4 ms·
Separating offense from conviction reliably is impossible, so we'll never know for sure. One thing you can do though is measure more quantifiable predictors of
by goodside 16y ago
Separating offense from conviction reliably is impossible, so we'll never know for sure.
One thing you can do though is measure more quantifiable predictors of criminality, such as IQ, and compare estimates based on attractiveness to actual performance. In most such cases the answer is that ugly people are dumber than average and pretty people are smarter than average, but people's subjective estimates are biased much more strongly by attractiveness than they should be given the actual strength of the correlation. A similar phenomenon is seen for height, also.
- bd 16y agoThe effect of intelligence may be even harder to assess reliably - are smarter people simply committing less crimes or are they just harder to be found / convicted?
- goodside 16y agoThey commit less crime, or at least less violent crime. The evidence is unambiguous. They have fewer behavioral problems in school, they measure lower on biological correlates of aggression (serum testosterone/DHT and cortical norepinephrine), and, perhaps most importantly, they are less likely to be victims (most murders are committed against people of similar socioeconomic status to the perpetrator).
- bd 16y agoAll this sounds reasonable. Though given how huge proportion of all crimes goes unsolved (even violent crimes are just about 45% "cleared" [1]), it also makes sense that ones who got caught will be on average less smart than ones who got away, thus skewing statistics in a major way. [1] http://www.fbi.gov/ucr/cius2008/offenses/clearances/index.html http://www.fbi.gov/ucr/cius2008/offenses/clearances/index.ht...
- Qz 16y agoThat's funny -- at one point I attended a boarding school for people with 'behavioral problems', and every person there was incredibly smart. I call shenanigans on this idea that smart people get into less trouble.
- bdr 16y agoWow, talk about a sampling bias.