In the 1960's, a wave of concern about violent crime swept through the United States. Similar concerns had troubled the Nation before. The Government responded, as it had in the past, by appointing Federal commissions to study the nature, causes, and treatment of crime. Three of these commissions were The President's Commission on Law Enforcement and the Administration of Justice (1967), The National Commission on the Causes and Prevention of Violence (1969), and The National Advisory Commission on Criminal Justice Standards and Goals (1973). Each of these commissions produced a set of multivolume reports containing comprehensive reviews of the character, causes, and proposed remedies for a wide range of crime problems. Among other issues, the commissions devoted separate volumes to topics such as drunkenness, drug abuse, and juvenile delinquency, but not one devoted a full volume to youth gangs or treated gangs as a significant part of the national crime problem. Although gangs were mentioned briefly in some of the reports, all three commissions conveyed similar messages: youth gangs are not now and should not become a major object of concern; violence by youth gangs does not pose a significant threat to the populace; what violence by gangs may exist or might develop can quite easily be converted into constructive channels, primarily through social services provided by community-based agencies. The role of law enforcement in gang control was ignored.

In succeeding years, a radical change occurred. In the 1990's, government officials at all levels and the public at large became acutely aware of the reality of violent youth gangs and youth gang crime. Gang activity was perceived as pervasive, threatening, and increasing. This heightened concern over youth gangs was manifested by a wide variety of developments. Hundreds of local police departments established a gang officer or gang unit as part of their operations. Regional organizations of gang officers were established, met regularly, and exchanged information. Major Federal agencies, including the Federal Bureau of Investigation, the National Drug Intelligence Center, and the Bureau of Alcohol, Tobacco and Firearms, conducted surveys and issued reports on youth gangs. The U.S. Department of Justice established a National Youth Gang Center, and several divisions of the Department of Justice and other cabinet-level departments implemented gang programs. Regional law enforcement organizations conducted periodic surveys of gang problems in their jurisdictions. Several States organized commissions and conducted studies of gangs. There was an enormous proliferation of efforts designed specifically or in part to curb gang crime. Thousands of new or continuing projects, programs, and procedures were carried out by city and county law enforcement agencies; courts; prosecutors; corrections and probation departments; parole officers; public and private social welfare agencies; clinical and health agencies; city, county, State, and Federal governments; churches; schools; and others. The projects included:

• Recognition by the public health establishment that gang homicide is a leading cause of death among younger age groups, which resulted in responding to gang violence as a public health problem.

• Cooperation between academic researchers and Federal, State, and local agencies in conducting studies and surveys of gang prevalence and characteristics.

• Implementation of demonstration projects to test the effectiveness of different gang control strategies.

• Establishment of an academic journal devoted exclusively to gangs.

There can be little doubt that these developments were responses to major increases in the number of youth gangs and the seriousness of gang crime during the past 25 years. The perception that gangs and gang violence were increasing was widespread among both public officials and citizens. Unfortunately, this perception was based on impressions and fragmentary information rather than concrete evidence. A vital element was lacking—accurate quantitative information on the magnitude and locations of these increases. As of 1995, more than 700 academic studies and reports of youth gangs had been published, but none provided systematic, quantitative, long-term, national-level data on changes in the numbers and locations of gangs. A few studies included some trend data, but none provided a comprehensive long-term picture of national-level gang trends.¹

The major reason these kinds of trend data were not presented was the absence of a body of quantitative information on youth gang characteristics collected during a specific time period in the past that could serve as a statistical baseline against which to measure changes occurring in subsequent periods. This information was provided by the first national-level youth gang survey, conducted in the 1970's under the auspices of the U.S. Department of Justice's Office of Juvenile Justice and Delinquency Prevention (OJJDP).² Quantitative data on the number of gang-problem cities and their locations, number of gangs, number of gang members, number of homicides, and many other characteristics were provided for the period from 1970 to 1980. The trend data presented here use these statistics as a basis for measuring the scope and direction of subsequent developments.

