The first nationwide survey of youth gangs in the United States was undertaken in the early 1970's. Information was gathered in the course of face-to-face interviews with 67 local service workers in 12 of the Nation's 15 largest cities. In six of these cities, those interviewed agreed that their cities faced gang problems of varying degrees of seriousness. In six others, there was disagreement about the presence of gang problems. The survey was later expanded to cover a total of 23 cities and 2 counties. According to the respondents, gang problems were present in 9 of these communities, and absent in 14. Additional data collected in both the 1970's and 1990's documented the existence of almost 300 cities with gang problems in the 1970's.

By the end of the 20th century, significant changes had occurred in the youth gang situation. The number of cities reporting youth gang problems had risen from 6 to more than 25,000. The number of youth gang surveys had also burgeoned, and the Federal Government was conducting national surveys annually.

The major objective of this Report is to provide concrete information on statistical trends in the development of youth gang problems during the last three decades of the 20th century. The description and analysis of these trends were made possible by the baseline data developed by the survey in the 1970's. The survey presented information on approximately 25 topics relevant to youth gangs. Given the resources available to the present study, it was not possible to obtain numerical trend data on each of these 25 topics because of the large number of topics and the intrinsic difficulty of collecting and analyzing trend data on such topics as gang member arrests for criminal activity, including gang homicides; numbers and changes in numbers of gang members in each gang city during three decades; the number of gang members incarcerated in jails or prisons; school experience of gang members, including dropout rates; and other similarly complex sets of statistics.

Instead, the study focused on a single topic that serves as a clearly defined unit for which information is readily available—a unit familiar to all and whose definition is not controversial—and it accumulated as much information as possible for this unit. The unit chosen was the gang problem locality—a city, town, village, township, county, or parish whose knowledgeable authorities reported the existence of gang problems.

The selection of a single unit and the availability of baseline data from the 1970's made possible what previously had not been possible—a method for providing concrete and detailed information on long-term trends in the prevalence of gang problems and a solid basis for ascertaining trends in the future. Although the major unit of analysis is a simple one, the findings resulting from its use are quite complex, as shown by the many tables and figures presenting data on gang localities—their populations, regional locations, prevalence compared with all localities and with gang-free localities, prevalence trends over a three-decade period, rankings by State, concentration in counties, and growth prospects, among others.

The main body of this Report covers the 25-year period between 1970 and 1995. As originally planned, the date for ending data collection was to have been December 31, 1995. However, during the time period required for reviewing and revising the Report, several new studies containing important new data were issued. The new data, for example, indicated the existence of previously unreported gang localities whose number exceeded the pre-1996 number by almost 70 percent. In order to capitalize on the extended scope and character of these data, the study period was extended to mid-1998. The new findings are summarized in the last chapter of this Report.

The historical perspective used in this study provides evidence for a major conclusion: the United States, during a time period comprising roughly the last three decades of the 20th century, experienced gang problems in more identified localities than at any other time in history. If the past is any guide, this period, during which the number of gang localities reached an unprecedented level, will be followed by a period of reduced prevalence. Using the data and methods of this study to obtain detailed, long-term information on gang locality numbers and trends will enable future researchers to determine with considerable precision the character and magnitude of future developments and to provide reliable answers to the critical question—"Is the gang situation getting better or worse?"

Walter B. Miller, Ph.D.
Cambridge, MA