Crime Change Patterns in American Cities

A longitudinal analysis of the changes in crime rates between the late 1950s and 1970 in the 155 largest U.S. cities indicates that (1) there is virtually no relationship between the four major crimes against persons although there is strong mutual association between the four major crimes against property; (2) changes in racial composition account for more variation in changes in crime than do changes in income, thereby suggesting the need for more detailed studies of the noneconomic ways in which a city's racial composition affects crime incidence; and (3) there is a modest relationship between changes in the number of male youth and changes in certain types of crime. Since these findings are based on highly aggregated data with both known and unknown measurement errors, they must be treated cautiously.