This Report attempts to provide the missing evidence for one major aspect of the growth in gang problems—the localities where gangs are found. It addresses a series of questions about youth gang localities:

• How many localities in the United States reported gang problems during the past 25 years?

• Which localities reported gang problems?

• Where are the localities?

• How large are the localities?

• Where have youth gang problems increased during the past 25 years and where have they decreased?

• How large have these increases or decreases been?

• Which localities reported the presence of gang problems during earlier years but not during later years?

• How can these trends be explained?

• Which States, counties, cities, and regions had the largest number of cities with youth gang problems?

Viewed solely in the context of the academic study of youth gangs, such information has limited value. However, answers to the question "What can be done to reduce the growth of gang problems and the massive social costs they entail?" are of far greater value to society at large. The data presented here help answer this more urgent question in several ways.

These findings will replace the general impression of substantial increases in gang problems with concrete numerical data on the specific magnitude of the increases. This information can serve a vital purpose in establishing policy priorities. For example, it can help determine what portion of limited crime control resources should be allocated to the reduction of gang violence compared with the portion allocated to other pressing crime problems. These findings can also assist in setting priorities within the enterprise of youth gang control by identifying the "hotspots" of gang activity; the cities, counties, States, or regions in which gangs are most heavily concentrated; and the localities in which gang problems are increasing most rapidly.

Another use for these data is more directly related to issues of gang control. Identifying U.S. localities that report youth gang problems also identifies those localities that do not report such problems and raises the question of why they do not report them. As will be discussed later in this Report, there are two logical answers: either problems are present but not reported, or problems are absent. If the second answer is valid, the identification of localities (cities, counties, regions) with few or no gang problems will help researchers recognize the characteristics of localities without gang problems, compare them with localities that do have such problems, and determine how some localities avoid such problems.

Scope of the Report

A comprehensive report on youth gang trends during the past 25 years would include findings on changes in the number of youth gangs; the number of gang members; the ethnic, racial, age, and gender composition of gangs; the volume of gang crime, including gang homicides; and other topics included in the 18 general information categories examined in the first national gang survey report.³

By contrast, the scope of the present Report is limited to a single information category—localities. Four types of locality units are distinguished—States, cities, counties, and regions. In some instances, locality subtypes are distinguished. The reasons for limiting the scope of this Report to localities are primarily logistical, but this focus also has conceptual strengths.

The Growth of Youth Gang Problems in the United States: 1970–98 may be considered one of a series of updates of the 1982 National Youth Gang Survey report. Baseline data in that report allow analysis of trends in the number of gangs, the number of gang members, the number of gang homicides, and other topics. Producing a volume that covered all of these topics would be a formidable and lengthy task. The present Report limits its coverage to the presentation of detailed information on 25-year trends in the locations of gang problems with selected updates provided in the final chapter. Viewed as a research procedure, using the gang-problem locality as a major data category is probably as efficient as, if not more efficient than, using other categories such as the number of gangs or gang members in developing a comprehensive picture of long-term trends in youth gang problems.

Terms and Definitions

Before presenting the findings of this Report, it is necessary to define and discuss some of the terms used here. The major unit of analysis, as just noted, is the "youth-gang-problem locality." The term "locality" refers to the major types of named place units found in the United States and includes States, counties, cities, towns, villages, boroughs, townships, regions, and subregions. A youth-gang-problem locality is a locality for which knowledgeable authorities have reported the existence of one or more youth groups that they are willing to designate as youth gangs and that pose a recognized crime problem.

For purposes of brevity, several shorthand terms are used in the tables and text. The terms "gang city" and "gang-problem city" are used as shorthand for "youth-gang-problem city," "gang county" and "gang-problem county" for "youth-gang-problem county," and "gang locality" and "gang-problem locality" for "youth-gang-problem locality." Similar conventions apply to States, regions, and other types of localities. The terms "city," "municipality," or "municipal unit" are sometimes used to refer not only to larger populated places but also to villages, towns, boroughs, and townships. Similarly, the term "county" may refer to parishes and boroughs, which fill the functions of counties in some States. The term "gang" will sometimes be used as shorthand for "youth gang." It is important to keep in mind that when these shorthand terms are used they actually refer to the more extended and accurate terms they replace.

