*This report was originally written in October, 1997 and subsequently published by the Office of Juvenile Justice and Delinquency Prevention (go to the following link to read this version of the report:: I have updated some of the original citations. A similar program has been launched in San Francisco, focusing on post-adjudicated youth.  Called “New Options,” I will be posting an article about this on my web site soon.




This evaluation consists of a comparison between a random sample of cases referred to DDAP between June, 1993 and December, 1994 (n=271) and a control group of youths placed in detention who were not referred to DDAP (n=271).  The follow-up period was approximately three years.  The independent variables used for this comparison include the following: age of referral, race, gender, risk scores, recidivism, previous referrals, the nature of previous referrals, previous placements, subsequent referrals, subsequent violent offenses, subsequent placements, and subsequent sustained petitions. 

The most important of the preliminary findings are as follows:  


                                                             MAJOR FINDINGS

I. Group Differences


~         The overall recidivism rate of the DDAP group was 34%, compared to a rate of 60% for the control group;


~         The risk scores for the DDAP group were significantly more likely to be greater than 10 than was the case for the control group (86% vs. 59%);


~         No differences were found between the two groups as far as the racial distribution was concerned; 89% of the DDAP group was minorities, compared to 86% of the control group;


~         There were some significant gender differences, as 23% of the control group were girls, compared to only 16% of the DDAP group;


~         There was also significant age at referral differences, as 27% of the control group were 14 or under, compared to 15% of the DDAP group;


~         Whereas 39% of the control group had three or more prior referrals, only 20% of the DDAP group did;


~         Interestingly, there was no difference between the two groups as far as the nature of the prior referrals.  Specifically, they were about equally as likely to have been previously referred to court for a serious violent crime (23% of the DDAP group vs. 28% of the control group) and for some other serious crime (30% of the DDAP group vs. 24% of the control group);


~         While 27% of the DDAP group had at least one prior placement, only 16% of the control group did;


~         There were significant differences between the two groups as far as the number of subsequent referrals is concerned; thus, only 14% of the DDAP group had two or more subsequent referrals, while half of the control group did;


~         However, the two groups did not differ significantly in the likelihood of receiving subsequent placements, although the control group was slightly more likely to (18% for DDAP vs. 24% for the control group);


~         The two groups did differ very significantly in terms of the likelihood of being returned to court on a violent crime.  Thus, whereas only 9% of the DDAP group came back to court on a violent crime charge, one-fourth of the control group did;


~         Finally, the two groups differed significantly in terms of the number of subsequent petitions were concerned: while only 5% of the DDAP group had two or more subsequent petitions, 22% of the control group did;


~         Taking referral age into consideration, and realizing that the higher recidivism rate of the control group could be attributed to the fact that they were more likely to be younger, and thus having more time to come back on subsequent referrals (and conversely older youths may not have returned simply because they became adults), age was held constant and comparisons were then made between the two groups.  Significantly, when holding referral age constant, the control group still had a much higher recidivism rate (among those 14 and under, the recidivism rate for the control group was 74%, compared to a 44% rate for the DDAP group; among those 15 and older, these rates were 55% and 33% respectively);


~         Holding constant such variables as risk scores, prior referrals, prior placements, the nature of previous offenses, race, and gender, with only one exception, the control group had a higher recidivism rate (the one exception was that when considering girls only, the differences were not significant, even though the control group had a higher recidivism rate); thus, for instance, among those with high risk scores, the recidivism rate for DDAP was 33%, compared to 58% for the control group; three or more prior referrals, the recidivism rate for the DDAP group was 50%, compared to 71% for the control group; among black youths, the recidivism rate for DDAP was 39% and 66% for the control group; in fact, one of the most significant findings is that risk scores did not predict recidivism and that, in fact, those with risk scores of 10 or more had a lower recidivism rate (45%) than those with risk scores of less than 10 (55%), although this was not statistically significant;


~         It is interesting to note that even among youths you would predict would not become recidivists (e.g., those with no prior referrals, those with minor prior offenses, those with no previous placements), the recidivism rate for the control group remains significantly higher than for the DDAP group.


~         This research suggests that the apparent success of DDAP may stem largely from the fact that it exists outside of the formal juvenile justice framework; that is, it is a true alternative to the normal institutional framework.


II. Other Noteworthy Findings


~         When using subsequent petitions as a measure of recidivism, instead of subsequent referrals, the DDAP group had a lower overall recidivism rate than the control group (24% of the DDAP group had at least one subsequent petition to court vs. 48% of the control group);


~         If we use subsequent placements as a measure of recidivism, the DDAP group remained more successful, as only 18% of this group had at least one subsequent placement, compared to 24% of the control group, although this relationship was not statistically significant;


~         Risk scores were significantly related to previous referrals and the nature of previous offenses (those with risk scores in excess of 10 were more likely to have prior referrals and more serious prior referrals);


~         However, risk scores did not predict subsequent violent referrals, subsequent petitions to court and subsequent placements.


~         Background data on the DDAP group reveals the following noteworthy characteristics, all of which would predict a strong likelihood of becoming chronic offenders:


~         They tended to come from one of four mostly poor neighborhoods, with the general area of Potrero Hills and Bayview Hunter areas being the most common;


~         The most common family arrangement was living with the mother;


~         Only a slight majority were attending school at the time they entered DDAP and about one-fourth had been either expelled or suspended at least once;


~         Drug use was common, with marijuana being the most commonly used; almost half had used within 90 days prior to entering the program and about one-third used drugs on a daily basis; 


~         Just over 40% were living in poverty conditions, indicated by notations on the intake forms that they were living in government housing or receiving some sort of public assistance.


~         DDAP clients were not minor offenders with little previous contact with the juvenile justice system, as the majority had at least one prior referral and the nature of their prior offenses were usually serious, with over half having committed a serious offense (mostly drugs and property crimes), while their most recent offense was about equally distributed between serious violent crimes (32%) and drug offenses (26%); just over one-fourth had previous placements; not surprisingly, many had very high risk scores: 34% had scores of 15 or more, with 10% had scores of 20 or higher.


~         One final significant finding is that, for the DDAP group the strongest predictor of recidivism was poverty.  


~         In reviewing the case files it was discovered that there was in too many instances missing data from the intake forms and also that certain kinds of information that should be part of the intake form were not being collected (e.g., background information about the parents and other family members).





An Assessment of the Detention Diversion Advocacy Project:  Final Report


Part I: Review of the Literature


Juvenile delinquency has continued to be viewed as a major social problem, especially in more recent years as more and more young people are joining gangs and engaging in violence.  The solutions being offered by politicians and criminal justice officials continue to be, with few exceptions, variations on conservative models of crime control, namely, “getting tough” on delinquents through greater use of incarceration and the certification of delinquents as adults.

One of the major consequences of this approach has been the overcrowding of virtually all the major correctional institutions in America, both at the adult and the juvenile level (Krisberg and Austin, 1993).  Not only this, previous research has clearly demonstrated that the punitive approach to delinquency is, in the majority of cases, nonproductive and may actually create more problems (Lemert, 1951; Schur, 1971; Schwartz, 1989; Krisberg and Austin, 1993; Miller, 1990).  

It should also be noted that the recent get tough policies have resulted in greater and greater proportions of minorities being incarcerated within the juvenile correctional system (Krisberg and Austin, 1993; Pope and Feyerherm, 1993; Wordes, Bynum and Corley, 1994).  The study by Wordes, Bynum and Corley is especially relevant here for they addressed the specific issue of minority overrepresentation in detention populations.  Analyzing data from five Michigan counties, they discovered that even when several other variables were considered, minorities were more likely to be detained.

Several alternatives to the use of secure facilities are being attempted. Most of these alternatives are part of what as come to be known as “diversion” programs.  As noted below, however, these are not really “alternatives,” in many cases, because youths who are “diverted” to them might not have been formally handled under normal circumstances.  Such is an example of what many have called “net widening” (Frazier and Lee, 1992; Rojek, 1982). 




