Intelligence and Crime

The central question of IQ-crime studies is whether individuals with less intelligence, on average, commit more crime than those with more intelligence. That is, are IQ and crime negatively correlated? The best answer, drawn from previous research, is a qualified "yes." Delinquents and criminals average IQ scores 8 to 10 points lower than noncriminals, which is about one-half a standard deviation. IQ and criminal behavior are negatively correlated at about r = -.20 (Hirschi and Hindelang; Wilson and Herrnstein). Here are five well-known studies that illustrate the correlation between IQ and crime.

Terrie Moffitt and colleagues studied 4,552 Danish men born at the end of World War II. They examined intelligence test scores collected by the Danish army (for screening potential draftees) and criminal records drawn from the Danish National Police Register. The men who committed two or more criminal offenses by age twenty had IQ scores on average a full standard deviation below nonoffenders, and IQ and criminal offenses were significantly and negatively correlated at r = -.19.

Donald Lynam and colleagues studied 430 seventh-grade boys in Pittsburgh, Pennsylvania. They measured both IQ and self-reported participation in delinquent acts. Those boys who committed serious delinquent acts, such as stealing cars, breaking and entering, or selling drugs, scored 8–10 IQ points lower than boys who had not. IQ scores and delinquency were correlated at r = -.22, with the correlation between verbal IQ and delinquency being much stronger than the correlation with performance IQ (r = -.33 versus -.06).

Hakan Stattin and Ingrid Klackenberg-Larsson followed 122 Swedish males from ages three though thirty. They measured IQ at ages three, five, eight, eleven, fourteen, and seventeen and counted the number of registered criminal offenses through age thirty. Frequent offenders, those men with four or more criminal offenses, averaged IQ scores of only 91 points; sporadic offenders averaged 97 IQ points; and nonoffenders averaged a full 102 points. Remarkably, IQ at age three significantly correlated with registered crime at (Spearman's) rho = -.25. IQ at the later ages also correlated with crime at around rho = -.20.

Scott Menard and Barbara Morse studied 257 high school students in San Diego, California, measuring both IQ and self-reported delinquency. IQ was correlated with nonserious crime—such as petty theft, liquor violations, vandalism, truancy, and running away—at r = -.08. IQ was correlated with serious crime—such as gang fights, auto theft, grand theft, and robbery—at r = -.16.

Deborah Denno analyzed data from 987 African American school children in Philadelphia. Her data contained multiple measures of intelligence collected at ages four, seven, and thirteen as well as officially recorded criminal offenses. Chronic, violent offenders consistently had low IQ scores. For example, female chronic offenders were almost four times less likely to be in the top third of verbal-IQ test scores than female nonoffenders. Similarly, male violent offenders scored 10 to 17 percentile points lower on measures of vocabulary, reading, and language than nonoffenders.

In addition to finding a robust IQ-crime correlation, studies have turned up two other empirical regularities worth noting. The first regards two different types of IQ measures: performance IQ (PIQ) versus verbal IQ (VIQ). Performance IQ is measured with nonverbal tests of attention to detail, manual design construction, and visual puzzle solving. Verbal IQ is measured with tests of general factual knowledge, abstract reasoning, mental arithmetic, and vocabulary. Studies have consistently found that criminals have PIQ scores close to the general population but VIQ scores substantially lower. This PIQ > VIQ finding holds even when controlling for race, class, and reading ability (Moffitt), suggesting that verbal intelligence is a more important correlate of criminal behavior than other types of intelligence.

The second regularity regards official versus self-reported measures of crime. While IQ consistently correlates with both of these measures, the correlation between IQ and official measures tends to be somewhat stronger than the correlation with self-reported crime (Hirschi and Hindelang).