The term "new" when applied to a locality incorporates a time dimension. The 25-year span of the main part of this Report is divided into three periods—two decades, the 1970's and 1980's, and a 6-year period, 1990 to 1995. The final chapter updates selected data through 1998. For a designated period, a "new" gang locality is one in which a gang problem is reported for the first time during that period. This term is discussed further in the second chapter entitled "Gang Localities in the United States: A Quarter-Century Summary."

Groups Not Counted as Youth Gangs

Determining which of the many groups cited by the various reporting sources could legitimately be considered youth gangs was a prerequisite to this study. This issue is discussed in some detail in the next section. In most instances, the author had no way of knowing exactly how the reporters defined or conceived of the groups on which they reported and could not assume that all citations of youth gangs referred to similar units.

However, in examining the reporting documents, it was usually quite clear which kinds of units would not be counted as youth gangs. In conformity with the usage adopted by a national survey conducted by the National Youth Gang Center in 1995, several groups designated as "gangs," "street gangs," or "criminal street gangs" were not considered youth gangs for purposes of this Report. These groups are motorcycle gangs, including Hell's Angels and Devils Disciples; hate or ideological gangs, including Skinheads and Neo-Nazis; prison gangs, including Nuestra Familia and the Black Guerilla Family; and other types of adult gangs, including drug operations, syndicates, and organized crime gangs. A major objective was to maintain a distinction between youth gangs (ages 12 to 24) and exclusively adult gangs. The kind of unit sought and counted is the traditional area-based adolescent and young adult street gang whose violent activities include assaultive and predatory crimes. Excluding the many hate gangs, prison gangs, and other adult gangs simplifies the task of associating gangs with localities but at the same time substantially reduces the total number of gangs that are considered by this Report.

Data Collection and Analysis: Problems and Limitations

Most readers of a Report on 25-year growth trends of gang problems in U.S. localities are interested primarily in the specific findings concerning the numbers, locations, and characteristics of gang-problem localities and the ways in which they have changed during the past 25 years.

A smaller group of readers are also particularly concerned with the precise details of how the data were collected, compiled, and analyzed; the soundness of the base data; the degree of comparability between present data and the data of similar studies; and the possibility of replicating the study in order to compare its results with findings from other localities and times.

The target audience for this Report is the first group of readers, and most of the Report deals with specific findings. However, the interests of the second group are also addressed. The methods of the study are presented in two places—a relatively brief, less technical discussion in the present chapter and an expanded treatment of more technical issues, details, and procedures in three separate appendixes that focus on research methods. Readers more interested in specific findings than in methods can skim the present section and skip appendixes A, B, and C.

The major objective of this study is to replace with concrete information a general impression that gang problems in the United States have been increasing during some unknown period of time, at some unknown rate, for an unknown number of localities. The data include answers to the kind of questions listed in previous section, primarily in the form of numerical charts, tables, and graphs. This objective entailed some difficult data-gathering problems. Unlike enterprises such as the U.S. Census, which periodically collects detailed information on U.S. citizens, there is no centralized source of information on youth gangs—their numbers, their locations, or their criminal activities. Data on the location of gangs are reported by many different sources in many different places, with little uniformity in reporting methods.

Given this situation, the primary information-gathering challenge faced by this study was that of constructing a reasonably accurate and comprehensive picture of the location of youth gang problems nationwide, based on materials that were for the most part scattered, incomplete, and hard to obtain. Under these circumstances, conventional social research methods (such as the survey research techniques used in opinion polling, most sociological surveys, and the collection of census data) could not be used. Added to these difficulties were problems arising from the long timespan to be examined in ascertaining change trends during a 25-year period. Sources of information such as local law enforcement agencies rarely maintain accurate 20- or 25-year-old records that can be readily retrieved.