Diversion is "an attempt to divert, or channel out, youthful offenders from the juvenile justice system" (Bynum and Thompson, 1996:430).  This is based upon the belief that juvenile justice system processing does more harm than good (Lundman, 1993:89).  The basis of the argument is that courts may inadvertently stigmatize some youths for relatively petty acts which might best be handled outside the court system, or ignored entirely.  Diversion programs are designed, in part, to deal with the problem of overcrowded juvenile courts and correctional institutions (including detention), so that greater attention can be devoted to more serious offenders.

The diversionary tactics have a strong theoretical background which is based firmly on “labeling” principles.  These principles have evolved from Tannenbaum who initially wrote in 1938 on the “dramatization of evil,”  and Becker's notion that social groups create deviance by labeling acts as “deviant” and treating those so labeled as “outsiders” (1963), and Lemert's classic statements about labeling leading to "secondary deviance" (1951).  Thus, the legal interaction by the juvenile justice system may actually perpetuate delinquency by processing cases of children and youths who's problems might be ignored, normalized in their original settings, or better dealt with in more informal settings within the community.

One of the most immediate responses to the labeling perspective was the President's Commission on Law Enforcement and Administration of Justice report (1967) that called for the creation of “Youth Services Bureaus” to develop alternative programs for juvenile offenders in local communities.  The establishment of these Youth Services Bureaus began a move toward diverting youths, especially status offenders and other non-serious delinquents, from the juvenile court.  These bureaus were quickly established in virtually every community regardless of size.

Unfortunately, the Youth Services Bureau concept was far from clear and unambiguous.  As Gibbons and Krohn have observed:  “For one thing, the recommendation that community services be coordinated by the bureau assumed that there was a wealth of services to be coordinated when, in fact, the lack of such agencies and services had been an impediment to successful juvenile court work” in the first place (1991:313).

It should not be surprising that conflicting expectations, findings, and conclusions would emerge from such a widespread, disjointed, and complicated social experiment.  There are many studies which show that diversion programs are successful in reducing subsequent deviance. These findings are balanced at least equally, however, by findings of no impact.  There are, additionally, reports that find diversion programs to have detrimental properties (Polk, 1995).

Proponents of diversion programs cite numerous studies such as the diversion project in Colorado that involved comparisons between an experimental group of diverted youths and a control group who received regular juvenile justice system handling.  The diversion program administered individual, parental, and/or family counseling to the diversion cases, resulting in significantly lower recidivism rates (Pogebrin, Poole and Regoli, 1984; see also Frazier and Cochran, 1986; Gilbert, 1977). 

The Adolescent Diversion Project (Davidson et al. 1990) was a large scale diversion project in Michigan that in addition to status offenses, included juveniles accused of criminal acts.  This study concluded that diversion can safely be extended to other than status and minor offenders.  Although most of the project's offenders admitted to criminal acts, each of the various diversion programs reported lower recidivism rates than normal court processing cases.

The most successful diversion projects have been those that provide more intensive services, which is consistent with Dryfoos' findings that successful programs involve more comprehensive services (1990).  Especially important is the use of experienced youth workers. For example, a project in St. Louis found that the youths who were under the direction of the most experienced youth workers were able to bring about greater behavioral changes on the part of their youths, while the inexperienced youth workers did not (Feldman, Caplinger, and Wodarski, 1983).

Opponents of diversionary projects, on the other hand, cite studies that show diversion projects to be unsuccessful (Rojek & Erickson 1981-82). In one analysis of a police diversion program, it was discovered that diversion served to aggravate rather than deter recidivism (Lincoln 1976).  Elliott and his colleagues (1978) report that receiving service, regardless of whether the intervention was in a traditional juvenile justice setting or in some alternative program, resulted in an increase in the youths' levels of perceived labeling and self reported delinquency.  This finding was supported by Lincoln's study (1976; see also Lipsey et al., 1981).

The issue of increased or decreased recidivism rates is coupled with the concerns over prejudice, discrimination, civil rights violations and net widening.  The issue of net widening has perhaps received the most attention.  Ideally, a true diversion program (and the original concept behind diversion) seeks to take those who would ordinarily be processed within the juvenile justice system and place them into some alternative program.  So, for instance, if normally you would have 1,000 youths processed within the system, a true diversion would be to take, say, 300 of these youths and place them in some alternative.  Essentially, “net widening” would occur if (using these same numbers) the alternative programs would be serving 300 that would not have heretofore been part of the 1,000 normally processed.  Thus, instead of having 1,000 youths being dealt with (300 in diversion programs and 700 within the juvenile justice system), you have instead a total of 1,300 being processed (1,000 + 300).  In this example, you have a “net gain” or a “net widening” of 300 youths.

Part of the net widening problem stems from the fact that as many as half of all referrals have come, not from the police and juvenile court intake officers, but from schools, parents and welfare agencies.  Ironically, these are the people and local agencies that were supposed to serve youth in the first place, in lieu of legal processing (Polk, 1995:372).

Two additional problems include certain constitutional issues and charges of differential treatment based upon race.  In the case of the former, Bullington et al. (1978) found that one of the biggest problems of diversion involved what often amounts to “disposition without adjudication.”  As for the race issue, Bortner et al. (1985) found evidence of systematic differential treatment of blacks, with black females, for instance, being more likely to be incarcerated for status offenses, while their white counterparts were more likely to be diverted elsewhere.

If diversion is to work effectively, it is of paramount importance that youths are really diverted from the system itself, rather than being part of some net widening process.  This becomes especially important when dealing with the ever-increasing problem of overcrowding within the detention system of the juvenile justice system.  If a particular detention center (or any other correctional center) is plagued by chronic overcrowding, then the obvious solution is to either (1) increase the space available (e.g., add rooms to the current structure or build a new one) or (2) remove a certain percentage from the correctional system in question and deal with them in some true “alternative” program.  The program being evaluated here seeks to do the latter.


Part II: The Detention Diversion Advocacy Project (DDAP)


The Detention Diversion Advocacy Project (DDAP) is a program sponsored by the Center on Juvenile and Criminal Justice in San Francisco, California.  The program's major goal is to reduce the number of youth in court ordered detention and provide them with culturally relevant community-based services and supervision.  Youths selected are those that are likely to be detained pending their adjudication.  DDAP provides an intensive level of community-based monitoring and advocacy that is not presently available.

Disposition case advocacy is the concept that describes the type of approach being used in this program.  This method has been defined as “the efforts of lay persons or nonlegal experts acting on behalf of youthful offenders at disposition hearings” (Macallair, 1994:84).  It is based in part on the more general concept of "case management" which has been defined as a “client-level strategy for promoting the coordination of human services, opportunities, or benefits.”  Case management seeks to achieve two major outcomes: 1) “the integration of services across a cluster of organizations” and 2) continuity of care (Moxley, 1989:11).  The main focus of case management is to develop a network of human services that integrates the development of client skills and the involvement of different social networks and multiple service providers (Moxley, 1989: 21).

The program is designed to accomplish the following goals:


1.         to provide multi-level interventions to divert youth from secure detention facilities;

2.         to demonstrate that community-based interventions are an effective alternative to secure custody and that the needs of both the youths and the community can be met at a cost savings to the public;

3.         to reduce disproportionate minority incarceration.


Part of the impetus for the establishment of this program is the fact that the San Francisco juvenile detention system has been the focus of criticism for the past 40 years.  Since 1951 a long series of reports have documented San Francisco's overuse of detention and its failure to develop suitable alternative options.  The two most recent reports were completed by the Jefferson Associates in 1988 and the National Council on Crime and Delinquency (NCCD) in 1990.  The researchers at NCCD noted that because of the absence of alternatives, San Francisco has a detention rate that ranks third in the state (Steinhart and Steele, 1988).  This same study found that more than one-half of all the referrals (63%) were African-American youth, far in excess of their proportion in the general population.  A 1994 study by the Center on Juvenile and Criminal Justice found that the overall incarceration rate for African-American males in San Francisco was 8,331 per 100,000, compared to a rate of 3,822 for the country as a whole and 851 in South Africa (Hewitt, Shorter, and Godfrey, 1994; Hewitt, Kubota, and Schiraldi, 1992).  One of the major reason for the high use of incarceration is the absence of intermediate options.

