PRELIMINARY RESULTS OF

THE 1993 PRISON WARDEN SURVEY

Principal Investigator: George W. Knox, Ph.D.

                          Director

                     National Gang Crime Research Center

November 29, 1993

ABSTRACT

  Described here are the results from a mail questionnaire research project. All state adult correctional institutions were included the sample frame. The usable sample consisted of N = 174 respondents. This represents over a fourth of all such institutions in the sample universe. It also included responses from 41 states.

    The research methodology is described, and basically involved a short time frame from data collection to providing this preliminary report.

    The descriptive statistical findings are reported for all quantitative variables in the survey. Many issues --- including the gang issue --- are discussed and examined here as they affect adult state correctional institutions in the United States.

     Additional analysis was reported on the differences, if any, between the correctional institutions by various factors and problems in terms of whether the same facilities were accredited by ACA. For most such variables and problems examined, no significant differences emerged.

    A more detailed discussion and analysis was also undertaken and reported on addressing the issue of estimating the scope of the "gang problem" in American adult state correctional institutions. Looking at the "gang problem" as consisting of several different variables, that is that it is a "multi-dimensional" problem, helps to clarify the issue. What it also means is that the "gang problem" is somewhat higher than is actually estimated by a single simplistic question such as "do you have a gang problem in your facility".

     Recommendations for future research and improving the focus of gang analysis are also included in this report.

- ii -

PRELIMINARY RESULTS OF THE 1993 PRISON WARDEN SURVEY

INTRODUCTION

     The preliminary results from the 1993 Adult Corrections Survey are reported here. Descriptive statistical findings are provided for all quantitative variables from this survey. More detailed analysis is also provided for some of the issues examined in the survey.

THE SURVEY

     The 1993 Adult Corrections Survey is a seven-page questionnaire. Nearly two-hundred quantitative variables are taken from this survey and the results are presented in this report. However, there are other "open-ended" questions in the survey which are not analyzed nor reported here. The reason is that these questions elicit narrative information --- words essentially ---and a content analysis is required to meaningfully interpret this kind of variable. No content analysis is included in this preliminary report. Rather, this report focuses exclusively on the forced-choice questions in the survey and on those questions that elicit numbers only.

     The reason for this delimitation of the focus of the report is a very practical one: basic results of the survey were promised to the respondents in a very short period of time. That promise has been made and honored by this report .

METHODOLOGY AND SAMPLING

     The methodology used here is that of survey research, more specifically the use of a mail questionnaire. As a probono research project, not specifically funded, the costs of the project (mailing, photoduplication, etc) were borne by the research team itself. Every member of the research team was able to include several questions on the survey instrument. The research associates listed for this project were at the time either graduate students or other researchers associated with the National Gang Crime Research Center.

      The sampling technique used for this survey research was that of a saturation sample method. Every adult state correctional institution listed in the Directory of the American Correctional Association was included in the sample. This included all institutions and boot camps. Approximately N = 600 such survey instruments, with cover letters, and return envelopes were mailed out on 9-27-93 and on 10-4-93. By the time frame of 11-8-93 some N = 174 institutions had completed the questionnaires and returned them in the return envelope provided. This is the sample used for the present analysis.

      While other surveys may be expected to continue to "arrive", most of the data in survey research projects such as this come in right away or it does not come in at all. Thus, it is unlikely that a few additional surveys that may arrive after the arbitrary "cut-off" date of 11-8-93 would substantially change the nature of the findings reported here.

      One methodological problem that this kind of project faces by seeking a rapid turnaround in terms of data collection and reporting of results, is that some states have bureaucratized the ability of wardens or superintendents in terms of being able to respond to such surveys. California, Minnesota, Illinois, Michigan, New York, and Virginia have policies in place that require even the warden to be subject to strict "research review and prior approval" policies. While it is our position that a warden or superintendent is a public official, and that this kind of research is therefore exempt from such "human subject review" processes that were designed to protect the rights of inmates (e.g., from medical and psychological experimentation), there is no doubt that this kind of policy inhibited to some extent the response rate to the present survey.