The inquiry, then, was unable to use the safe and familiar research methods customarily employed and required methods that would make it possible to gather information from a wide range of diverse sources that used different data-collection procedures. These methods required considerable improvisation and risk taking. Some of the problems involved are discussed in the following sections.

Definition Problems and Accuracy of Information

Any study that counts discrete entities needs a clear conception of the unit to be counted. In conventional data-gathering operations such as the Federal census, the nature of the primary unit of analysis—the individual—is quite clear. In the present instance, the nature of the primary unit is far from clear. Given the present objective of measuring trends in the number of gang-problem localities over a 25-year period, the designation and nature of the measurement unit posed a major conceptual problem—What unit is to be counted, and why? To achieve the purposes of the study, the unit is defined here as a locality in which problems with youth gangs have been reported by local authorities.

One may ask why gang localities are not counted directly instead of by secondhand or relayed information, but it would be impractical and too expensive to do so. One approach to counting gang localities directly would require the following tasks:

• Develop or adopt a definition of the term "youth gang" that would identify and describe directly comparable groups.

• Identify and count all localities in the United States with a potential for experiencing youth gang problems.

• Select from this universe a valid population or statistically representative sample.⁴

• Dispatch field workers to all chosen localities to perform the following tasks:
  • Familiarize themselves with the locality.

  • Locate all candidate groups through direct observation or reliable information from local authorities.

  • Decide whether or not the identified groups meet the agreed-on definition of youth gang.

  • Ascertain whether the groups designated as youth gangs are considered by local authorities to pose a problem.

With approximately 37,000 cities, towns, and counties in the United States, the execution of these tasks would require a massive organizational effort, a large staff, and huge expenditures. Arranging for such resources would in itself be an arduous, if not impossible, task. Moreover, the series of tasks outlined above could not be executed without a successful completion of the first task—the development of a widely accepted definition of youth gang—which could pose a major obstacle.

The definition problem is not trivial. How to define a youth gang is one of the most contentious issues in the field of youth crime.⁵ Policymakers, law enforcement personnel, social service agencies, researchers, and other groups have not been able to reach consensus on this issue during the past 25 years, and current efforts to reach this goal have thus far met with only limited success.⁶

There is little disagreement among those who study or deal with gangs that the availability and widespread use of a uniform definition would be extremely useful for a variety of important purposes, but few are willing to relinquish and replace the definitions that have become established within their agencies and are intimately related to agency operations.

It would thus appear that postponing the collection of data on the prevalence of gang problems until a uniform definition of youth gang is developed and widely adopted would preclude the acquisition of vital data. This study, therefore, uses the locality for which authorities report the presence of problems with youth gangs as a surrogate unit.

Using this surrogate results in a significantly lower degree of accuracy than would be achieved if a unit based on a widely accepted uniform definition were used. Nevertheless, the risk of receiving information on a wide variety of disparate groups in answer to the question "Does your locality have a youth gang problem?" might be considerably less than it first appears.

As part of the first major national survey of youth gangs, conducted in the 1970's (Miller, 1982), 309 respondents representing 121 agencies in 26 localities were asked for the criteria they considered essential for designating a group as a youth gang. Respondents represented a wide range of agencies and groups, including police agencies, youth outreach services, courts (e.g., judges, prosecutors, defenders), probation departments, parole operations, youth corrections and detention facilities, school security officers, public school staff (e.g., teachers, principals, guidance personnel), government and legislative officials, and current and past gang members. Respondents represented different age groups, genders, regions, residential areas, races, religions, and national backgrounds.⁷

Despite the diversity of the 300 respondents who provided criteria they considered essential to the definition of a youth gang, they showed a surprising degree of agreement. Six criteria were cited most frequently. Ninety percent of the respondents agreed on three of the six, and 85 percent on all six.⁸ Although 300 respondents generally agreed, it has been difficult to get agreement among the executives and policymakers who participated in efforts to develop common definitions of youth gangs. Most of these individuals held executive or administrative positions in their respective organizations and were thus committed to supporting the policies that were built into their organizational operating procedures and, in most cases, that had been followed for many years.