The racial disproportionality issue in San Francisco is not unique, for this is a serious problem nationwide.  As Krisberg and Austin have noted (1993:118-127) minorities now outnumber white youths among those incarcerated on any given day.  In 1985 for example, minorities constituted 52% of the total juvenile incarcerated population, while in 1989 this percentage had risen to 60.  This fact cannot be attributed merely to the fact that minorities stand a greater chance of being arrested, for careful research has noted that even following arrest, minorities fare worse than whites are every subsequent stage of the juvenile justice system processing; regardless of the charge, they are more likely to be detained and sentenced to a correctional facility. 

In addition to the high levels of minority incarceration, youths in the juvenile justice system often reflect a plethora of high risk elements.  Risk factors include inadequate family support, school failure, negative peer associations, and insufficient utilization of community-based services.  Because most adjudicated youths are released from secure detention without community follow-up or supervision, high risk factors remain unaddressed (see also, Dryfoos, 1990).


It has been noted that the overuse of detention is partly attributed to the failure of the probation department staff to consider alternative options at the time a youth is taken into custody.  Indeed, the juvenile probation department in San Francisco has consistently recommended detention in the majority of the cases; one study found a 77% detention rate (Steinhart and Steele, 1988; Macallair, 1994).  However, as demonstrated in other jurisdictions, when community agencies are present to advocate for alternatives to detention, secure custody rates decline.  A recent study by the Massachusetts Department of Youth Services found that advocacy on behalf of youth at detention hearings by community agencies leads to significant reductions in unnecessary detentions.  When advocacy is combined with intensive case management, youths receive a wide range of quality services (Krisberg et al., 1988; Austin et al., 1991). 

The ability of case advocacy and case management to promote detention alternatives was recently demonstrated by the National Center on Institutions and Alternatives (NCIA).  Under contract with New York City's Spofford Detention Center, NCIA significantly augmented the efforts of that city's Department of Juvenile Justice to reduce the number of youth in detention and expand the range of alternative options (Jefferson and Associates, 1987; this is also documented in Krisberg and Austin, 1993:178-181).  A similar case management system has been in use in Florida through the Associated Marine Institutes (Krisberg and Austin, 1993: 178-181).  The Key Program, Inc., also uses the case management approach where in this instance the youth are closely supervised, meaning that they are monitored on a 24-hour basis and must conform to some very strict rules concerning work, school, counseling, victim restitution, etc. (Ibid.).

Additional evidence in support of the use of case advocacy comes from a study by the Rand Corporation (Greenwood and Turner, 1991).  This study compared two groups of randomly selected youths, a control group that was recommended by their probation officers for incarceration and an experimental group that received disposition reports by case advocates.  Of those who received case advocacy disposition reports, 72% were diverted from institutional care, compared to 49% of the control group.  The Rand study also found tremendous resistance from juvenile justice officials, especially probation officers, to alternative dispositions, especially those coming from case advocates.  It appeared that the probation staff resented the intrusion into what had heretofore been considered their own “turf” (Greenwood and Turner, 1991: 92).


The funding under the 1992-93 Children's Services Plan allowed for start-up costs and an initial collaboration of five San Francisco agencies: OMI Pilgrim Community Center, the Potrero Hill Neighborhood House, the Vietnamese Youth Development Center, Horizons Unlimited and the Center on Juvenile and Criminal Justice (CJCJ).  Additional cooperation has also come from the Juvenile Probation Department, the Public Defender's Office. the San Francisco Education Services Corporation, the Log Cabin Ranch aftercare program and the Omega Boys Club.  Funding has been provided through several local sources, including the City of San Francisco which, as a result of a vote on a proposition by the voters, provides 1% of all city revenues to fund services for children.  This collaboration led to the establishment of the Detention Diversion Advocacy Project (DDAP) to provide case advocacy and case management for youths who would otherwise be detained.  The Center for Juvenile and Criminal Justice (CJCJ), a non-profit organization that promotes and sponsors alternatives to incarceration for both youth and adults, oversees the DDAP project.

As of the end of 1995, a total of 612 clients had been referred to DDAP, of which 77% were accepted (470).  Of the remaining cases, a total of 358 were accepted by the court and placed within the DDAP program for supervision.  

The DDAP program involves two primary components:



1.  Detention Advocacy:  This component involves identifying youth likely to be detained pending their adjudication.  Once a potential client is identified, DDAP case managers present a release plan to the judge.  The plan includes a list of appropriate community services that will be accessed on the youth's behalf.  Additionally, the plan includes specified objectives as a means to evaluate the youth's progress while in the program.  Emphasis is placed on maintaining the youth at home and if the home is not a viable option, the project staff will identify and secure a suitable alternative.  If the plan is deemed acceptable by a judge, the youth is released to DDAP's supervision.

2.  Case Management:  The case management model provides frequent and consistent support and supervision to youth and their families.  The purpose of case management is to link youths to community-based services and closely monitor their progress.  Case management services are “field-oriented,” requiring the case manager to have daily contact with the youth, his or her family, and significant others.  Contact includes a minimum of three in-person meetings a week.  Additional services are provided to the youth's family members, particularly parents and guardians, in areas such as securing employment, day care, drug treatment services, and income support.


Clients are primarily identified through referrals from the public defender's office, the probation department, community agencies, and parents.  Admission to DDAP is restricted to youths currently held, or likely to be held, in secure detention.  The youths selected are those deemed to be “high risk” in terms of their chance of engaging in subsequent criminal activity.  The selection is based upon a risk assessment instrument developed by the National Council on Crime and Delinquency.  The target population are those whose risk assessment scores indicate that they would ordinarily be detained.  This is what Miller has terms the “deep-end” approach (Miller, 1991).  This is very important, for by focusing on detained youth the project ensures that it remains a true diversion alternative, rather than net widening.  Youths are screened by DDAP staff to determine if they are likely to be detained and whether they present an acceptable risk to the community.[1]  

Client screening involves gathering background information from probation reports, psychological evaluations, police reports, school reports, and other pertinent documents.  Interviews are conducted with youths, family members, and adult professionals to determine the types of services required.  Once a potential client is evaluated, DDAP staff presents a comprehensive community service plan at the detention hearing and requests that the judge release the youth to DDAP custody.

Because the project deals only with youths who are awaiting adjudication or final disposition, their appropriateness for the project is based on whether they can reside in the community under supervision without unreasonable risk and their likelihood of attending their court hearings.  This is similar in principle of what often occurs in the adult system when someone is released on bail pending their court hearings (e.g., arraignments, trial).


The primary goal of the project is to design and implement individualized community service plans that address a wide range of personal and social needs.  Services that address specific linguistic or medical needs are located by case managers.  Along with the youth's participation, the quality and level of services is monitored by DDAP staff.  It should be noted that the purpose of multiple collaboratives is to insure that the project is able to represent and address the needs of the various communities within San Francisco in the most culturally appropriate manner.  Since youth services in San Francisco have been historically fragmented by ethnicity, race, and community, a more unified approach is being tried with DDAP in that it has become a neutral site within the city and staffed by representatives from CJCJ and several other community-based service agencies (e.g., Horizon's Unlimited, Potrero Hill Neighborhood House, Vietnamese Youth Development Center).

More specific goals include: (1) ensuring that a high proportion of the program clients are not rearrested while participating in the program; (2) achieving a high court reappearance rate; (3) reducing the population of the Youth Guidance Center; (4) reducing the proportion of minority youths in detention.  Currently, the Youth Guidance Center is the only place of detention in the city.  It has a capacity of 137, but the daily population typically ranges from 140-150.  The average length of stay is around 11-12 days.


Part III: Data and Sampling Procedures


The data from the two samples were made available from print-outs obtained from the San Francisco Department of Juvenile Probation.  Systematic random sampling techniques were used to select the control and DDAP group for comparison.  A printout showing selected socio-demographic and legal variables was obtained for all youths who spent three or more days in detention during the calendar year 1994.  Originally, only DDAP referrals during 1994 were to be used (n=189), but it was decided to draw additional names from referrals during the second half of 1993 in order to have a larger sample.  A total of 271 from each group were eventually selected (total n = 542).  Each of the printouts contained such information as referral data, age, race, gender, prior referrals (including the charges), prior risk scores, prior placements, subsequent referrals (including the charges), subsequent placements, and subsequent petitions.


