No abstract

No abstract

Previous research on the perceived certainty of punishment indicates that individuals with experience in committing crimes perceive arrest as less certain than do those without such experience. Studies assessing the influence of experiencing formal sanctions on perceptions of risk have produced mixed results. Most studies however, have not considered the experience of sanctions in conjunction with the frequency of criminal behavior. With a sample of 1,046 incarcerated felons, we examined relationships among perceived risk of arrest, arrest history, and frequency of committing crimes. Our findings suggest that it is important to measure the ratio of arrests to crimes and that perceptions of risk are formed in a manner consistent with a rational choice perspective, even in a sample of serious offenders.

This article reports on the first results of the Second International Self-Report Delinquency Study (ISRD-2), a large international collaborative study of delinquency and victimization of 12—15-year-old students. The analysis is based on a subsample of the data set: 43,968 respondents from 63 cities and 31 countries. The prevalence rates of the major categories of delinquency, both for individual countries as well as for 6 country clusters, are presented as well as data for victimization experiences (theft and robbery/extortion). Using different measures, significant differences in level and type of offending are found between country clusters, with the Western European and Anglo-Saxon countries generally (but not always) scoring highest, followed by Northern Europe, Latin American and Mediterranean countries, with post-socialist countries at the bottom. The results for victimization experiences do not follow this pattern. The second part of the article compares ISRD-2 offending and victimization rates with two other main sources of internationally available crime-related statistics: International Crime Victim Survey (ICVS) data and European Sourcebook (i.e. police-based) data. The analyses show a moderate level of support for a convergence of different measures. The article concludes with implications and suggestions for further research.

No abstract