Definitional differences that would probably seem trivial or insignificant to outsiders appear highly significant to executives whose organizations have a major investment in maintaining their conventions and who would incur substantial costs if new and different definitions were adopted. By contrast, most of the individuals who provided information on gang problems for the present study held lower level positions and thus resembled the 300 respondents in the 1970's survey more closely than the higher level individuals who thus far have been unable to achieve agreement on gang-related definitions.

If one assumes that the 85–90 percent agreement level shown by respondents in the 1970's is close to the level that existed among the approximately 1,500 individuals who reported gang problems in their localities between 1970 and 1995 in the present study, the groups they designated as youth gangs, while obviously not identical, would be similar enough to constitute comparable units for present purposes. Insofar as this assumption is valid, it serves to support the accuracy of the present findings, unless and until they are disproved by further research.

More Gangs or More Information?

The substantial increase in the number of gang-problem localities during a period when the volume of information on gangs was also increasing raises the possibility that at least some part of the apparent increase resulted from the increased volume of information rather than from an actual increase in the number of gangs and gang localities. To what extent the increases represent actual developments and to what extent the increases represent more available information cannot be determined definitively. One can argue that the relationship between increased information and the actual numbers was either insignificant or significant.

The "insignificant relationship" argument maintains that when the number of gangs is relatively low, attention to gangs and the number of enterprises counting gangs are also low, leading to an undercount of the true number of gangs. When the number of gangs starts to rise, the attention level also rises, leading to more and better quality data collection and thus more accurate reporting of the true number of gangs. The "significant relationship" argument maintains that once the increase in number of gangs reaches a certain level, the increase in attention to gangs outpaces the increase in number, in turn generating more attention and more gang counting, leading to exaggerated figures on the number of gangs.

The dates of the principal data sources used in the present study clearly document an increase in the number of gang-prevalence studies. Three major reports were issued between 1975 and 1980, 4 between 1981 and 1985, 6 between 1986 and 1990, and 15 between 1991 and 1995. In addition to the increasing number of large-scale surveys, other factors that could contribute to an accelerated production of gang-related information have been suggested. These include:

• An expanded use of computers for recording and disseminating criminal information, including information on gangs.

• The greater geographical mobility of gangs, resulting in increased information sharing across State lines and among agencies.

• An increase in the number of specialized gang units and officers, generating an increased volume of information in order to accommodate the needs of the additional units.

However, several factors offset the possible impact of the increasing number of studies and volume of data on reported increases in the prevalence of gang problems. First, looking at data sources other than the major surveys, particularly media sources, reveals a more even flow of information over time. The data collected by the National Youth Gang Center's 1995 survey provide additional evidence that the growth in the number of localities with youth gangs was considerably more gradual than previous evidence had indicated (National Youth Gang Center, 1977). Second, several of the reports issued during particular time periods report the existence of gangs or gang problems during previous periods. A good example is the supplementary homicide data of the Uniform Crime Reports (Fox, 1994), which records gang killings, and thus the presence of gangs, on a yearly basis for each year between 1976 and 1992.

A third possible offset involves definition issues. During periods when gangs are less prevalent and less well defined, investigators in some localities may be more likely to count as "gangs" youth groups such as casual street corner assemblages elsewhere classified as "disruptive local groups" and not gangs (Miller, 1982, pp. 8–20). During periods when gangs are more prevalent, more readily identifiable, and easier to count, investigators are less likely to dip into the pool of nongang groups to find true gangs. If one thinks of a line dividing the many thousands of adolescent groups into gang and nongang categories, that line tends to shift to include larger numbers of youth groups in the gang category during times of lower gang prevalence and to designate smaller numbers of gangs during periods of higher prevalence.