Additional information was obtained for the DDAP sample through intake forms that are filled out for each client by a case worker (unfortunately time did not permit collecting this information for the control group).  The type of information available on these intake forms include neighborhood, highest grade completed, number of times expelled or suspended, whether or not the youth is attending, parents living with youth or not, drug usage, and several poverty indicators (e.g., living in public housing, receiving welfare assistance, etc.).


Key Concepts Defined


Among the concepts used in this study, the following definitions are used.  First, the dependent variable in the study was recidivism and this was operationally defined as a referral to the juvenile court on a new offense (technical violations, if not accompanied by a specific delinquent charge, were not counted as recidivists) subsequent to the original referral to either DDAP or to the control group.  Data were not available (because of either time or monetary restraints) on other possible measures of recidivism (e.g., police contacts, arrests as an adult, etc.).  Recidivism was subsequently subdivided into serious recidivists and minor recidivists.  The former included those referred to court on felonies (e.g., robbery, murder, burglary, grand theft) or other serious charges (e.g., drugs), while the latter included misdemeanors (e.g., petty larceny, simple assault).  Other measures of recidivism were also used, namely subsequent petitions and subsequent out-of-home placements.

Among the most important independent variables were the nature of previous offenses for those with prior referrals.  This was operationalized as (1) serious violent (robbery, murder, assault with a deadly weapon, rape, etc.), (2) serious other (burglary, grand theft, drugs, etc.), (3) minor offenses (e.g., petty larceny, simple assault, disturbing the peace, etc.), and (4) technical offenses (violating a court order, etc.).


Another key independent variable was risk scores which were listed on various print-outs of data compiled by the Juvenile Probation Department.  These were divided into four major categories: (1) under 10, (2) 10-14, (3) 15-19, (4) 20 or more.  Since it was discovered that in filing out the “risk assessment” forms, some intake workers stopped adding up the points once a youth reached a score of 10 (so that an exact 10 could theoretically be a 15 or a 20 or more), it was decided to operationalize this variable as either low risk (under 10) or high risk (10 or higher), because according to the risk assessment form used (which was originally designed by NCCD and is used throughout the country) a score of 10 or more indicates that the youth is either a “danger to himself or others” or is likely to “abscond” or “without adequate adult supervision” and is ordered to remain in detention.


Part IV.  Findings


A Socio-Demographic Profile of the DDAP Group


Since more detailed information was available, via intake forms, about the youths handled by DDAP during 1994 (the first full year of operation, n=189) a summary of these 189 cases handled by this program will be reviewed first.  From these forms, several socio-demographic variables are summarized for each case.   As already noted, intake forms from 1993 were too incomplete to be included here. 


As noted in Table 1, the majority of the subjects were members of minority groups, with African-Americans accounting for just over half (56%) of the sample.  Asians and Hispanics constituted almost 30 percent of the group (29.6%), while white youths were almost 13 percent of the group (12.7%).   Also, most were male (82.5%) and they tended to come from one of four neighborhoods (some of these areas border each other so were combined to simplify the presentation): Potrero/Bayview Hunter, Excelsior/Visitacion, Mission, and the Haight district. The twin neighborhoods of Potrero Hills and Bayview Hunter were most common, constituting one-fourth of all DDAP clients. 

The most common family arrangement was living with mother only, not surprising given the more general family backgrounds of delinquent youth these days.  The data also indicate that school was a problem for these youths, as only slight majority (53.6%) were attending school at the time they entered DDAP and about one-fourth had been either expelled or suspended at least once.

Not surprisingly, drug use was common, as over half had used drugs (53.5%) and almost half (46%) had used them at least once during the 90 days prior to DDAP referral, and about one-third used drugs on a daily basis. Marijuana was the most popular drug used.  Most of the DDAP clients were not employed and just over 40% were living in poverty conditions, indicated by notations on the intake forms that they were living in government housing or receiving some sort of public assistance.


Group Comparisons (DDAP Group vs. Control Group)



As noted in Table 2, there were several significant differences between the DDAP sample and the control group sample.  However, many of the differences are somewhat surprising.  In the first place, a greater percentage of the control group were younger (over one-fourth were 14 and under).  Second, the control group had a higher percentage of females.  This in itself would lead one to predict a lower recidivism rate, since females traditionally have a lower rate of criminal involvement than males.  Third, the DDAP group was significantly more likely to have a risk score of 10 or higher.  This should, ideally, cause one to predict a higher recidivism rate among the DDAP group, which was not the case, as already mentioned. Fourth, the control group had a higher percentage of those with three of more referrals.  This seems to conflict with the lower overall risk scores of the control group, since presumably multiple referrals might indicate a strong probability of recidivism and hence "risk."  Fifth, as far as the nature of the prior offenses, no significant differences were found between the two groups: among those with previous referrals, each group was about equally as likely to have prior serious offenses.  Sixth, surprisingly the DDAP group was more likely to have had prior placements. 

Most of the above differences might have led one to predict a slightly higher recidivism rate among the DDAP group (e.g., more males, higher risk scores, prior placements), yet the most significant finding of this entire study was the fact that the DDAP group had a much lower recidivism rate.  In fact, the overall recidivism rate for the control group was almost double that for the DDAP group. Moreover, there was quite a large difference for the rate of serious recidivism (defined as those with subsequent referrals for major felonies).  Further, the control group was more than three times more likely to have had two or more subsequent referrals and, more dramatically, the control group was almost three times more likely to be referred for a violent crime. Further, the control group was about four times more likely to have two or more subsequent petitions and slightly more likely to have subsequent placements.  All of these relationships were statistically significant at the .001 level, except for gender differences (p. <.05) and subsequent placements (not significant).[2] 


Comparisons were made controlling for age, for the main reason that older youth had a lower recidivism rate, which could reflect the fact that so many turned 18 and had no  subsequent referrals to juvenile court for that reason alone.  Examining Tables 3 and 4, however, show that the comparisons between the two groups remained essentially the same.  Recidivism rates were still significantly higher for the control group.

Controlling for additional variables (see Table 5), we still find that the DDAP clients fared significantly better.  So, for instance, among those with high risk scores (10 or higher), the overall recidivism rate for the DDAP clients was 34%, compared to 59% for the control group.  Almost identical differences were found for those with low risk scores (under 10).  The highest recidivism rates for the DDAP group were found among those with three or more referrals (exactly 50%); but they were still significantly lower than the control group (70.5%).          

What is perhaps most interesting from the data presented in Table 5 is that for those for whom one would normally predict a low rate of recidivism (e.g., low risk scores, no previous referrals, no prior placements, with minor previous offenses, and girls) in each case the recidivism rate was higher among the control group.  All but the relationship for girls were statistically significant (although for the girls, as indicated, the difference between the DDAP group and the control group was substantial).  In fact, in almost every case the recidivism rate for the control group was double that of the DDAP group.  This finding lends support to the frequent charge that the juvenile justice system takes youths with a low potential for delinquency and only makes matters worse (Becker, 1963; Schwartz, 1989; Schur, 1971). 


Table 6 breaks down these rates more fully, comparing recidivism rates by risk scores for the two groups.  As shown in this table, the rate for serious recidivists was consistently higher among the control group for most of the variables noted here. Thus, 44% of the “high risk” control group youths were serious recidivists, compared to only 23% of the control group.  More critical was that among “low risk” groups, only 13% of the DDAP youths were serious recidivists compared to almost half (49%) of the control group.  In other words, not only was the recidivism rate consistently higher among the control group, the youths left within the juvenile justice system who did return were far more likely to return on a more serious charge, even when they were considered to be “low risk (see also Tables 2-4).


Other Measures of Recidivism


Obviously the term “recidivism” can be defined in many different ways, as can the concept of success.  One possible measure is “subsequent petitions.”  Actually this could be considered a much better measure of success or failure because so many referrals are never petitioned to court for more formal hearings.  One could conclude that the petitioned cases are the cases with the most evidence or are the most serious and that those not petitioned to court  have little or no evidence or are not serious enough to warrant further court action.

