The preceding chapters cover the period between 1970 and the end of 1995. Tables, figures, and text discussions are based on prevalence figures and trend calculations derived from data collected during this 25-year timespan. During the time required for reviewing, revising, and editing the Report, new information on gang localities became available. Three national-level reports and a larger number of local reports were published either in printed form or as Internet documents. Cities, towns, villages, and counties not previously reported were added to the study's database as the new information became available. This chapter presents a selected set of tables and figures incorporating 1998 data and using the expanded time period—28 years—made possible by these data. By mid-1998, approximately 1,060 new gang cities and 450 new gang counties had been tabulated, coded, and added to the master list of gang localities shown in appendixes D and E. Table 25 specifies the sources of information for the gang-problem cities, towns, and villages newly reported during 1996, 1997, and 1998.

Table 25 shows clearly that State and national surveys provided the names of most of the gang cities—approximately 87 percent—reported after 1995. Results of four national surveys were reported—two by the National Youth Gang Center (1997, 1999) for OJJDP and one each by the National Drug Intelligence Center (1996) and the Gang Resistance Education and Training (G.R.E.A.T.) Branch of the U.S. Bureau of Alcohol, Tobacco and Firearms (1998). Of these four surveys, a clear majority of the data (94 percent) were provided by the two national surveys conducted by NYGC in 1995 and 1996. The 865 gang cities identified by these surveys accounted for about 80 percent of all gang cities reported after 1995. Information supplied by police departments and sheriff's offices provided about 10 percent of the post-1995 gang-city reports, the second largest source. State surveys conducted in Florida, Georgia, and Utah contributed another 60 gang cities.

This chapter examines some of the effects of adding approximately 1,500 new gang-problem localities (1,000 cities and 500 counties) to the major dataset of the study. A selected set of tables and figures from the first eight chapters are updated below, using gang locality data reported subsequent to 1995. This chapter consists of a limited set of selected tables and figures that document major findings.

Before presenting the new findings, it will be useful to clarify their context. Two issues will be discussed: the major objectives of the 1998 update and the relation of the updated findings to those in the rest of the Report.

Objectives of the 1998 Update

This chapter has three major objectives: first, to update selected findings on the standings of gang localities as of 1998 (e.g., the numbers and populations of localities reporting gang problems as of 1998 and their regional locations); second, to update the trend analyses (e.g., changes in the numbers and populations of reported gang localities between the 1970's and 1998 and changes in the numbers of gang localities in different population categories); and third, to readdress, in the light of new data, the "more gangs or more information" issue discussed in the chapter entitled "An Explosion of Youth Gang Problems in the United States."

Findings with respect to standings reflect the addition of the 1,500 previously unreported gang localities to the totals reported prior to 1996. Updating the trend analyses raises the issue of what baseline time period should be used. Baseline data for the 1970's derived from figures available in 1995 are less complete than those available in 1998, and all subsequent trend analyses will use 1970's figures based on 1998 data as a baseline, unless otherwise specified.

The third objective is to answer the difficult questions posed early in this Report, including to what extent increases over time in the numbers and populations of gang localities are a product of actual developments and to what extent the increases are products of increased availability of information. The discussion in "An Explosion of Youth Gang Problems in the United States" concluded that a definitive answer was not possible given the information available at the time. Although it is still not possible to provide a completely accurate and incontrovertible answer, the uncertainty associated with this issue can be substantially decreased by incorporating new information obtained after 1995. The new data make it possible to identify sources of information more accurately, obtain fuller details on data-gathering methods, and track the procedures used in reporting new gang localities with greater precision than was previously possible.

Relationship Between Findings in the Earlier Chapters and the 1998 Update

The addition of about 1,500 new gang-problem localities to the dataset that provided the basis for the analyses in the earlier chapters obviously changes the size of the dataset, enlarging it by about 70 percent. It follows that the numerical totals in the tables and figures in this chapter are significantly larger than the totals in the corresponding figures and tables in the rest of the Report. This raises a question of comparability—How does the use of the larger numerical set affect the findings presented in the Report? The simple answer is "not very much." Comparing the major findings derived from 1995 data with those derived from 1998 data reveals many differences in the figures supporting the general conclusions but few differences in the substance of the conclusions themselves. For example, the major finding earlier in the Report on changes in the number of gang cities between the 1970's and 1990's shows an increase of about 7.5 times (640 percent), while the corresponding finding in this chapter shows an increase of 9.5 times (840 percent). The specific numbers used to document the finding are less important than its basic thrust—that there was a very substantial increase in the number of gang cities between the 1970's and 1990's. Similarly, the major Report finding on changes in the populations of gang cities shows a three-decade increase of about 3.5 times, while the corresponding figure in this chapter is 2.5 times. The general magnitude of these figures falls within a similar numerical range.