Finally, the technical and situational factors that accompanied an increase in the volume of recorded gang information could be explained, as in the case of other factors mentioned here, either as direct responses to actual increases in gang problems or as new technologies or conditions that increased the volume of gang information relatively independent of actual increases in gang problems.

Despite these offsetting factors and arguments, the possibility remains that recorded increases in gang localities reflect, at least in part, an increase in gang reporting activity. This issue is discussed further under "Data-Collection Methods and the Growth of Gang Problems."

Possible Explanations for Growth Trends

The data presented here document an explosive growth in the number of U.S. localities reporting problems with youth gangs between 1970 and 1995. Clear evidence of so large an increase in the number of localities reporting youth gangs—a sign of commensurate growth in serious problems of gang crime and violence—leads to questions about how the growth can be explained and what can be done about it. Unfortunately, the scope of this study and the resources allotted to it make it impossible to present solidly grounded answers to either of these vital questions. To do so would require large-scale research enterprises—each with data collection and analytic methods geared specifically to the question at issue, with sizable expenditures for personnel and research resources.

Given these circumstances, one possible course of action would be simply to note the importance of these questions and exclude them from discussion on the grounds that they fall outside the scope of this study. This course of action was followed with respect to the question of what can be done about the youth gang problem. No explanation of how gang problems might be prevented or controlled is provided. However, as to the question of how the growth of gang problems might be explained, a decision was made to include a relatively brief exposition. A discussion of seven possible reasons for the growth of gang problems appears in the "Summaries and Explanations" chapter.

There are several reasons to attempt to explain the growth of gang problems. First, evidence that such large increases have occurred leads to the need to explain the increases. Second, finding effective methods of amelioration requires a solid understanding of the causes of the problems.

Explaining the growth of the youth gang problem involves a number of risks. The first concerns the character and quality of the supporting evidence. The explanations presented here represent a set of tentative hypotheses, based on the author's knowledge of relevant studies and his long personal experience with these issues.

A second risk involves the highly sensitive and controversial nature of the youth gang problem. It involves issues such as race, ethnic status, family structure, and social class—the consideration of which is significantly influenced by one's ideological orientation (Miller, 1973). Given the intensity of ideological orientations, any proposed set of explanations may be met with disagreement or rejection by some readers. Strongly held ideological convictions make it virtually certain that no body of supporting evidence, however detailed, specific, and voluminous, would have the power to ensure a reasonable degree of agreement (see Miller, 1973).

There are, however, also benefits to presenting a particular set of explanations. Disagreement, if sufficiently intense, can serve as a useful impetus for undertaking further research that would produce a better grounded and more reliable framework for explaining the growth of gang problems.

Data Sources

An extensive and detailed description of the study's data-collection methods is presented in appendix A of this Report. However, it is useful, before presenting specific findings, to include a brief outline of these methods to better understand the relationship between the data-collection methods and findings of the study.

Information on gang-problem localities was obtained from six types of data sources: National Youth Gang Survey reports, media sources, databases, interviews, conferences, and routine police reports. Three phases of data collection were distinguished: mid- to late 1970's, 1980's to mid-1990's, and mid-to late 1990's. Different combinations of sources were used during each of these periods.

Data were entered into a youth gang database starting with the year 1970 and were entered continually through 1995. Gang reports available over a multidecade period made it possible to develop the long-term trend findings reported here. In its final form, the database included about 10,000 records, making the use of a wide variety of sorting, categorizing, and analytic methods possible.

Topics

This Report describes 25-year trends in youth gang problem localities in the following substantive chapters: "Gang Localities in the United States: A Quarter-Century Summary," "Gang Cities," "Gang Counties," "Regional Trends in Gang Cities," "Trends in Size of Gang Cities," "Summary and Explanations," and "Trend Prospects for Gang Localities."