So, as shown in Table 7, the differences between the DDAP group and the Control Group remain significant.  Whereas only about 24% of the DDAP group had at least one subsequent petition, almost half (47.8%) of the control group did.  When considering age, those 14 and under had a significantly higher recidivism rate, as was the case for males, those with three or more prior referrals and those with the most serious prior referrals.  Race and prior placements did not correlate with subsequent petitions.  Also, risk scores did not predict subsequent petitions. 


Still another possible measure of recidivism was being referred to court on a charge of violence. Table 8 shows these data.  Once again, group differences surfaced, but this time the differences were more dramatic as the control group was almost three times more likely than the DDAP group to return to court on a charge of violence.  Also, note that there was a rather large difference when considering age, as more than one-third of those 14 and under had subsequent referrals for violence, compared to only 12% of the older youths.  Additionally, note that the differences between white and black youth were negligible; both were significantly higher than other racial groups.  Not surprisingly, the greater the number of prior referrals and the more serious the prior referrals, the greater was the likelihood of being returned on a charge of violence.  One might speculate that keeping a youth within the juvenile justice system tends to create a strong likelihood of subsequent violence.  More will be said about this in the next section.

Finally, as shown in Table 9, subsequent out-of-home placement as a measure of recidivism is considered.  While the differences between DDAP and Control group youths are not statistically significant, the control group nevertheless was slightly more likely to have subsequent placements.  Subsequent placements were also more likely for younger youths and for youths with the most priors, the most serious priors and those with prior placements.




Summary of the Findings


It seems obvious that the youths referred to DDAP have been more successful than those not referred, at least as far as the key measure of success used here is concerned, namely, recidivism.  As has been noted so often, however, the term recidivism can have different meanings and can be measured in several different ways.  Given the available data, at least three popular measures can be used: a referral to court on a new offense (and this can be further broken down to a focus on violent offenses, as we have done here), a referral that results in an actual petition to go before the judge for possible adjudication, and a referral that results in some sort of out-of-home placement (e.g., group home, institution, etc.)  By each of these three measures, the DDAP group is decidedly more successful.

To counter the possibility that we are comparing “apples and oranges” several kinds of data were examined and the conclusion is inescapable that the DDAP group was, overall, not significantly different from the control group. In fact on several measures the DDAP group had characteristics that would lead to the prediction that the recidivism rates would be higher.  In other words, the "apples and oranges" argument does not hold up and in fact one could even reverse the use of the metaphor: the DDAP group consisted of more serious offenders than the control group.


The data suggest that the risk scores themselves are relatively poor predictors of outcome, unless, as already noted, many of the assessment forms were not filled out properly.  Since this is not known for sure, we have to base our conclusion on existing risk scores.  However, the problem with the risk scores may be irrelevant anyway, because there is an even more important fact to consider.  It cannot be denied that DDAP accepted many cases that conventional wisdom suggested were “dangerous” youth who posed a “threat to public safety” (whether this assessment is based on risk scores or some other criteria) and were successful.  DDAP took youthful offenders who otherwise would sit in detention for several days or even weeks, intensely supervised them over an extended period of time, and placed them into community-based programs (that incidentally are not formally connected with the juvenile justice system) which enabled these youths to lead productive lives without trouble with the police.  We don't know, of course, how many subsequently became adult offenders later in life; but by the same token, we don't know how many of the control group became adult offenders.

There may be several reasons for the apparent success of this program.  From the data collected here and information from previous research, three reasons seem of paramount importance.

First, the caseloads of the DDAP case workers are extremely low in comparison to normal probation officers.  The DDAP workers average about ten cases each.  Regular probation officers in major urban areas have caseloads ranging from 50 to 150.  Smaller caseloads typically results in more intensive supervision, and more intensive supervision means that the case worker is constantly “on top of things” with regard to their clients.  Indeed, with small case loads they can spend more quality time with their clients in the field (e.g., in their homes, on the street corners, at school), rather than endless hours in an office doing paperwork, on the phone and other bureaucratic chores.


Second, DDAP is a program that is “out of the mainstream” of the juvenile justice system, that is, it is a true “alternative” rather than one of many bureaucratic extensions of the system.  This means that normal bureaucratic restrictions do not generally apply.  For instance, the qualifications for being a case worker with DDAP are not as strict as you might find within the juvenile justice system (e.g., age restrictions, educational requirements, arrest records, “street” experience, etc.).  From casual observations of some of these case workers, this researcher was impressed with their dedication and passion to helping youth.  Moreover, the backgrounds of these workers were similar to the backgrounds of some of their clients (e.g., similar race, neighborhood of origins, language, etc.). 

Third, the physical location of DDAP seemed to this observer “user friendly” and lacked the usual “macho” appearance of the formal system. There are no bars, no concrete buildings, no devices for screening for weapons as you enter the building, no cells for “lockdown,” etc. Further, the DDAP workers are not “officers of the court” with powers of arrest, and the usual accoutrements of such occupations (e.g., badges, guns).  


There could also be a possible fourth explanation - but one we can only speculate on at this time because we lack the data to draw such a conclusion.  It could be that given the low case loads, DDAP case workers are more likely than regular probation officers to be “on top of the case,” that is, be in constant contact with the youth and thus be able to “nip in the bud” potential problems.  Also, some police officers, when facing a possible arrest situation, learns that the youth is a DDAP case (presuming the officer knows about DDAP), may be in a position to contact the case worker who might be able to persuade the officer that the situation could be handled without a formal arrest.  We have no way of knowing whether or not this occurs with any degree of regularity.  Even if it did, such a procedure may be a positive sign, since youths from more privileged backgrounds are often treated this way by the police, if it is believed that someone in authority can handle the youth informally.  Many a youth has been saved the stigma of formal juvenile processing by such intervention by significant adults in their lives.




Given the apparent success of DDAP additional funding and an overall expansion of the program seems warranted; specifically it is highly recommended that more youths, especially those who have high risk scores or who otherwise would be predicted to be highly likely to engage in repetitive serious delinquency.  It seems obvious that many so-called “dangerous” youth can be handled within their own community.  The findings here provide support the conclusions of other researchers on this matter of the question of “dangerous” youth (see, e.g., Miller, 1990, 1996). 

The record keeping procedures on the part of DDAP workers needs to be improved; specifically, more detailed information should be collected, including information about what kinds of programming each client participated in.  Too many intake forms, and other information contained in case files, were incomplete or missing completely (this was especially true for the 1993 cases).


This research should be considered only a beginning.  Additional funding for more research on this program should be forthcoming.  Specifically, subsequent research should be conducted that would include in-depth interviews of samples of program participants and family members, DDAP case workers and heads of agencies whose programs DDAP clients were referred to.  Larger samples from both DDAP and control groups should be drawn and studied.  Also, sampled cases should be followed up to their adult years to find out how many became adult offenders, especially those who ended up incarcerated in adult institutions.  What is especially important for future research to examine is to look carefully at what accounts for the apparent success of DDAP. What sorts of programs were the non-recidivists referred to?  What did their case workers do? In short, what did DDAP do on behalf of these youths that are not normally done for them?

A serious re-assessment of the existing risk assessment instrument seems warranted, with special attention given to accuracy in filling out the form upon referral to juvenile court; subsequent research should seriously address this problem, with emphasis on the predictability of the risk assessment instrument itself.  We must remember that placing a youth in detention, or even referring to the court system itself, is a serious step that can have far-reaching, often negative, consequences in the life of a youth.  A “risk assessment” instrument using a numerical scale, even if it is based upon “scientific research,” should not be passively accepted as the “last word” or written in stone.  The assignment of a risk score (or any other “objective” numerical score) should not go unquestioned.

Finally, it is suggested that this program remain a true alternative to the formal juvenile justice system.  To put it bluntly, we should never allow this program to become bureaucratized nor to be co-opted or otherwise “invaded” or “taken over” by the existing juvenile justice system or any other existing bureaucracy.  The best crime control is often done in the community, without the interference of any formal, bureaucratic system.  Many a good idea has been corrupted by formalization and powerful interests (Elias, 1993).  The lives of our youth, our families and our communities are at stake.  The formal systems of social control have not proven themselves to be very successful.