Another good example of the similarities between findings earlier in the Report and in this chapter is found in the analysis of the regional distribution of gang cities, where, despite a difference of more than 1,000 cities between the rest of the Report and this chapter in the total number of gang cities in U.S. regions, the regional rankings and trend findings are almost identical. Perhaps the best evidence of the high-level degree of correspondence between the findings is that none of the numerous analyses developed in this chapter required the change of a single word in the one-paragraph summaries of the major findings.

The basic similarity between findings in the earlier chapters and in this chapter confirms and strengthens all of the findings. Because this chapter's dataset substantially enlarges the information base of the rest of the Report without significantly altering its conclusions, it enhances the depth and accuracy of the analysis. Readers should also remember that the first 8 chapters are far more comprehensive than this chapter; they are 5 times as long and contain 50 tables and figures, compared with 11 in the present chapter. The other chapters specify the central problem of the study and cover many topics not included here. The rest of the Report also includes extensive materials on data gathering and analytic methods both in the text and in a separate appendix and provides definitions for the major terms and concepts of the study. This chapter provides updated materials, corroborative evidence, and some important new findings, but the other chapters provide the major contribution of the study.

To facilitate comparability with the pre-1996 findings, updated tables and figures will include notes that refer readers to the original versions in earlier chapters.

Gang Locality Totals

The following list displays the total number of new gang-problem localities reported at any time during the period between January 1970 and June 1999, using the same format as the list in the chapter entitled "Gang Localities in the United States: A Quarter-Century Summary."

• Total number of identified gang cities, towns, and villages—2,547.

• Counties reporting gang problems, no cities specified (county sources)—42.

• Counties containing specified gang cities (city or county sources)—1,110.

• Total number of identified gang counties—1,152.

• Total number of discrete gang cities and counties—2,589.

• Total number of identified gang cities and counties (city and county sources)—3,699.

Local authorities in about 2,550 identified cities, towns, and villages and about 1,150 identified counties in the United States reported problems with youth gangs at some point between 1970 and mid-1998. These numbers represent about 7 percent of the number of U.S. municipalities in 1990 and about 38 percent of all counties. The total number of identified gang cities and counties was approximately 3,700. These prevalence figures are substantially higher than figures reported in any currently available survey and represent the largest number of identified youth gang localities ever reported in the United States.³⁹

Comparing the 1998 data with the corresponding 1995 data shows that the total number of gang cities in 1998 was 71 percent larger than the number in 1995, the number of gang counties was 63 percent larger, and the combined city-county total was 69 percent larger. The 7 percent figure for the number of gang cities, while higher than the 4 percent reported at the end of 1995, still represents a small proportion of all municipal localities in the United States. As noted in the second chapter of this Report, a much more meaningful measure of the prevalence of gang problems is the percentage of the urban population affected rather than the number of cities. At the end of 1995, the population of gang cities was about 103 million, or about 50 percent of the total city population (see figure 2); the 1998 data reported this figure to be about 122 million, about 60 percent of the total city population. Thus, in 1998, the population of cities, towns, and villages reporting youth gang problems represented a substantial majority of the total urban population.

The new gang counties reported in 1996, 1997, and 1998 constitute a total of 1,152 counties with a combined population of approximately 200 million—about 90 percent of the total county population. This number is about 10 percent higher than the equivalent figure for 1995 and indicates that only about 10 percent of the Nation's county population was free of youth gang problems in 1998.

Gang Cities

Table 26 is the updated version of table 8. Table 26 lists the number of new gang-problem cities reported for each State during the 1970's, 1980's, and most of the 1990's and ranks the 50 States and the District of Columbia according to the total number of gang cities reported for each by the middle of 1998.

Table 26 shows a cumulative total of 2,547 gang cities reported for the three-decade period—an increase of 9.4 times (843 percent) over the 270 cities reported for the 1970's. Figures in table 8, based on 1995 data, showed a cumulative total of 1,487 gang cities for the three-decade period—an increase of 7.4 times (640 percent) over the 201 cities reported for the 1970's. Thus, the more complete data available by 1998 show a larger rate of increase—9.4 times compared with 7.4 times—than the rate shown by the 1995 data. Additional information on changes in the numbers and percentages of gang cities between the 1970's and 1998 is provided in "Trends in Size of Gang Cities."

Note: The States are ranked by the cumulative number of new gang cities per State in 1998. Figures for 1998 represent only the 5 months between January and May. For corresponding 1995 data, see table 8. In a few instances, the number of cities tabulated in table 26 does not correspond exactly to the number of cities listed in appendix D because of changes in the number of localities reported subsequent to the completion of the final analyses.

