Crime Trends in Southern and Nonsouthern Cities: A Twenty-Year Perspective

Abstract

Using annual data from 1951-70, for 467 U.S. cities, this study reexamines the extent to which reported crime offenses between the South and non-South continue to manifest regional differences. Alternative structuralist and normative hypotheses are evaluated with respect to both property and personal crimes. The data generally support a structural interpretation of converging crime trends with some lag among selected person crime categories. The effects of census regions and states are also examined, and here the data tentatively indicate a small but potentially increasing state effect. One of the lingering beliefs about the South as a distinct region concerns its allegedly high crime rates. Crime differentials between the South and the non-South were first noted in a group of early studies by Hoffman (1925), Brearley (1932), and Porterfield (1949), and more recently corroborated by Hackney (1969) and Gastil (1971). Likewise, Reed (1972) has amusingly entitled one of his chapters, To Live and Die in Dixie; an obvious clue to his thesis that the traditional pattern of violence in the South has persisted into more modern times. Yet in spite of this evidence, there are compelling reasons for believing that crime differentials between the South and the non-South should be and have been converging over time. The principal basis for this contention lies in the work of Porterfield and Talbert (1954), McKinney and Thompson (1965), and McKinney and Bourque (1971). These investigations argue that rapid social change in the South in recent years has had, and will continue to have the effect of dissolving hitherto existing interregional differences-e.g., crime patterns. Our purpose in this paper, therefore, is to examine in greater detail the timedependent comparisons between the South and non-South in levels of crime. In contrast to previous work whose evidence is inconclusive owing to the cross-sectional nature of their analyses, we instead employ a data set of municipal crime reports extending from 1951-70. Moreover, unlike Hackney (1969) and Gastil (1971), our attention goes beyond concern for differential homicide rates to include variations in most major criminal offenses reported to the FBI. Finally, the analysis attempts to refine our understanding of regional variations by employing a nested design of analysis of variance. *Revised version of a paper presented at the annual meeting of the Southern Sociological Society, April 1974. The research reported here was supported by funds from the Faculty Research Council of the University of North Carolina at Chapel Hill and the Council of Urban Studies Program of the University of North Carolina. I also wish to acknowledge the computer programming assistance provided by Dick Hoskins.