The findings reported here have very broad implications in terms of national and local policies.  During the past couple of decades we have witnessed a resurgence of a conservative agenda toward crime and delinquency.  The various “get tough” policies (e.g., mandatory sentencing, increasing waivers of juveniles to the adult system, the “war on drugs,” the increase in executions) have had a negligible effect on the crime problem.  In fact, what such policies have done is to fill prisons, jails and juvenile institutions beyond their capacity.  Such policies have, however, expanded an already huge “crime control industry” that has a vested interest in perpetuating the problem because it is more profitable to do so (Shelden and Brown, 2000; Reiman, 1995; Donziger, 1996; Austin and Irwin, 2001).

More importantly, however, is the fact that we as a society continue to insist that the problems of crime and delinquency (and related problems, such as drug and alcohol abuse) be handled primarily as a legal, rather than a social or even a medical problem (Prothrow-Stith, 1991).  Rarely do these approaches address any of the major root causes of crime, such as poverty, lack of educational opportunities, unemployment, racial inequality, etc.  In fact, it can be argued that such policies have exacerbated these problems (Miller, 1996).  It is significant that when looking at the DDAP group alone, the one variable that is most strongly associated with recidivism is poverty.  Hence, expanding programs like DDAP without simultaneously addressing issues like poverty and inequality will not be enough.


This last point is further underscored by the results of a study that focused on one of the areas with the highest concentration of DDAP clients, the Bayview area of San Francisco.  The study (Bowser, Fullilove and Fullilove, 1990), which focused on crack cocaine, found that the heaviest concentration of sales was almost exclusively in the poorest housing projects of this community where residents have been on the margins of the economy since the 1960s.  The study also found that this mostly African-American community “has been cut off from the city's economic life” as the prospects for good jobs has steadily shrunk.  The community itself has deteriorated to such an extent that there is little control by the residents.  The rising rates of drug sales (especially crack) have occurred simultaneously with increases in unprotected sex and the concomitant rise in cases of sexually transmitted diseases, including AIDS.  The authors concluded that these problems have been influenced by “long-term community economic conditions and opportunities.”  They see their parents and neighbors, and not surprisingly see themselves, cut off from the surrounding community with little hope for the future (Bowser, Fullilove and Fullilove, 1990: 57-63).  


The recently passed Juvenile Crime Bill is illustrative of this trend.  This bill represents a classic example of our continuous “war on crime” that is based less on facts than on media hype, exaggerations, stereotypes and anecdotal evidence.  The bill seeks to treat as adults teenagers as young as 13, first at the federal level and ultimately at the state level, since the states will be tempted by the lure of money.  There will be some effects, however, not often recognized by the designers of such bills (who, after all, never seem to look beyond the next election).  For one, without any doubt the vast majority of those juveniles waived to adult court will be African-American males.  Too often the stereotypes of the “dangerous criminal” have a black or brown face.  Second, getting tough on crime has a sort of “trickle down” effect.  Since we are already super-tough on the most serious crimes (especially violent crimes) this type of legislation has had the effect of increasing the penalties on lesser crimes or even prompting attempts by criminal justice officials to crack down on relatively minor crimes.  Thus, for instance, we have seen this effect in California's “Three Strikes and You're Out” legislation where a good number prosecuted under the new law are relatively minor offenders - and once again mainly African-Americans (and this has been the result all over the country).  In Las Vegas recently we have witnessed the macho posturing of a local district attorney who wanted desperately to prosecute a 9-year-old “felon” who wrote his name and the name of some of his friends on a new sidewalk!

Part of the problem is that such policies rarely address directly the actual needs of children and their families and to do so prior to their contact with the juvenile justice system.  And after they do come into contact with the juvenile justice system, it is apparent that the problem too often gets worse, as the data from this study clearly shows.  There is abundant research that shows very clearly the "risk factors" associated with delinquency and the kinds of programs that are effective.  And the most effective programs have little or no connection with the more established juvenile justice system (Dryfoos, 1990; Chesney-Lind and Shelden, 2004; Shelden, Tracy and Brown, 2004; Hawkins, Catalano, and Miller, 1992).


One of the main reasons for the success of such alternative programs is the fact that they tend to stress the strengths of at-risk youths, rather than the weaknesses.  Indeed, the more traditional, conservative approaches tend to emphasize the negative aspects of at-risk youth and focus on all of their faults and negative behaviors.  It is typical of the “get tough” approach to do so since, after all, it stems from a negative view of human nature - for instance, that in order to make people behave in a certain way you need something negative, like a threat, and you need to label someone as “bad.”  What the more effective programs do so well is to focus on building strengths so as to add to the resiliency of these youths (Hawkins, Catalano, and Miller, 1992; Hawkins and Weis, 1985).  Perhaps it is too easy to focus on the negative.  Maybe the reason we do stems in part from the media's preoccupation with the negative side of life (witness the never-ending violent programming during the local news hour on television).  We as a society seem to have a fixation on the “quick fix.”  Indeed, we suffer from what may be called the Excedrin Syndrome as we continue to search for quick and easy solutions to very complex problems.

It is the opinion of this author that in the final analysis, the solutions to these and related problems will not, sadly, come from entrenched bureaucracies too often plagued by “politics” and with the classic problem of “goal displacement,” that is, the tendency to lose sight of the original goals (in this case reducing delinquency and crime) and instead focusing mostly on the survival of the bureaucracy itself (Merton, 1968).  The solutions will come from local communities and groups of activists with few formal ties to the established bureaucracies.  It will take a lot of hard work on the part of dedicated people who have genuine compassion for our youth.  And we can expect a great deal of resistance from the more established bureaucracies, in particular the juvenile justice system. 

One final comment is in order.  Those who read this report must not conclude that the program evaluated here is the solution to all of our problems.  It is most assuredly not a “quick fix.”  More research on this type of program, under varying conditions and in different geographical locations, needs to be done.  Ideally we should conduct a longitudinal study that utilizes a true experimental design where youths would be randomly assigned to DDAP and regular juvenile justice programming.  More extensive and long-term follow-ups like these can shed a great deal more light than a brief and limited study such as this one.  We should consider the present study to be merely a beginning, rather than the last word on this problem.






Table 1.  Socio-Demographic Characteristics of the 1994 DDAP Cases (n=189)


%                    N




White                                                               12.7                 24

African-American                                             56.1                 106

Hispanic                                                           13.2                 25

Asian                                                                16.4                 31

Other                                                               1.6                    3




Male                                                                82.5                 156

Female                                                             17.5                 33


Neighborhood (n=188)


Potrero/Bayview Hunter                                   24.9                 47

Excelsior/Visitacion                                           14.8                 28

Mission                                                            12.7                 24

Haight                                                              12.2                 23

OMI                                                                6.9                  13

Tenderloin                                                        5.8                  11

Other                                                               22.2                 42


Family (n=188)


Intact (both natural parents)                              27.0                 51

Mother only                                                      46.0                 87

Father only                                                       8.5                  16

Relative                                                            13.2                 25

Other                                                               4.7                    9


Living Arrangements (n=188)


With parent(s)                                                  82.5                 156

Relative                                                            14.8                 28

Other                                                               2.1                    4


Highest Grade (n=178)


8th or less                                                         14.0                 25

9th                                                                   30.3                 54

10th                                                                 32.6                 58

11th                                                                 13.5                 24

12th                                                                 2.2                    4

GED                                                                7.3                  13


Attending School (n=181)


Yes                                                                  53.6                 97


Ever Expelled/Suspended (n=187)   


Yes                                                                  24.1                 45


Drug of Choice (n=187)


None                                                                46.5                 87

Marijuana                                                         44.9                 84

Other                                                               8.6                  16


Used last 90 days (n=187)


Yes                                                                  46.0                 86                  


Frequency of use (n=79)


Daily                                                                32.9                 26

Once a week                                                    19.0                 15

Twice a week                                                   43.0                 34

Once a month or less often                                5.1                    4




Yes                                                                  10.7                 20


Poverty Indicator


Yes                                                                  41.5                78





Table 2.  General Comparisons Between DDAP Sample and Control Group (n=542)


DDAP Control                        Significance


% 14 and Under                                               15.1                 27.3                             p. <.001


% Minorities                                                     88.9                 85.6                                 ns


% Female                                                         15.9                 22.9                             p. <.05


Risk Score 10 or more                                     84.2                 59.4                             p. <.001


% with 3 or more priors                                    19.9                 38.7                             p. <.001


Nature of Prior offenses:

% with prior serious violence                             23.2                 30.3

% with prior serious other                                 28.4                 23.6                                 ns


% with prior placements                                    27.3                 15.5                             p. <.001



% serious recidivist                                           23.6                 45.8

% total recidivist                                               34.3                 60.1                             p. <.001


% with 2 or more subsequent

  referrals                                                          14.4                 50.2                             p. <.001


% with subsequent violence                               9.2                  24.7                             p. <.001


% with 2 or more sub. petitions                         5.2                  21.5                             p. <.001


% with subsequent placements                          18.1                 24.0                             p. <.001






Table 3.  Group Comparisons for Referrals Age 14 and under (n=115).