Gang City Rankings

In the 1970's, the States occupying the top six ranks in the number of reported gang cities were California, Illinois, Massachusetts, Texas, Pennsylvania, and Arizona. The 1998 rankings list the top six States as California, Illinois, Texas, Florida, Ohio, and Washington. Note that in the 1970's, two of the top six were eastern States; in 1998, none were eastern. The table also indicates a major change in the concentration of gang cities. In the 1970's, gang cities in the top 10 States accounted for 90 percent of all gang cities but only 47 percent by 1998. This 43-percentage-point difference shows that in the 1970's, gang cities were heavily concentrated in the top 10 States, while in 1998, they were spread much more evenly throughout the Nation. In the 1970's, only 6 States reported 10 or more gang cities, while in 1998, 41 States reported 10 or more.

Decade Trends: Cities

Findings based on 1995 data show that there was relatively little growth between the 1970's and 1980's in the number of new gang cities, followed by a sharp increase between the 1980's and 1990's (see figure 5). These data support a widely held notion that gang-city growth accelerated suddenly and rapidly in the 1990's, following relatively little growth during the previous two decades. Findings based on 1998 data, which are more comprehensive and thus more accurate, challenge this picture of the three-decade growth curve.

Figure 20 (1998 data) shows the number of new gang cities reported for the three decades between the 1970's and 1990's, and table 27 compares these numbers with those shown in figure 5 (1995 data).

Instead of gradual growth between the 1970's and 1980's, followed by a sharp spurt between the 1980's and 1990's, these data show a much steadier and more gradual rate of growth over the three decades. The 1995 data show a 33-percent increase in the number of new gang cities between the 1970's and 1980's and a 282-percent increase between the 1980's and 1990's—a difference of 249 percent. By contrast, the 1998 data show a 137-percent increase between the 1970's and 1980's and a 156-percent increase between the 1980's and 1990's—a difference of only 19 percent. These differences are shown even more clearly by the magnitude of change figures; the 1995 data put the difference between the two decades at 2.5 times, whereas the difference shown by the 1998 figures is only 0.2 times. This difference between the finding based on 1995 data and the present finding is one of the exceptions to the pattern of similarity between findings in the rest of the Report and this chapter.

* The magnitude of change is the number of new gang cities in the later decade divided by the number of new gang cities in the earlier decade.

Gang Counties

Table 28, like its counterpart, table 10, lists the number of counties reporting gang problems for each of the 50 States and the District of Columbia and ranks the States according to the cumulative number of gang counties in each State as of 1998. The cumulative number of gang counties reported during the three-decade period (1,152) includes about 40 percent of all the counties in the United States. These 1,152 counties represent an increase of about 11.4 times over the 101 counties reported for the 1970's. The equivalent figures in table 10 (1995 data) are 73 counties in the 1970's and 706 in the 1990's—an increase of about 9.7 times. The gang county increase shown by the 1998 data was thus somewhat larger than the increase shown by the 1995 data.

These data are consistent with the finding that the geographic concentration of gang localities lessened considerably between the 1970's and 1990's. In the 1970's, counties with larger numbers of gang cities were concentrated in a relatively small number of States—principally Los Angeles and Orange Counties in California, Cook and Du Page Counties in Illinois, Middlesex County in Massachusetts, and Dallas County in Texas. The 1998 data show gang counties spread more widely throughout the States.

Note: The States are ranked by the cumulative number of new gang counties per State in 1998. Figures for 1998 represent only the 5 months between January and May. For corresponding 1995 data, see table 10. In a few instances, the number of cities tabulated in table 28 does not correspond exactly to the number of cities listed in appendix D because of the change in the number of localities reported subsequent to the completion of the final analyses.

Gang County Rankings

In the 1970's, the States of California, Texas, New Jersey, Pennsylvania, and New York occupied the top five ranks among States reporting gang counties. It is noteworthy that three of these five were northeastern States. In 1998, the top States were Texas, Georgia, California, Illinois, Florida, and Ohio. Three of these are southern States, and none are northeastern. The county data add further support to findings on the concentration of gang localities. In the 1970's, the number of gang counties in the top 10 States included more than 70 percent of all gang counties; in 1998, the top 10 States included less than 40 percent. In the 1970's, only 7 States reported 5 or more gang counties; by 1998, 48 States reported 5 or more. As noted above, gang counties, like gang cities, were spread much more widely throughout the Nation during the later years.

Decade Trends: Counties

Figure 21 displays the number of new gang counties reported during the 1970's, 1980's, and 1990's. The figure shows that 101 new gang counties were reported in the 1970's, 356 in the 1980's, and 695 between 1990 and 1998. The equivalent figures based on 1995 data were 73, 174, and 495.

Three-decade trends in the number of gang counties resemble those of the gang cities. Figure 22 compares the three-decade trend line derived from 1998 data with the equivalent line derived from 1995 data. As in the case of the gang cities, the 1995 data show a gradual increase between the 1970's and 1980's and a somewhat steeper increase between the 1980's and 1990's. The trend line based on 1998 data, by contrast, is almost completely straight, showing that the magnitude of increase from the 1970's to 1980's was almost the same as that of the increase from the 1980's to the 1990's. These data strengthen the finding that the growth of gang localities over the three-decade period was relatively even, instead of showing a gradual increase between the first and second decade and a sharp spurt between the second and third.