     The questions and issues addressed in this survey are those that have appeared in the recent criminological and correctional literature. None of the questions in the survey address the religious beliefs, the sexual preferences, or very private aspects of the respondents. The questions address issues about problems such as gangs. The questions measure beliefs about the problems, and estimates of the scope and extent of the problems. Other questions simply elicit what is basically public information: population size, facility type, date the physical plant was constructed, etc.

DESCRIPTIVE FINDINGS FROM THE SURVEY

     The procedure here will be to present the descriptive statistical findings from the survey in the same order in which the questions appeared on the survey instrument. An important analytical factor is that in the findings described in this section, the results reflect the entire sample size. That is, in this section no effort is made to partial out certain factors or to control for other variables in the presentation of results. This section, then, describes the raw data.

     The respondents were asked to rate the extent to which they believe that the lack of post-release employment assistance contributes to the recidivism rate of the inmates released from their facilities. On a scale between zero (No Problem) and nine (Large Problem), 70.8 percent of the respondents gave a rating of five or higher. The mean or average value was 5.6 on this scale. Obviously, most state corrections facility administrators truly believe that to some extent the lack of post-release employment assistance is indeed a factor exacerbating the recidivism problem.

     When asked whether crimes committed by gang members while in prison should be subject to federal prosecution, 45.3 percent of the respondents said "yes". Some 54.7 percent disagreed with extending federal jurisdiction to the prosecution of crimes committed by gang members while in state custody.

       The wardens and superintendents of the state correctional adult institutions responding to the survey were also asked "What percentage increase in your budget would be necessary to assure no overcrowding, adequate staff training and inmate services?". Those that claimed they needed "no" increase in their budget --- to assure that overcrowding was not a problem, to assure that adequate staff training was not a problem, and to assure that providing inmate services was not a problem --- were clearly the exception and accounted for a very small percentage of the respondents. In fact, 67.8 percent of the respondents indicated that they would need their budgets increased 25 percent or more. Indeed, some 40.1 percent indicated they would need their budgets raised fifty percent or more. The mean or average value was 42.3 percent.

     The wardens were also asked "Do you believe prison gangs could be more effectively controlled if gang members could be transferred to a central-national federal unit?". Some 41.8 percent agreed with this idea. Some 58.2 percent said "no", they do not believe in this idea.

     The wardens were also asked for separate estimates among male and female inmates for gang density. Gang density means the percentage of inmates who are gang members. The specific language of the item was "Among staff who know about gang members, what is the current estimate of what percentage gang members are of the total inmate population?". The reason for this particular wording is that as an outgrowth of prison gang problems, the modern adult correctional organization now includes some gang specialist role. The gang specialist in corrections is known by a variety of titles: gang coordinator, gang intelligence officer, etc, whose duty it is to monitor gang activity. Overall, among all respondents to the survey, the results for males ranged between a low of zero percent to a high of 75 percent for this gang density variable. Overall, among all respondents the range for female inmates was zero percent to a high of 20 percent. The mean or average for males was 12.26 percent; and 2.3 percent for females.

     When asked if their staff receive formalized training in dealing with the gang problem, some 46.8 percent of the respondents indicated "yes". Thus, some 53.2 percent of the wardens report that their staff do not receive formalized training in dealing with the gang problem. Some N = 79 respondents provided information on the nature of such gang training that was provided; and such "gang training" ranged from a one hour session to a 48 hour training program. However, 88.6 percent of those reporting data for the hours of such gang training indicated that their officers received eight hours or less of such "gang training". Thus, the vast majority of such training appears to consist of a low intensity one day training exercise of unknown quality. The average or mean number of gang training hours was 5.8 hours.

