The Boston Special Youth Program ran from 1954 through 1957 in Roxbury MA. It received funding from the Boston United Community Services and the National Institute of Mental Health. The goal of the pproject was to prevent delinquency with a specific focus on gang delinquency. Detached workers diverted youth from involvement in delinquency and gangs, provided services to the families of youth and provided community resources that could sustain the efforts after the funding had ended. Walter Miller, a Harvard anthropologist was hired to monitor the program and conduct an evaluation.

Miller and his team collected a voluminous amount of information about the youth, their delinquent and non-delinquent activities and their families. Seven gangs were identified in Roxbury, and each was assigned a detached worker. Four sets of records were kept: 1) contact cards that provided extensive data regarding interactions between a gang member, their gang and an outreach worker, 2) a journal of Miller's notes, 3) outreach worker notes from interviews and group meetings and 4) court records that included gang member's criminal histories including neighborhood level crime data.

More than a dozen sources were used to identify and locate the 480 project participants, including social media. Of the 231 gang members, we found 171 (74%). Seventy were deceased. We conducted interviews with 8 of these individuals. All the subjects were in their 1970's and some declined to participate, fearful of a "scam".

A second pproject was the analysis of the contact cards collected by social workers during the study. All the cards were scanned and convert to .pdf images and transcribers entered the data into spreadsheets. There are 79,671 contact cards, but a large number are duplicates. Another group of cards included information about nongang members. Our goal was to construct affiliation networks among the individuals and co-offending networks over the years of the study. After removing duplicate cards, 33,653 remained that included 166 unique individuals from the seven gangs that constitute the intensive study groups. There is considerable variation in the extent to which the full roster of gang members appear in the contact cards. For most of the gangs, the individuals in the roster appear at least once. There is also variation in the extent to which each gang constitutes the network. Three gangs account for 89% of the cards, whereas the other four gangs account for the remaining 11%. Figure 1 shows the connections between gang members for each gang. A connection between two nodes indicates that these individuals were mentioned on the same card.

The third project used historical arrest data to examine life course offending trajectories of gang and nongang members. Out of 489 intense and control group members, we had enough information to request criminal records for 384 individuals. Of the 211 intense group gang members whose criminal records we requested 27% had criminal records compared to 36% of the control group members.

The 120 individuals who possessed criminal records were arrested 1,966 times. From these records we estimated group-based trajectory models of arrest from age 8 to 70 (Figure 2). Non-offenders comprise 69% of the sample, 13% of the sample consists of very low-rate offenders who were arrested no more than twice over the life-cou8rse, often late in life. A third group, comprising 11% of the sample were moderately criminally active in their teens and early twenties, with arrests tapering off by the 40's. Eight percent of the sample were chronic offenders who were arrested on average 50 times from age 8 to 70. This small group accounted for 74% of all arrests in the sample. This group was entirely male, whereas 23% of the non-offenders were female. No gang network characteristics predicted life-course arrest trajectories.

The key finding from this study is that gang membership in the 1950's increased delinquent involvement, but the effects were smaller and of shorter duration than for contemporary gang members.

Learn more about the project and Walter Miller's unpublished book, City Gangs, at:

http://ccj.asu.edu/gangresearch

Last Modified: 11/19/2016
Modified by: Scott H Decker