

There is always a serious point-usually about empirical methodology-even in
Mr Levitt's seemingly frivolous projects. The paper on sumo wrestling is
interesting not just for its persuasive demonstration that contests are
often rigged, but also for the way Mr Levitt and his co-author, Mark
Duggan, set about showing this. The incentive structure in the sport means
that wrestlers on the margin of achieving a winning record have far more to
gain from winning than opponents, not on the margin, stand to lose if
defeated. That opens the window to corruption. Wrestlers on the margin duly
win more often than their record would lead you to expect-and the authors
show that increased effort cannot be the explanation. Also, wrestlers who
win under these conditions lose more than they should when they later face
the same opponents, suggesting "repayment in kind". At times of intense
media scrutiny, evidence of fixing disappears. The research is instructive
for economists sceptical that clear-cut evidence of corruption can ever be
found. Imaginative prospecting for data, together with great ingenuity in
drawing warranted inferences from them-the Levitt hallmarks-can reveal far
more than you might suppose.

Mr Levitt's research on crime has earned him occasional spells of petty
notoriety in the wider world. The paper on legalised abortion and crime is
a case in point. Mr Levitt, along with co-author John Donahue, boldly
surmised that legalised abortion might have reduced the number of unwanted
children born to parents likely to raise criminal offspring. The evidence,
as it turned out, strongly supported that guess. The states that first
allowed legal abortion in 1970 (three years before Roe v Wade) were the
first to experience the subsequent downturn in crime; states with high
abortion rates experienced bigger reductions in crime. The data are tested
this way and that, and the conclusion stands up. The authors reckon that
legalised abortion may account for half of the fall in crime of the 1990s.

Mr Levitt has also ruffled some feathers with his work on imprisonment. A
chief finding is that prison works. It reduces crime, and not just because
it keeps people who would otherwise be committing crimes off the streets.
It also deters, something which many right-thinking people wish not to
believe. Technically speaking, a big challenge in this area of research was
to deal with a classic instance of the so-called simultaneity problem:
incarceration rates affect crime rates, but the converse is also true.
Disentangling the two relationships is impossible unless a third variable
with the right statistical properties, a so-called instrument, can be found
and exploited.

In one strand of investigation, Mr Levitt used prison-overcrowding
litigation-an improbable but statistically effective instrument-to do the
job. Properly analysed, the data then show that reducing the prison
population by one (saving roughly $30,000 a year) increases the number of
crimes committed by 15 a year (costing roughly $45,000 a year). In separate
work, equally unpalatable to many people, Mr Levitt shows that juveniles
respond to the disincentive effects of punishment in much the same way that
adults do. Kinder regimes, in other words, promote juvenile crime.

Despite such provocative and uncomfortable findings, few if any of Mr
Levitt's peers will deny he deserves the Clark medal. That may say as much
as the award itself.

R. A. Hettinga <mailto: rah at>
The Internet Bearer Underwriting Corporation <>
44 Farquhar Street, Boston, MA 02131 USA
"... however it may deserve respect for its usefulness and antiquity,
[predicting the end of the world] has not been found agreeable to
experience." -- Edward Gibbon, 'Decline and Fall of the Roman Empire'

More information about the FoRK mailing list