This document is copyrighted by the American Psychological Association or one of its allied publishers This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. Journal of Abnormal Psychology 2000, Vol. 109, No. 3, 396-402 Copyright 2000 by the American Psychological Association, Inc. 0021-843X/00/$5.00 DOI: 10.1037/0021-843X.109.3.396 Cognitive Abilities in Adolescent-Limited and Life-Course-Persistent Criminal Offenders M. Brent Donnellan, Xiaojia Ge, and Ernst Wenk University of California, Davis T. E. Moffitt's (1993a) hypothesis that adolescent-limited criminal offenders will have higher scores on tests of cognitive ability than life-course-persistent offenders was tested with 12 tests of cognitive ability given to a large and diverse sample of delinquent juveniles whose arrest records were collected over 20 years. This is the first investigation to empirically evaluate this proposal with longitudinal data obtained from a sample for a long enough time to distinguish life course patterns of crime. This study provided only partial support for Moffitt's hypothesis because the results varied by ethnicity. We found relatively consistent support for the hypothesis for Caucasians and Hispanics but no support for the hypothesis for African Americans. These findings are interpreted in terms of differences in developmental contexts for individual ethnic groups. The relationship between cognitive ability and crime has long been recognized by social scientists. In a comprehensive review, Hirschi and Hindelang (1977) summarized work on differences in cognitive abilities between delinquent juveniles and nondelinquent juveniles. They concluded that, on average, delinquent juveniles score significantly lower on tests of cognitive abilities than non- delinquent juveniles. Other researchers have reported similar IQ differences between delinquent and nondelinquent juveniles in longitudinal studies of crime and delinquency (e.g., West & Far- rington, 1973; Wolfgang, Figlio, & Sellin, 1972). A similar relationship between cognitive abilities and crime has also been found in more recent studies. Kandel et al. (1989), for example, reported that men who had an antisocial parent and lower IQ scores were at greater risk for criminal activity than other men. White, Moffitt, and Silva (1989) found similar IQ differences in the Dunedin sample, even when familial criminality was statisti- cally controlled. These authors interpreted this finding to suggest that low IQ by itself is a risk factor for criminal activity, a conclusion also drawn by Farrington (1995) on the basis of data from the Cambridge Longitudinal Study. Moreover, Moffitt and Silva (1988) demonstrated that individ- uals with lower IQ scores self-report more delinquent acts than individuals with higher IQ scores. Because they did not use official police records to measure delinquency, their results suggest that M. Brent Donnellan, Xiaojia Ge, and Ernst Wenk, Human and Com- munity Development, University of California, Davis. This study is based on data originally collected by Ernst Wenk, who was supported by a grant from the National Institute of Justice (Award No. 90-IJ-CX-0061), a Guggenheim Fellowship, and Arizona State University. We acknowledge financial support from the University of California, Davis, and from the California Experimental Station (CA-D*-HCD-6092- H). Portions of this paper appeared in an earlier form as a poster presented at the Seventh Biennial Meeting of the Society for Research on Adoles- cence in San Diego, California, on March 1, 1998. Correspondence concerning this article should be addressed to M. Brent Donnellan, Department of Human and Community Development, Univer- sity of California, Davis, One Shields Avenue, Davis, California 95616. the IQ-crime relationship is not simply a matter of greater police detection of less intelligent offenders. Furthermore, Lynam, Mof- fitt, and Stouthamer-Loeber (1993) found that the self-reported IQ-delinquency relationship holds while controlling for ethnicity, socioeconomic status (SES), and IQ test motivation in the Pitts- burgh Youth Study. Although most reviewed studies provide evidence for differ- ences in cognitive abilities between criminals and noncriminals, a dual taxonomy recently proposed by Moffitt (1993a) argues that differences in cognitive ability may exist between criminals who engage in criminal activity during their adolescence and those who engage in life-long criminal activity. According to this proposal, life-course-persistent criminals are more likely to be characterized by a "neuropsychological deficit" that can be detected by dimin- ished performance on tests of cognitive ability. This neuropsycho- logical deficit helps to channel the individual into criminal ability. and prevents him or her from finding viable alternatives to crime. In contrast, the other type of criminal offender is classified as adolescent limited because it is made up of delinquent juveniles who engage in transient antisocial behavior during their adoles- cence. These adolescent-limited offenders, according to Moffitt (1993a), do not suffer from the lower cognitive abilities that characterize life-course-persistent offenders. As such, these ado- lescents do not have cognitive liabilities that can prevent them from desisting from criminal behavior once they reach adulthood. Moffitt's (1993a) theoretical proposal has not been tested em- pirically, however, largely because of the dearth of investigations that have monitored individuals long enough to distinguish Mof- fitt's typology of criminal careers. This study was designed to fill this gap in knowledge by testing a hypothesis derived from Mof- fitt's (1993a) proposal that chronic criminal offenders will have lower scores on tests of cognitive ability than adolescent-limited offenders. The study was based on data from 20 years of arrest records from a large and diverse sample of juvenile offenders who were first assessed between 1964 and 1965 with a comprehensive battery of cognitive tests. The unique features of this data set provide a rare opportunity to explore the links between crime and cognitive ability for at least 396 COGNITIVE ABILITIES AND CRIME 397 three reasons. First, this investigation monitors sample members into their 30s, making it possible to clearly differentiate Moffitt's (1993a) two types of criminals. With several notable exceptions (e.g., Farrington, 1995; Robins, 1966; Sampson & Laub, 1993), few longitudinal studies of crime and delinquency have monitored sample members into their late 30s, making the evaluation of long-term differences in criminal careers difficult (i.e., differences between truly life-course-persistent offenders and adolescent- limited offenders). Second, whereas previous studies of crime and cognitive ability have typically used only one or two measures of cognitive ability, such as verbal IQ, this study examines a wider spectrum of cognitive abilities assessed with multiple instruments. The breadth of coverage from diverse measures provides an op- portunity to evaluate the robustness of cognitive ability differences between criminal career patterns across different tests. Third, whereas previous research has primarily investigated links be- tween cognitive ability and crime in mostly ethnically homoge- neous samples (e.g., Kandel et al., 1989; Moffitt, Gabrielli, Med- nick, & Schulsinger, 1981; Moffitt & Silva, 1988; West & Farrington, 1973; White et al., 1989; for exceptions, see Lynam et al., 1993; Wolfgang et al., 1972), this study was conducted with an ethnically diverse sample. The large size and diversity of the sample allow us to test the research hypothesis within separate ethnic groups. This feature is particularly important because pre- vious research and theory on the links between crime and cognitive ability have largely ignored the role of ethnicity. She of its allied publishers. r and is not to be disseminated broadly.