     Further, among those respondents reporting their staff do receive such gang training, the data indicate that for just over half of the sample (56.5%) that only five of their staff actually received such training. Additional analysis of this data may prove worthwhile by looking at the multiplicative relationship between number of staff training and the hours of such gang training; however, that is beyond the scope and purpose of this preliminary report.

     When asked if gang members have been a problem in terms of assaults on their staff, some 18 percent of the wardens responding to this survey indicated "yes".

     Some 32.5 percent of the respondents indicated that gang members have been a problem in terms of threats against staff.

     Some 55.9 percent of the respondents reported that racial conflicts are a problem among the inmates in their facility.

     Some 56.7 percent of the respondents reported that white inmates have a separate gang in their facilities.

     The respondents were asked if in their opinion, giving staff recognition to prison gang leaders is similar to negotiating with terrorists. Some 59.3 percent of the respondents agreed with this analogy.

     Some 104 of the 174 respondents provided data in response to the question "In what year did gangs first become recognized as a problem in your facility". Those that claimed to have no gang problem therefore provided no such data. Thus, at least 59.7 percent of these adult state correctional facilities do report having a gang problem in the sense of providing an exact year in which the gang problem first appeared. The range for this factor varied between a low of 1956 to a high of 1993. It is interesting to note, however, that 80.8 percent of those reporting such yearly data indicated that the gang problem first arose in their facility since the year 1984. In fact, 48.1 percent of these respondents indicated the gang problem arose in the 1990's (1990 or after). These findings are consistent with similar research on law enforcement agencies showing the recent general proliferation of the gang problem nationwide. The mean or average for this variable on the year gangs were first recognized as a problem was 1987.

      Some 69.3 percent of the respondents to the survey endorsed the belief that federal agencies should play a greater role in the investigation and prosecution of gang crimes.

      Only 10.5 percent of the respondents indicated that their state has a separate correctional facility for confidential informants. These are sometimes called "snitch farms".

     Only 2.3 percent of the respondents to the survey answered "yes" to the question "do you as an administrator receive any pressure from state officials to 'play down' gang activity".

      The survey also asked if during the last twelve month period whether there have been any disturbances related to gang members in their facility. Some 31.6 percent reported that such disturbances by gang members had occurred. Similarly, 35.1 percent of the respondents reported that during the last twelve months that there have been disturbances related to racial conflict.

      The survey included a question on sexual assault. It asked the respondents how many sexual assaults were actually reported by inmates during the last twelve month period. Data on this item was available for 167 correctional administrators. Some 54.5 percent indicated "none". However, this factor ranged from a low of zero such incidents to a high of 51 such incidents. The mean or average number of such sexual assaults was 2.22 for the sample. A follow-up question to this "official rate" focused on the "estimated rate" of sexual assaults. It was phrased "estimate the total number of sexual assaults that occurred in your facility during the last twelve month period whether or not they were actually reported". The findings for this variable were obviously somewhat higher, and ranged from a low of zero such incidents (46.4%) to a high of 510 such incidents. Some 14.6 percent of the respondents estimated ten or more such sexual assaults. In fact, the mean or average for this variable was 9.7 such incidents.

      Only 9.4 percent of the respondents to the survey (N=16) reported that any criminal prosecutions have resulted from sexual assaults within their facility within the last twelve month period.

      The survey asked the correctional administrators whether they believed that correctional officers may face a greater actual day-to-day danger of physical confrontation than do police officers on the street. Some 71.1 percent of the respondents indicated support for such an assertion. This is not to say that correctional officers face more objective life-threatening danger than do police officers on the street. As one California correctional officer remarked to the principal investigator about this issue, "a prison is probably the best place to get stabbed or assaulted if you have to get it anywhere....because emergency medical services are readily and quickly available....unlike a police officer who patrols a large geographical area and can be victimized in a location some distance from the nearest trauma center, thereby increasing the danger factor".

