The stability of self-control Michael G. Turner a, * , Alex R. Piquero b a Department of Criminal Justice, University of North Carolina at Charlotte, 232 Garinger, 9201 University City Boulevard, Charlotte, NC 28223-0001, USA b University of Florida, Center for Studies in Criminology and Law, P.O. Box 115950, 201 Walker Hall, Gainesville, FL 32611-5950, USA Abstract Gottfredson and Hirschi’s General Theory of Crime contends that low self-control interacts with opportunity to produce criminal and analogous behaviors. Although several theoretical and empirical attempts have been aimed at assessing the general theory, researchers have been slow to examine one of the central postulates of the general theory: the stability postulate. Gottfredson and Hirschi contend that once established by ages eight to ten, self- control remains relatively stable over the life-course. In the only study to address this question, Arneklev, Cochran, and Gainey found that self-control levels were relatively stable in a four month test – retest among college students. In this article, the work of Arneklev et al. is extended and examination is made of the stability postulate in childhood, adolescence, and early adulthood. Using a national probability sample, as well as behavioral and attitudinal measures of self-control, the results offer mixed support for Gottfredson and Hirschi’s stability postulate. Theoretical and future research directions are advanced. D 2002 Elsevier Science Ltd. All rights reserved. Introduction Since its publication in 1990, Gottfredson and Hirschi’s A General Theory of Crime has spurred a significant amount of theoretical (Akers, 1991; Bar- low, 1991; Geis, 2000) and empirical attention (see Pratt & Cullen, 2000). For example, scholars have examined several of the theory’s key assumptions and hypotheses including the relationship between self- control and criminal (Burton, Evans, Cullen, Olivares, & Dunaway, 1999; Grasmick, Tittle, Bursik, & Arne- klev, 1993; Keane, Maxim, & Teevan, 1993; Long- shore, 1998; Longshore & Turner, 1998; Nagin & Paternoster, 1993; Piquero & Tibbetts, 1996; Sellers, 1999) and analogous (Evans, Cullen, Burton, Dun- away, & Benson, 1997; Paternoster & Brame, 1998; Vazsonyi, Pickering, Junger, & Hessing, 2001) behav- iors, onset age and specialization (Piquero, Paternos- ter, Brame, Mazerolle, & Dean, 1999), white collar crime (Benson & Moore, 1992), and victimization (Schreck, 1999). Scholars have also debated the measurement properties of self-control (Longshore, Turner, & Stein, 1996; Longshore, Turner, & Stein, 1998; Piquero, MacIntosh, & Hickman, 2000; Piquero & Rosay, 1998). In general, empirical evidence con- firms the theory’s central hypothesis that low self- control is related to a variety of criminal and deviant acts, though it is not the sole predictor of these acts (see review in Pratt & Cullen, 2000). At the same time, scholars have been slow to empirically examine one of the most central postulates of the General Theory of Crime: the stability pos- tulate. Hirschi and Gottfredson (2000, pp. 58 – 59) contend that the stability assumption, that is ‘‘the correlation between identical measures repeated over time’’ (Jensen, 1980), derives from two empirical 0047-2352/02/$ – see front matter D 2002 Elsevier Science Ltd. All rights reserved. PII:S0047-2352(02)00169-1 * Corresponding author. Tel.: +1-704-687-6161; fax: +1-704-687-3349. E-mail address: mgturner@email.uncc.edu (M.G. Turner). Journal of Criminal Justice 30 (2002) 457 – 471