Gang Cities, by Region

Table 29, an update of table 15, displays the four major regions of the United States, the Northeast, Midwest, South, and West, ranked by the magnitude of change between the 1970's and 1998.

The most noteworthy feature of table 29 is how closely the general trends it portrays parallel those of table 15, although the 1998 analysis includes more than 1,000 additional cities. The ranks for 1970's cities are identical in the two tables, and the "magnitude of increase" rankings are also identical. Even the magnitude of change figures in tables 15 and 31 fall within the same general range (1995 data): South, 31.6; Midwest, 25.9; Northeast, 5.6; West, 3.2; and all regions, 7.4. Based on 1998 data, the corresponding figures are South, 33.2; Midwest, 22.4; Northeast, 8.3; West, 4.3; and all regions, 9.4. The only difference in ranking between the two tables is found in the rankings for 1998 where the South and Midwest exchange places. This similarity of results, despite differences in the specific numbers derived from 1995 and 1998 data, documents how the findings in the rest of the Report and in this chapter reinforce one another.

* The regions are ranked by magnitude of increase, which is the number of gang cities in 1998 divided by the number of gang cities in the 1970's.
Includes the District of Columbia.

Trends in Size of Gang Cities

Efforts to prevent the further spread of youth gang crime require as much information as possible on the characteristics of those localities where youth gang problems are most likely to be found. Information on the size of localities that have youth gang problems or are likely to develop or sustain gang problems is particularly valuable. Data collected between 1996 and 1998 strengthen and expand findings on gang locality sizes and population trends presented in the "Trends in Size of Gang Cities" chapter of this Report.

Table 30 updates and amplifies table 18, using 1998 data. It assigns numbers and percentages of gang cities in the 1970's and 1998 to 11 population categories, displays the magnitude and percentage increases between the 1970's and 1998, and shows the differences between 1970's and 1998 percentages.

As noted earlier, the average increase in the number of gang cities in all population categories between the 1970's and 1998 was 9.4 times (840 percent). Increases for all cities with populations larger than 50,000 fell below this average; cities smaller than 50,000 fell above the average. Cities with populations between 1,000 and 10,000 showed the largest numerical increases; the number of cities with populations of 5,000 and 10,000 inhabitants increased 27.4 times (2,643 percent), and the number of cities with populations between 1,000 and 5,000 increased 32.3 times (3,125 percent).

The largest percentage increase occurred in the 1,000 to 4,000 category. In the 1970's, this category contained 4.4 percent of all gang cities; by 1998, the percentage had risen to 15.2, a difference of 10.8 percentage points. The second largest increase occurred in the 5,000 to 9,000 category, which had 5.2 percent of all gang cities in the 1970's and 15.1 percent in 1998, an increase of 9.9 percentage points. The largest decrease occurred in the 50,000 to 99,000 category, where the percentage fell from 22.2 percent in the 1970's to 13.0 percent in 1998, a drop of 9.2 percentage points.

Note: For corresponding 1995 data, see table 18.

* The magnitude of increase is the number of gang cities in 1998 divided by the number of gang cities in the 1970's.
Populations were not available for eight cities.

The data in table 30 confirm and amplify the findings in the "Trends in Size of Gang Cities" chapter that show the numbers and percentages of gang cities in smaller size cities and towns increased much more substantially during the three-decade period than those for the larger cities, particularly in the population categories between 10,000 and 50,000. In these population categories, measures to inhibit further growth of gang problems assume the greatest urgency.

Data presented thus far have dealt primarily with various categories of gang cities in the context of the numbers and populations of those cities reporting gang problems. Table 31 examines both the numbers and populations of gang cities in the context of all U.S. cities. The table compares numbers and populations of gang cities with those of all cities in eight population categories as of 1998.

Previous analyses have focused on increases in the prevalence of gang problems in smaller cities; the data of table 31 focus on the gang-problem situation of larger cities. In the present analysis, cities with populations larger than 25,000 are considered "larger" cities and cities under 25,000 are considered "smaller."

Looking first at the numerical prevalence of the larger cities shows that, in 1998, larger gang cities constituted only 3 percent of all U.S. cities of any size but 77 percent of all larger U.S. cities. Population figures, as would be expected, show considerably higher numbers and percentages. The total population of the larger gang cities accounted for 88 percent of the population of all gang cities, 86 percent of the population of larger U.S. cities, and 52 percent of the population of all U.S. cities.

Summarizing numbers and populations of gang cities shows that larger cities made up 43 percent of the number of all gang cities but 88 percent of their population; 77 percent of the number of cities with populations greater than 25,000 but 86 percent of their population; and 3 percent of the number of all U.S. cities but 52 percent of their population. Thus, despite the impressive increases in the prevalence of gang problems in smaller cities during the three-decade study period, gang problems remained primarily a big-city phenomenon.