DDAP Control                        Significance


% Minorities                                                     87.8                 90.5                                 ns


% Female                                                         26.8                 31.1                                 ns


% Risk Score 10 or more                                 83.3                 47.3                             p. <.01


% 3 or more prior referrals                               9.8                  35.1                             p. <.001


Nature of Prior offenses:

% with prior serious violence                             22.0                 21.6

% with prior serious other                                 24.4                 24.3                                 ns


% with prior placements                                    17.1                 10.8                                 ns




% serious recidivist                                           31.7                 63.5

% total recidivist                                               43.9                 74.3                             p. <.001


% with 2 or more subsequent

  referrals                                                          17.1                 63.5                             p. <.001


% with subsequent violence                               24.4                 40.5                             ns (p. <.09)


% with 2 or more subsequent petitions              2.4                  32.4                             p. <.001


% with subsequent placements                          26.8                 31.1                                 ns





Table 4.  Group Comparisons for Referrals Age 15 and older (n=427).


DDAP Control                        Significance


% Minorities                                                     89.1                 83.8                                 ns


% Female                                                         13.9                 19.8                                 ns


% Risk Score 10 or more                                 87.0                 64.0                             p. <.01


% 3 or more prior referrals                               21.7                 40.1                             p. <.001


Nature of Prior offenses:

% with prior serious violence                             23.5                 33.5

% with prior serious other                                 29.1                 23.4                                 ns


% with prior placements                                    29.1                 17.3                             p. <.01




% serious recidivist                                           22.2                 39.1

% total recidivist                                               32.6                 54.8                             p. <.001


% with 2 or more subsequent

  referrals                                                          13.9                 45.2                             p. <.001


% with subsequent violence                               6.5                  18.8                             p. <.001


% with 2 or more subsequent petitions              5.7                  17.3                             p. <.001


% with subsequent placements                          16.5                 21.3                                 ns





Table 5.  Recidivism Rates, DDAP and Control, by Selected Variables.


Recidivism Rates


DDAP             Control            Significance


High Risk Scores (10 +)                                                                                   32.8%              58.4%              p. <.001


Low Risk Scores (under 10)                                                                             31.4                 62.7                 p. <.02


With Three or More Prior Referrals                                                                  50.0                 70.5                 p. <.02


With No Prior Referrals                                                                                    25.0                 43.0                 p. <.01


With Previous Placements                                                                                 33.8                 66.7                 p. <.001


With no Prior Placements                                                                                  34.5                 58.8                 p. <.001


With Serious Prior Offenses                                                                              42.1                 70.5                 p. <.001


With Serious Violent Prior Offenses                                                                  42.9                 68.3                 p. <.01


With Minor Prior Offenses                                                                                28.1                 69.0                 p. <.01


Blacks only                                                                                                       38.5                 66.4                 p. <.001


Males only                                                                                                        35.1                 63.6                 p. <.001


Girls only                                                                                                          30.2                 48.4                    ns


15 and over                                                                                                      32.6                 54.8                 p. <.001


14 and under                                                                                                    43.9                 74.3                 p. <.001





Table 6.  Risk Scores and Recidivism Rates, DDAP vs. Control Group.


DDAP             Control            Significance

Low Risk


  Recidivist                                                                                                        31%                 63%                 p. <.01


High Risk


  Recidivist                                                                                                        33%                 58%                 p. <.001 



Low Risk


  Serious recidivist                                                                                             13%                 49%

  Minor recidivist                                                                                               16                    14                    p. <.001


High Risk


  Serious recidivist                                                                                             23%                 44%

  Minor recidivist                                                                                               11                    15                    p. <.001





Table 7.  Subsequent Petitions as a Measure of Recidivism.


Percent with One or More Subsequent Petitions     Significance


DDAP                                                                                      23.6

Control                                                                                     47.8                                                     p. <.001


14 & Under                                                                              45.2

Over 14                                                                                   33.1                                                     p. <.05


White                                                                                       40.6

Black                                                                                       35.8

Other                                                                                       33.5                                                       ns


Male                                                                                        37.8

Female                                                                                     26.7                                                     p. <.05


High Risk (10 or more)                                                             35.2

Low Risk (under 10)                                                                38.9                                                       ns


Prior Referrals

  None                                                                                      26.6

  One                                                                                        29.4

  Two                                                                                       42.2

  Three or more                                                                        49.1                                                     p. <.001


Nature of priors

  Serious                                                                                   43.2

  Minor                                                                                     31.1                                                     p. <.001


Prior Placements                                                                       35.7

No Prior Placements                                                                 35.7                                                       ns






Table 8.  Subsequent Violence as a Measure of Recidivism.


Percent with Subsequent Violence            Significance


DDAP                                                                                                  9.2

Control                                                                                                 24.7                                         p. <.001


14 & Under                                                                                          34.8

Over 14                                                                                               12.2                                         p. <.001


White                                                                                                   17.4

Black                                                                                                   20.6

Other                                                                                                   10.7                                         p. <.05


Male                                                                                                    18.1

Female                                                                                                 12.4                                           ns


High Risk                                                                                             14.1

Low Risk                                                                                             19.3                                           ns


Prior referrals

  None                                                                                                  13.0

  One                                                                                                    13.4

  Two                                                                                                   18.8

  Three or more                                                                                    23.9                                         p <.05


Nature of priors

  Serious                                                                                               20.6

  Minor                                                                                                 14.8                                         p <.05


Prior Placements                                                                                   15.5

No Prior Placements                                                                             17.4                                           ns





Table 9.  Subsequent Placements as a Measure of Recidivism.


Percent with Subsequent Placements                       Significance


DDAP                                                                                      18.1

Control                                                                                     24.0                                                       ns


14 & Under                                                                              29.6

Over 14                                                                                   18.7                                                     p. <.05


White                                                                                       24.6

Black                                                                                       22.6

Other                                                                                       16.9                                                       ns


Male                                                                                        21.5

Female                                                                                     19.0                                                       ns


High Risk (10 or more)                                                             22.8

Low Risk (under 10)                                                                15.9                                                       ns


Prior Referrals

  None                                                                                      13.0

  One                                                                                        13.4

  Two                                                                                       23.4

  Three or more                                                                        35.8                                                     p. <.001


Nature of priors

  Serious                                                                                   27.3

  Minor                                                                                     18.0                                                        ns


Prior Placements                                                                       29.3

No Prior Placements                                                                 18.8                                                     p. <.05








Austin, J., Elms, W., Krisberg, B, and P. Steele (1991).  Unlocking Juvenile Corrections: Evaluating the Massachusetts Department of Youth Services.  San Francisco: National Council on Crime and Delinquency.


Austin, J. and J. Irwin (2001).  It's About Time: America's Imprisonment Binge (3rd  ed.). Belmont, CA: Wadsworth.


Becker, H. S. (1963).  Outsiders.  New York: Free Press.


Bortner, M.A., M.L. Sunderland, and R. Winn (1985).  “Race and the Impact of Juvenile Deinstitutionalization.” Crime and Delinquency 31 (January): 35-46.


