« A new ethnic classification | Gene Expression Front Page | Brain teaser »
October 30, 2003

Sociological Dogmas
Sociological Versus Psychological Theories of Crime

.. . . sociological theories [of crime] are not usually put in a way that makes them testable. If unemployment is a powerful factor, does it work immediately? Is there a delay, and if so how long - 5 years, 10 years, 20 years? Similar questions can be asked about poverty. And how do we define poverty? Is it absolute or relative? Arbitrary assumptions are usually made when sociologists are confronted with contrary evidence, but still there is no theory precise enough to permit exact quantitative predictions. Such facts as are available certainly do not support common-sense ideas of this kind. Differences in personal wealth - a favorite sociological cause of crime - have declined considerably since the turn of the century, but crime has gone up several hundred percent. Poverty? From 1979 to 1987 there has been a particularly steep rise in crime, but poverty has decreased dramatically (Eysenck & Gudjonsson 1989). Unemployment? Some authors have analyzed U.S. crimes and found an inverse relation between unemployment and crime. Gross national product, an indicator of national wealth? Ellis and Patterson (in press) found it to correlate positively and highly with criminality in a sample of 13 industrial nations (.68 with total theft). The evidence, if anything, is strongly opposed to sociological theories. Naturally, poverty, unemployment, and wide differences in wealth are undesirable and ought to be eradicated or at least diminished; however, doing so might increase rather than diminish crime, counter-intuitive as such a prediction, might seem. Possibly, of course, the regression is curvilinear. That is, perhaps great poverty, great unemployment, and great differences in wealth lead to low crime (as, e.g., in the early days of the Weimer Republic?); middling poverty, medium unemployment, and middling differences in wealth lead to high crime (as in the United States of America?); and little poverty, little unemployment, and few differences in wealth lead to low crime (as in Switzerland?). My point is that sociological theories are hunches rather than theories - not based on thorough statistical analysis of historical records, and too inexact to be testable.
The second argument against sociological theories is that social "causes" of crime, even if they could be proved to exist, must act through psychological pathways. Individuals react differently to poverty, unemployment, and inequality; clearly, the personality and intelligence of the individuals concerned filter objective conditions and determine their perception. Poverty may cause some people to rebel against society, blame the government, and seek refuge in crime, while others blame themselves, their lack of cognitive ability, their ignorance, and their lack of skill, and regard unemployment as just punishment. This great diversity to stress is well documented (Lazarus & Folkman, 1989); to disregard relevant factual knowledge is unscientific in the extreme. Ultimately sociology must be a part of psychology, because it studies a limited set of factors that affect human behavior through psychological mechanisms.

Excerpt taken from Personality and Crime by HJ Eysenck. Chapter in Psychopathy - Theodore Millon (editor). All emphasis mine.

Posted by Jason Malloy at 08:16 AM

Some authors found an inverse relationship between unemployment and crime? That's weird - did they include data from the 1930s or something? I've seen recent data (1980-2000) that shows a strong positive correlation. Maybe I'll go dig it up.

Posted by: bbartlog at October 30, 2003 10:47 AM

Seems like a typical case of disciplinary aggrandizement. Every discipline, sub-discipline, and research program claims to be the key to solving the riddle of the universe.

Inability to make testable predictions is characteristic of all discussions of history and human behavior. By and large the search for a deterministic social or psychological physics has been abandoned. When the dust settles evolutionary psychology will end up where the rest of them did -- some interesting insights, but not a deterministic science.

The author does not really enumerate the particular sociological theories he is rejecting, but he leaps in any case to the rejection of all sociological theories as such. From the excerpt it seems that the guy is a criminologist with an axe to grind (criminology is an applied field), and as such not too likely to come up with a powerful new statement of the relationship between psychology and sociology. His final sentence could be reversed: all psychological behavior manifests itself in a social context. (I am not an advocate of sociological determinism, btw).

Posted by: Zizka at October 30, 2003 11:13 AM

" Every discipline, sub-discipline, and research program claims to be the key to solving the riddle of the universe."

It's like Matt Groening's 7th kind of college professor: "The country that controls magnesium controls the world!"

Posted by: jimbo at October 31, 2003 07:20 AM

In James Thurber (~1940 humorist) there was a Deep Thinker who would go around saying "The next world war will be fought, not over ideology or religion or gold or oil or territory, but -- phosphorus".

Posted by: Zizka at October 31, 2003 05:12 PM

I'm convinced that the single greatest factor in the level of crime in a society is not genes are poverty but the level of enforcement. Look at Russia. 20 years ago under the Soviet Union, you were safe walking the streets even though few people lived above what we would call the lower middle class. Now, there is little strict enforcement and crime is rampant. Cuba is also in the former phase right now.

Posted by: Unstoppable at November 1, 2003 09:41 AM