Note: For corresponding 1995 data, see table 21. "Cities" in this table includes U.S. cities and towns with municipal or township governments.

* Populations were not available for eight cities.

Sources: Data on numbers and populations are from Statistical Abstract of the United States, 1974, table 420, "County, Municipal, and Township Governments, 1972, and Their Population, 1970, by Population Size Groups" (p. 262); and Statistical Abstract of the United States, 1995, table 473, "County, Municipal, and Township Governments, 1992" (p. 298). Both documents were published by the U.S. Department of Commerce, Bureau of the Census, in Census of Governments, Vol. 1, Governmental Organization.

Major Findings

The substance of the data presented in the main Report was summarized in 19 major findings. The present section presents 16 additional findings. Findings that correspond to those based on 1995 data are updated using new data that became available after 1995.

• Between 1970 and 1998, problems with youth gangs were reported for 3,700 localities in the United States. These included about 2,550 identified cities, towns, and villages and about 1,150 identified counties. These gang locality figures are the highest ever reported in the Nation's history.

• During the 1970's, 270 gang cities, about 1 percent of all U.S. cities, reported gang problems. By 1998, the number of identified gang cities had risen to 2,547—about 7 percent of all U.S. cities—an increase of about 9.5 times, or 843 percent.

• In the 1970's, gang problems were reported by municipal units with a combined population of 48.6 million, or 27 percent of the total city population. By 1998, the combined population of all gang cities had risen to 122 million, about 60 percent of the all-city population. The magnitude of the increase was about 2.5 times, or 150 percent.

• The number of counties reporting gang problems rose from 101 in the 1970's, about 3 percent of all U.S. counties, to 1,152 in 1998, about 38 percent of all counties, an increase of about 11.5 times. In the 1970's, three of the five States reporting the most gang counties were in the Northeast; in the 1990's, none of the top-ranking States were in the Northeast.

• The population of gang counties rose from about 86 million in the 1970's, about 38 percent of the population of all counties, to about 200 million in 1998, about 90 percent of the all-county population—an increase of about 2.3 times, or 133 percent.

• Between the end of 1995 and mid-1998, about 1,550 previously unreported gang localities—about 1,100 cities and 450 counties—were reported by national surveys and other sources. The number of localities reported in this 2.5-year period was equal to approximately 70 percent of the gang localities reported during the 25-year period between 1970 and 1995.

• States reporting the largest number of gang cities in 1998 were California, Illinois, Texas, Florida, and Ohio. In the 1970's, eastern States occupied two of the top ranks; in 1998, none of the top-ranked cities were located in the East.

• Nationwide, there was a substantial decrease in the concentration of gang cities in the higher ranking States. In the 1970's, the top four States contained more than three-quarters of all gang cities; in the 1990's, the top four States contained about one-third. In the 1970's, only 8 States reported 5 or more gang cities; in the 1990's, all 50 States did so.

• In the 1970's, California, Texas, New Jersey, Pennsylvania, and New York ranked highest in the number of gang counties—with three of these five States in the Northeast. In 1998, the top-ranking States were Texas, Georgia, California, Illinois, and Florida, with southern replacing northeastern States as the majority among the top-ranking States.

• In the 1970's, gang counties were concentrated in a relatively small number of States, principally California and Texas. By 1998, gang counties were spread widely throughout the Nation. In the 1970's, only 5 States reported more than 5 gang counties; in 1998, 47 States did so. Gang counties were distributed more evenly than gang cities throughout the Nation. In 1998, about 40 percent of all gang cities were located in the top five gang-city States, while less than 25 percent of all gang counties were located in the top five States.

• The trend line showing the rate of increase in the number of gang cities between the 1970's and 1990's was significantly altered by the addition of the 1998 data. On the basis of 1995 data, the number of gang cities showed a relatively small increase (30 percent) between the 1970's and 1980's, followed by a sharp increase (281 percent) between the 1980's and 1990's. The trend derived from 1998 data is more balanced; instead of a rapid acceleration in the number of gang cities starting in the 1990's, the data show that the acceleration started during the 1980's and continued fairly steadily through the 1990's.

• The impact of the post-1995 data on the gang locality growth rate was even more pronounced in the case of the gang counties. The trend line representing the growth of gang counties between the 1970's and 1990's is almost completely straight, strengthening the finding that gang locality growth over the three decades was relatively gradual rather than abrupt. The magnitude of increase between the 1980's and 1990's was almost equal to that of the increase between the 1970's and 1980's.

• In the 1970's, the western region of the United States contained the largest number of gang cities, and the southern region contained the fewest. By 1998, the South had moved to second place with the largest regional growth in the number of gang cities—an increase of 33 times—while the number of gang cities in the West increased only 4 times.