Bowser, B. P., M. T. Fullilove and R. E. Fullilove (1990).  “African-American Youth and AIDS High Risk Behavior: the Social Context and Barriers to Prevention.”  Youth and Society 22, 1: 54-67.


Bullington, B., J. Sprowls, D. Katkin, and M. Phillips (1978).  “A Critique of Diversionary Juvenile Justice.”  Crime and Delinquency 24 (January): 59-71.


Bynum, J. E. and W. E. Thompson (1996).  Juvenile Delinquency: A Sociological Approach (3rd. ed.).  Needham Heights, MA: Allyn and Bacon.


Chesney-Lind, M. and R. G. Shelden (2004). Girls, Delinquency and Juvenile Justice. (3rd  ed.). Belmont, CA: Wadsworth.


Davidson, W. S. II, R. Redner, R. Admur and C. Mitchell (1990).  Alternative Treatments for Troubled Youth: The Case of Diversion from the Justice System.  New York: Plenum.


Donziger, S. (1996).  The Real War on Crime.  New York: Harper.


Dryfoos, J. (1990).  Adolescents at Risk.  New York: Oxford University Press.


Elias, R. (1993).  Victims Still: The Political Manipulation of Crime Victims.  Newbury Park, CA: Sage.


Feldman, R. A., T. E. Caplinger and J. S. Wodarski (1983). The St. Louis Conundrum.  Englewood Cliffs, N.J.: Prentice-Hall.


Frazier, C. E. and S. R. Lee (1992).  “Reducing Juvenile Detention Rates or Expanding the Official Control Nets: An Evaluation of Legislative Reform Effort."” Crime and Delinquency 38: 204-218.



Frazier, C.E. and J. K. Cochran (1986).  “Official Intervention, Diversion from the Juvenile Justice System, and Dynamics of Human Services Work: Effects of a Reform goal Based on Labeling Theory.” Crime and Delinquency 32: 157-176.


Gibbons, D. C. and M. D. Krohn (1991).  Delinquent Behavior (fifth edition).  Englewood Cliffs, NJ: Prentice-Hall.


Gilbert, G. R. (1977). “Alternate Routes: A Diversion Project in the Juvenile Justice System.” Evaluation Quarterly 1:301-318.


Greenwood, P. W. and S. Turner (1991).  Implementing and Managing Innovative Correctional     Programs: Lessons from OJJDP's Private Sector Initiative.  Santa Monica, CA:  Rand Corporation.


Hawkins, J.D., R.F. Catalano, and J. Y. Miller (1992).  “Risk and Protective Factors forAlcohol and Other Drug Problems in Adolescence and Early Adulthood:

Implications for Substance Abuse Prevention,” Psychological Bulletin 112: 64-105.


Hawkins, J.D. and J. G. Weis (1985). “The Social Development Model: An Integrated Approach to Delinquency Prevention.” Journal of Primary Prevention 6:73-79.


Hewitt, C., A. Shorter, and M. Godfrey (1994).  Race and Incarceration in San Francisco: Two Years Later. San Francisco, CA:  Center on Juvenile and Criminal Justice.


----------, K. Kubota, and V. Schiraldi (1992).  Race and Incarceration in San Francisco: Localizing Apartheid. San Francisco, CA:  Center on Juvenile and Criminal Justice.


Jefferson and Associates, Community Research and Associates (1987).  Creating a New Agenda for the Care and Treatment of San Francisco's Youthful Offenders: A Model Program. Author.


Krisberg, B. and J. Austin (1993).  Reinventing Juvenile Justice. Newbury Park, CA: Sage


Krisberg, B., J. Austin, K. Joe and P. Steele (1988).  A Court that Works: The Impact of Juvenile Court Sanctions.  San Francisco: National Council on Crime and Delinquency.


Lemert, E. (1951).  Social Pathology.  New York: McGraw-Hill.


Lincoln, S. B. (1976).  “Juvenile Referral and Recidivism.”  In R. M. Carter and M. W. Klein (eds.), Back on the Street: Diversion of Juvenile Offenders.  Englewood Cliffs, N.J.: Prentice-Hall.


Lipsey, M. W. (1981) “Evaluation of a Juvenile Diversion Program: Using Multiple Lines of Evidence.” Evaluation Review 5:283-306.


Lundman, R.J.  (1993).  Prevention and Control of Delinquency (2nd ed.).  New York:  Oxford University Press.


Macallair, D. (1994).  “Disposition Case Advocacy in San Francisco's Juvenile Justice System: A New Approach to Deinstitutionalization.”  Crime and Delinquency 40: 84-95.


Merton, R. K.  1968.  Social Theory and Social Structure. New York: Free Press.


Miller, J.  (1990). The Last One Over the Wall.  Columbus: Ohio State University Press.


------------  (1996). Search and Destroy: African-American Males and the Criminal Justice System.  New York: Cambridge University Press.


Moxley, R. (1989).  Case Management.  Beverly Hills, CA: Sage Publications.


Pogebrin, M. R., E. D. Poole, and R. M. Regoli (1984). “Constructing and Implementing a Model Juvenile Diversion Program.”  Youth and Society 15: 305-324.


Polk, K. (1995).  “Juvenile Diversion: A Look at the Record.” In P. M. Sharp and B. W. Hancock (eds.), Juvenile Delinquency.  Englewood Cliffs, NJ: Prentice-Hall.


Pope, C. and W. Feyerherm (1993).  Minorities and the Juvenile Justice System: an Executive Summary.  Rockville, MD: Office of Juvenile Justice and Delinquency Prevention.


President's Commission on Law Enforcement and Administration of Justice (1967).  Task Force Report: Juvenile Delinquency and Youth Crime.  Washington, D.C.: U.S Government Printing Office.


Prothrow-Stith, D. (1991).  Deadly Consequences.  New York: Harper/Collins.


Reiman, J. (1995).  The Rich Get Richer and the Poor Get Prison:  Ideology, Crime, and Criminal Justice. 4th ed.  Boston:  Allyn and Bacon.


Rojek, D. G. (1982).  “Juvenile Diversion: A Study of Community Cooptation.”  In D. G. Rojek and G. F. Jensen (eds.), Readings in Juvenile Delinquency.  Lexington, MA: D. C. Heath.


Rojek, D. G. and M. L. Erickson (1981-82).  “Reforming the Juvenile Justice System: the Diversion of Status Offenders.”  Law and Society Review 16: 241-64.


Schur, E. (1971).  Labeling Deviant Behavior.  New York: Harper and Row.


Schwartz, I. (1989).  (In)justice for Juveniles.  Lexington, MA: Lexington Books.


Shelden, R. G. and W. B. Brown (2000).  “The Crime Control Industry and the Management of the Surplus Population.” Critical Criminology (Autumn).


--------------, S. K. Tracy and W. B. Brown (2004). Youth Gangs in American Society. (3rd ed.) Belmont, CA: Wadsworth.


Steinhart, D. and P. Steele (1988).  San Francisco Detention Survey: Results and Recommendations.  San Francisco: National Council on Crime and Delinquency.


Tannenbaum, F. (1938).  Crime and the Community.  New York: Columbia University Press.


Wordes, M., T.S. Bynum and C.J. Corley (1994). “Locking Up Youth: The Impact of Race on Detention Decisions.” Journal of Research in Crime and Delinquency

31: 149-165.






[1]  It should be noted that while the goals are to deal with non-violent youth, nevertheless this does not preclude the inclusion of some youth who may exhibit violent behavior.  According the original project proposal, several factors will be considered to determine whether it would be in the interest of public safety to release the youth to the project.  These factors include: (1) the level of participation and role in the offense(s) charged, (2) the circumstances surrounding the alleged offense(s), (3) whether there is a pattern or history of violent behavior and, (4) the final disposition of previous referrals.  As noted in this report, several youths (including many who were successful in this program) had violent incidents in their backgrounds.

 [2]  It should be noted that when considering the entire group of youths referred to the court during 1994, from which the sample was drawn, the overall recidivism rate was 58%, remarkably close to the rate for the sample as a whole.  Also, the overall rate of subsequent petitions was 50%.