![Sample our Social Sciences journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/110822/TOC-Subject-Sample-2021-Social-Sciences.jpg"})
Abstract
Educational attainment and school bonding are established predictors of delinquent behavior. In spite of an abundance of research on the relationship between education and delinquency, there is little research that examines the impact of education on stability and change in criminal offending over the life course. This dearth of research is surprising given the increasing significance of post-secondary education in contemporary society and the prominence of the life course approach in the study of crime. The current study uses seven waves of data from the National Youth Survey to examine the impact of higher education on criminal offending over the life course. Findings indicate that college attendance and investment in higher education are negatively associated with criminal offending in adulthood. In addition, the protective effect of higher education is stronger for individuals who were more delinquent during adolescence. Study limitations and future research needs are discussed.
Notes
a Log transformed variable.
1It is possible that these youth do not necessarily represent the serious chronic offender population. Nevertheless, this subsample of delinquent youth show significantly (p < .001) higher rates of delinquency (mean = 8.018) than the rest of the sample (mean = 6.461) and include a substantial amount of serious and violent criminal offending.
2A value of 7.00 on the delinquency scale was chosen to be the cutoff value that distinguishes between those offenders who can be considered desisters and those who cannot. The baseline criminal offending value is 6.32, so this value allows for individuals to have committed one or two very minor offenses, such as being drunk in a public place, and still to be considered a desister in this study. Bushway et al. (Citation2003) point out that many studies of desistance include individuals who commit non-serious and infrequent acts of deviance are classified (perhaps inappropriately) as persistent offenders. This classification thus allows very minor and infrequent offending in an effort to avoid classifying minor and infrequent offenders as persistent offenders. Models using the zero-involvement definition and other minor variations were also assessed and results are similar (analyses not shown). See Schroeder et al. (Citation2007) for further information on this classification scheme.
3Respondents who report little or no offending at the most recent follow-up could be classified as desisters, but prior research has suggested that researchers be cautious to not include these individuals in the same category as respondents who have shown a more sustained pattern of crime cessation (see Bushway et al. Citation2003).
4The romantic partner measures are only asked to those respondents who have a spouse or a live-in boyfriend or girlfriend. The employment measures ask the respondents to report their experiences with the “major” job held in the previous year. Missing values on both measures are assigned a value of zero. Several other methods of managing missing data on these measures were also investigated, including mean substitution and listwise deletion. The results, however, were substantively similar.
5Ordinal logistic regression techniques are also appropriate techniques for investigations of life course criminal offending, but the ordinal models assume an underlying continuous variable captured by the dichotomous measures. In the current study, however, it is assumed that the three categories of offenders are qualitatively distinct and therefore do not represent an underlying continuous process captured by the variables.
c Variables are centered.
Note. Standardized coefficients are reported (+ p < .10; *p < .05; **p < .01; ***p < .001).
c Centered variable.
Note. Unstandardized beta values presented (*p < .05; **p < .01; ***p < .001).