• A major development during the study period was a substantial and progressive increase in the numbers and proportions of smaller gang cities. The average increase in the number of gang cities in all population categories between the 1970's and 1998 was 9.4 times (840 percent). Increases for all cities larger than 50,000 fell below this average; cities smaller than 50,000 fell above the average. Cities with 1,000 to 10,000 inhabitants showed the largest numerical increases; the number of cities with 5,000 to 10,000 inhabitants increased 27.4 times (2,643 percent); and the number of cities with 1,000 to 5,000 increased 32.5 times (3,125 percent).

• Between the 1970's and 1998, the size of the average gang city fell from 182,000 to 34,000, an 81-percent decline. The number of gang cities smaller than 25,000 rose from 35 to 57 percent of all gang cities, a difference of 22 percent, and the number of gang cities smaller than 10,000 rose from 9 to 32 percent, a difference of 23 percent. In 1998, more than one-half of the gang cities in the United States had populations of 25,000 or less, compared with about one-third in the 1970's. Of the 1,064 cities reported after 1995, almost 80 percent were smaller than 25,000.

• Despite striking increases in the numbers and populations of smaller gang cities, gang problems in the 1990's, as in the past, remained primarily a big-city phenomenon. Comparing the numbers and percentages of gang cities in designated population categories in 1998 with the numbers and percentages of all U.S. cities shows that cities with more than 25,000 inhabitants (larger cities) made up 43 percent of the number of all gang cities but 88 percent of their population; 77 percent of the number of the larger cities but 86 percent of their population; and 3 percent of the number of all U.S. cities but 52 percent of their population.

Data-Collection Methods and the Growth of Gang Problems

Data presented in this chapter, strengthening and amplifying the findings of the rest of the Report, paint an alarming picture of the extent, location, and growth rate of youth gang problems in the United States. The major findings just presented include statements that by 1998, 3,700 identified cities and counties had reported gang problems—the largest number ever reported; that the total population of all gang-problem cities included 60 percent of the total municipal population and the population of gang counties included 90 percent of the total county population; and that the number of gang cities with populations of 5,000 and 10,000 increased 32.5 times, or 3,125 percent, between 1970 and 1998.

How can one account for these unprecedented increases in the numbers and growth rates of gang-problem localities? The other eight chapters of the Report address this question on two levels; first, a set of explanations based primarily on social and cultural developments, and second, a discussion of the relationship between prevalence data and data-collection methods, under the heading "More Gangs or More Information?" This discussion concluded that in the absence of adequate supportive evidence, a definitive answer was not possible.

The availability of new information obtained during the post-1995 period created a new situation with respect to evidence on gang locality increases and data collection. Information acquired during the relatively short 2.5-year period made it possible to identify sources of information more accurately, to obtain more details on data-gathering methods, and to track the procedures used in reporting new gang localities more precisely than was previously possible.

This new information provides a basis for recasting the "more gangs or more information" issue. The central question addressed in the present section is, What is the relationship between the unprecedented growth in gang-problem localities during the past several decades and the methods used to obtain information on these localities? This revised formulation is more likely to engender concrete conclusions than the formulation in the "More Gangs or More Information?" section, and it also dovetails more closely with the nature of the new evidence.

The first major clue to understanding the relationship between information on gang-problem localities and data-collection methods may be found by dividing the decade of the 1990's into two periods—January 1, 1990, through December 31, 1995 (the last day of data gathering for the first eight chapters), and January 1, 1996, through May 31, 1998 (the reporting period for this chapter). During the first period, an average of 14.2 new gang cities per month were reported; during the second, the average was 36.5 per month, more than 2.5 times the number reported during the first period. Clearly, new gang cities were being reported at a disproportionately rapid rate during the post-1995 period.

What lay behind this striking development? Data collected during this period provide a body of evidence that helps to explain this phenomenon. Subsequent sections discuss this evidence under three headings: "Yearly Rate of Printed Reports Listing Identified Gang Localities," "Discovering Gang Cities Not Reported in the 1970's and 1980's," and "Sharp Increases in the Number of Gang Cities in Individual States."

Yearly Rate of Printed Reports Listing Identified Gang Localities

Between 1970 and 1998, approximately 50 reports listing identified gang localities were produced by various agencies and obtained by the present study. The average yearly frequency of these reports reveals a clear pattern. Between 1970 and 1990, four reports were issued, an average of 0.2 reports per year. During the period from 1970 to 1995, 26 reports were issued, raising the average to 1 per year. During the period from 1990 through 1995, an additional 22 reports were issued, an average of 3.7 per year. For the period covered by this chapter—January 1996 through May 1998—25 new reports were issued, an average of 10.4 per year.

The identities or types of these reports are presented in tables A-2 and A-3 in appendix A and in table 25. The numbers and types of agencies reporting gang localities during the most recent period were four national surveys, three State surveys, seven regional surveys, five task force reports, three commission reports, and three university-sponsored surveys.

