D. Deming wrote:
For those interested in statistical criminology, there is
an interesting article that appeared in the scholarly
journal "The Mankind Quarterly", vol. 35, no. 4, summer,
1995. The article is titled "Ideology and Censorship
in Behavior Genetics" by Glayde Whitney of Florida State
University in Tallahassee.
A most, umm, interesting journal. If I was looking for one word to
describe it I think that word would be "racist". In one of the other
issues I found an absolutely glowing and entirely uncritical review of
JP Rushton's "Big dick = little brain" theory about the relationship
between race and intelligence.
Whitney has a figure which shows that the murder rate
for the United States is actually one of the LOWEST,
when the race variable is not included. In other words,
he has compared the white homicide rate in the US with
the homicide rate in other countries which are
predominantly white (e.g., Sweden). Interestingly, in this
comparison, the highest rate is found in that enclave
of perfect socialism: Sweden. The US is actually lower
than England, Switzerland, Denmark, Scotland, France,
E. Germany, Czech., Israel, and Finland. The US rate is
only higher than W. Germany and Belgium, and only marginally
so at that.
Sigh. It must be my week for checking homicide rates. Whitney is
unable to read a table. He has grossly overstated the homicide rates
for all countries other than the United States. Here are his figures
and the correct ones from the source he cites (United Nations
Demographic Yearbook 1992)
Whitney correct
Germany 3.5 1.0
Belgium 3.8 1.6
US (whites) 4.5
England 4.5 0.5
Switzerland 4.6 not given
E. Germany 4.7 1.0
Denmark 4.9 not given
Scotland 4.9 1.5
France 5.0 1.1
Czechoslovakia 5.5 2.0
Israel 6.0 2.4
Finland 7.3 3.1
Sweden 7.4 1.3
US 9.1 9.1
Whitney argues that race rather than environmental factors determines
the crime rate because the US white homicide rate is similar that in
countries populated by white people. As can be seen by the above
table his argument is based on incorrect figures for the homicide
rate. The homicide rate for US whites is much higher than in Europe.
The difference can only be due to environmental factors. (Which might
include guns or might be something else.)
How could Whitney have got the figures so badly wrong? As far as I
can tell, he has added the figure for homicide to that for "other
external causes" and reported this as "murders". Except that he
didn't do this for the US. The reference does not report homicide
rates for Switzerland and Denmark, just a figure for homicide and
"other external causes" combined. It would be an understandable error
to report the incorrect figures for Switzerland and Denmark, but
getting the other ones wrong as well is inexcusable.
Whitney complains that his paper was rejected by Behavior Genetics
because it wasn't "politically correct". If the editor of that journal
rejected it on those grounds he/she was definitely wrong -- it should
have been rejected because it contained gross errors and and a
completely inadequate literature review. (The only study on the
relationship between race and crime he cites is JP Rushton's. He
ignores, for example, Centerwall's Atlanta study where black and white
domestic homicide rates were found to be the same once household
crowding was controlled for.)