Editors' NoteAs explained in the Introduction to this volume, the requirement that rational choice models of crime be specific to rather closely defined types of offense makes no assumptions about the degree of specialization or generalization in offenders' careers. Thus, even if an offender were a "generalist," the factors influencing decisions about some of the offenses he or she committed (residential burglary, for example) might be quite different from those factors influencing others of his or her offenses (shoplifting, for example). This being so, crime-specific modeling is needed if effective ways of preventing or controlling crime are to be developed. The virtues of rational choice approaches would be easier to communicate, however, the greater the degree of specialization that could be demonstrated. For this reason Kimberly Kempfs findings, using unrivalled data from the second Philadelphia birth cohort, of a greater degree of specialization in offending careers than has been shown before are helpful. As she points out, however, the issue is by no means settled. In particular, it would be interesting to see if greater specialization would be found if more complete information were available about the nature of offenses committed, if much finer subclassifications of offense types were used, and if offenders' careers were subdivided into more discrete periods. It is conceivable, for example, that a juvenile offender might for a few months (or even a year or two) have a small repertoire of offenses, including, let us say, shoplifting from local stores and smoking marijuana. Subsequently he or she might graduate to burglaries of local residences, and later still to mugging. Looked at over an entire career span of several years, this pattern of offending might well be characterized as generalist in nature. Considering, however, the complete range of crime types in which the offender could have become involved and contrasting this with the limited number of types actually committed at particular points in the offender's career, one could equally well make the case for a form of "serial specialization." The issues are definitional as well as empirical.