Paralleling previously reported trends, the number of reports produced between January 1996 and May 1998 was disproportionately high compared with the other periods. Rates for this period were 13 times higher than those for the 1970-90 period and almost 3 times higher than rates during the first half of the 1990's. It is unlikely that the number of authentically new gang cities could have increased as rapidly as the number of new reports. The evidence presented here provides a reasonable basis for concluding that during the time periods examined in this Report, the more frequent the issuance of new reports, the higher the likelihood of reported increases in the number of gang localities.⁴⁰

Discovering Gang Cities Not Reported in the 1970's and 1980's

As noted earlier, there was a substantial difference between the 1995 and 1998 data in the number of cities reporting gang problems in the 1970's and 1980's. For the 1970's, 1995 data recorded a cumulative total of 201 gang cities, while 1998 data showed a total of 270—an increase of 69, or about 34 percent. For the 1980's, 1995 data showed a figure of 267 gang cities compared with the 1998 figure of 640—an increase of 373, or about 140 percent. Again raising the question of the extent to which these increases were a result of actual increases in new gang cities or a result of expanded and improved data-gathering activity, it is logically impossible that previously unknown gang-problem cities had somehow come into being during the 15 to 25 years preceding the preparation of this Report, since their designation as gang cities was based on information obtained subsequent to the 1970's. The actual number of named cities under consideration did not change after the 1970's; only their designation as gang cities was changed. It follows that these increases had to be related to data collection activity and practices. What is involved here is a process of retrieving information on previously undetected gang-problem cities.

For both decades, the increases just cited are a direct product of national surveys conducted in 1995 and 1996 by the National Youth Gang Center (1997, 1999), funded by the U.S. Department of Justice. These surveys contained questions designed to elicit information on the existence of gang-problem localities in the past. In the 1995 survey, the major question used to obtain this information was phrased as follows: "On the basis of your personal knowledge and information you may be able to obtain from persons familiar with past time periods, please indicate the time period or periods when your jurisdiction experienced youth gang problems." The 1996 survey included two questions on prior gang problems. The first was, "Have you had youth gang problems in your jurisdiction prior to 1996?" The second was, "If you answered yes to question one, approximately what year did gangs begin to pose a problem in your jurisdiction?"

Respondents in 127 of the 1,492 cities that reported gang problems in the 1995 survey also reported that gang problems were present in their jurisdictions during the 1970's. Of these, 59 had not been recorded in the major dataset of this study. In the 1996 survey, 321 respondents who reported the presence of gang problems prior to 1996 also reported gang problems for the 1970's. Of these, 10 had not been recorded in the dataset. Thus, data derived from the two NYGC surveys accounted for the entire difference (69 cities) between the figures for the number of gang cities in the 1970's reported in table 8 (1995 data) and those in table 26 (1998 data).

For the 1980's, the figure of 373 for the number of gang cities not included in the 1995 tabulations was also derived from the two NYGC surveys. In the 1995 survey, respondents representing 1,492 urban jurisdictions reported the presence of youth gang problems. Of these, 545 respondents reported gang problems in their cities in the 1980's, of which 235 had not been recorded in this study's dataset. The 1996 survey reported 898 cities with gang problems during that year, with 331 respondents naming a year in the 1980's as the time these problems were first observed. Of these, 138 were not included in the study's 1995 dataset. Adding 235 and 138 produces 373, the number of cities added to the 1980's figures on the basis of 1998 data.

Thus, the addition of almost 450 gang cities to the number of cities reported for the 1970's and 1980's on the basis of 1998 data was due entirely to information obtained by the two NYGC surveys. It was the extensive coverage and inclusion of relevant inquiries that accounted for the increase, since no real increase could have occurred.

Sharp Increases in the Number of Gang Cities in Individual States

Tables 8 (1995 data) and 26 (1998 data) rank the 50 States according to the number of gang cities reported in 1995 and 1998, respectively. Of the 2 States that entered or left the top 10, Oklahoma (33 gang cities) fell from 9th place with 33 gang cities (1995 data) to 18th place with 45 cities (1998 data). Oregon, the new top 10 entry, rose precipitously, rising from 29th place with 16 cities (1995 data) to 9th place with 68 cities (1998 data). As noted earlier, the cumulative total of all gang cities in table 26 was more than 70 percent larger than the total shown in table 8. Individual States showed larger increases. A comparison of city rankings by State based on 1995 data with those based on 1998 data shows that the State of Oregon, with 16 gang cities and a rank of 29 in table 8, added 52 cities to reach a total of 68 in table 26—an increase of 4.25 times, or 325 percent. As a result, Oregon reached a rank of nine. What accounted for this precipitous increase? Table 32 lists the four States that showed the highest percentage increases in the number of gang cities (1998 data compared with 1995 data). These States were Oregon, Utah, Iowa, and Washington. Increases ranged from 171 percent (2.7 times) for Iowa and Washington to 325 percent (4.2 times) for Oregon. The basis for the increases is revealed clearly by the data in the last column of the table, which shows that more than 90 percent of the 156 cities newly reported for 1998 were obtained by the NYGC surveys. In the case of Iowa, all of the 24 new cities were reported in the NYGC surveys. These figures provide evidence that the locality coverage of the NYGC surveys was more extensive than that of previous surveys and suggest that the precipitous increases shown here were a direct product of the design and implementation of the NYGC surveys.

* National Youth Gang Center.

Summary

A major objective of this Report has been to replace a widespread but essentially undocumented impression that crime by youth gangs was proliferating and getting worse with concrete statistical evidence. In the course of the study, several other objectives assumed increasing importance. One of these objectives concerns methods of data collection. How does one go about acquiring reliable information on the changing prevalence of youth-gang localities over an extended time period (a difficult enterprise with few established guidelines)? Another objective concerns accuracy of information. On what grounds can one evaluate collected information as reasonably accurate and complete? Do the numbers of gang-problem localities tabulated here correspond reasonably well to the actual number of such localities, past and present?

Voluminous evidence with respect to the first objective is presented throughout the Report. On the most inclusive level, the 1998 data identify approximately 3,700 gang-problem localities—cities, towns, villages, counties, parishes—that reported gang problems by mid-1998. This compares with a figure of about 371 gang localities reported for the 1970's—an increase of about 10 times. Similar increases in numbers were reported for individual States, regions, and population categories. The presentation of these data fulfilled the major objective of the Report. The second objective, the development of methods for collecting reliable information on gang-problem localities, is discussed in considerable detail in the "An Explosion of Youth Gang Problems in the United States" chapter and appendix A.

The third objective, increasing the accuracy of collected data, is the principal concern of the section entitled "Data-Collection Methods and the Growth of Gang Problems." The issue of accuracy is examined by analyzing three sets of calculations: the yearly rates of reports listing identified gang-problem localities, the process of discovering gang-problem localities known to local observers but not included in nationally disseminated reports, and the phenomenon of States that showed exceptionally large increases in the number of gang cities during a relatively short time period.

Evidence based on the first set of calculations showed that the more frequently surveys were conducted during the 1970-98 study period, the greater the likelihood that new gang localities would be recorded. The second set of calculations showed that substantial increases in the numbers of gang localities reported for previous decades resulted from more extensive coverage and greater specificity of inquiries in more recent surveys. The third set of calculations showed that some of the most striking increases in the numbers of gang localities in particular States resulted from more extensive survey coverage rather than actual increases in the numbers of new gang localities.

The implications of these findings with respect to the accuracy of collected data can be summarized as follows: the more studies conducted, the greater the likelihood of accurate data; surveys inquiring about past time periods in addition to current periods fill information gaps and provide greater accuracy; the more extensive the coverage of existing localities, the greater the likelihood of accurate prevalence figures. Increases in both the numbers and quality of more recent surveys enhance confidence in the validity of the figures presented here and strengthen the likelihood that the data in this Report accurately represent empirical reality.

How does the three-decade period preceding the year 2000 fit into the overall history of the Nation's experience with youth gangs? The landscape of gang localities is constantly shifting; past research has rightly characterized the presence of gang problems as a wax-and-wane phenomenon. The prevalence of gangs during particular time periods and in particular localities continues to change. New gangs form, existing gangs divide, separate gangs consolidate, and older gangs dissolve.

In comparing different historical periods, three logical situations can be distinguished with respect to gang problems. First, there are periods during which the number of gangs that form and the number that dissolve are relatively equal, resulting in a fairly stable number of gang localities and little change in their prevalence. Second, there are periods when the number of dissolving gangs outnumbers those coming into existence, producing a decrease in prevalence. Third, there are periods during which the number of gangs forming and continuing to function outnumbers those that are dissolving, resulting in an overall increase of gangs and gang localities.

There can be little doubt that the time period covered by this study—from the beginning of the 1970's to the end of the 1990's—fits the third situation and can be fairly characterized as a period of unprecedented growth in the number of youth gangs and the number of localities with gang problems. While there have been some decreases in particular localities and during particular times, the overall trend has been one of expanding and continuing growth.

How long will this phase continue, and what will follow it? The "Trend Prospects for Gang Localities" chapter addresses this issue. It concludes that the next phase may be a leveling off of gang problems or even a reduction in the prevalence of gang problems in the United States.

One thing is sure. In the future, the Nation will not know which of these outcomes has in fact occurred unless the practice of conducting comprehensive and high-quality gang surveys, developed during the study period, continues. The data presented in this Report will enable future surveys to measure trends in the scope and character of youth gang problems in the United States with more precision than was previously possible.