      Some 90.7 percent of the respondents expressed the belief that correctional officers face ethical dilemmas in their everyday work.

     The respondents were also asked "To what extent does stress relate to your work environment" and were asked to indicate their answer on a scale of values between zero (NONE) and ten (A LOT). Some 87.2 percent provided a rating of "seven" or higher on this scale. The mean or average score on this variable was 8.

       The respondents were also asked "Are there any means available to you in your work place for you to reduce stress?". Some 55.7 percent of the respondents indicated that such means were available in their work place to reduce stress.

      The survey asked "what types of rehabilitation programs are available in your facility". The results show 97.7 percent of the respondents reported having "education (high school diploma or GED)" programs. Some 55.7 percent indicated that "college education" was available in their facility. Half (50%) reported offering "technical training". Some 55.2 percent indicated that inmates could receive an "A.A." degree. Some 79.3 percent reported offering "vocational education" programs. Some 19 percent indicated that inmates could receive a "B.A./B.S." degree. And 5.2 percent indicated that inmates could do "graduate work" in college in their facility.

     When asked to chose the one most effective type of rehabilitation program from the above list, most (62.8%) chose the general education (education, high school degree or GED) component. Some 4.4 percent indicated the "college education" component. Only .7 percent indicated the "technical training" component. Only 2.9 percent indicated the "A.A." degree component. Some 28.5 percent chose the component of "vocational education". No one chose the "B.A./B.S." college degree component as the most effective type of rehabilitation program in their facility. And only one respondent (.7 percent) chose the "graduate work" component. It would be valuable for future research to focus on whether such program resources are allocated and distributed along the same dimensions of perceived effectiveness within the same facilities. Some states are mandated in state laws to provide certain types of programs. Some 17.3 percent of the respondents reported that they would be the chief negotiator in a hostage situation.

      The data shows that 72.3 percent of the respondents never faced such a hostage situation. Among those facilities that did report that their facility had experienced a hostage situation involving inmates and staff, when asked "when was the last time" this occurred, the findings ranged from 1970 to 1993. Basically, about a fourth of the respondents (27.7%) report that their facility has experienced such a hostage situation.

     Some 45 percent of the respondents indicated that it is legally possible in their states to make inmates reimburse the state for the expenses of their incarceration . Such laws become attractive tools to deal with highly manipulate inmates. The Illinois version of this law was recently used to sue John Wayne Gacey who was collaborating with an outside contact selling his art work and "autographs". The recent case of a gang member locked up in the California Department of Corrections who earned a very substantial amount of money for his life story (Monster) is another case in point where such laws might be put to good use. The survey also asked how many staff members in their facility have tested positive for the PPD (tuberculosis) test in the last twelve month period. The results show a range between a low of zero (50.4%) to a high of 180 such cases. In fact, 9.8 percent of the respondents reported having ten or more staff who tested positive for the PPD test during the last twelve month period. Basically, about half of the respondents (49.6%) reported that one or more of their staff had tested positive for the PPD test during the last twelve month period, making this an issue of some concern.

The mean or average value for this variable was 4.57 staff members testing positive for the PPD test within the last twelve months. A total sum of 613 such officers testing positive for the PPD test was reflected in the sample.

     The survey also asked the respondents "in your opinion, have prison gangs tended to result in more improvised weapons production (e.g., shanks, etc) among inmates in your facility". Some 40.5 percent indicated support for this belief.

      About three-fourths (76.2%) of the respondents expressed support for the notion that tougher laws are needed to control the gang problem in prisons.

       Most of the respondents (87.8%) felt that their Department has an effective Affirmative Action program.

        The survey also asked "in your estimate, what percentage of all institutional management problems in your facility are caused by gangs or gang members". This is the gang disruption variable, and it measures the extent of the prison gang problem in terms of overall inmate management problems. The results for this factor varied between a low of zero percent (29.6%) to a high of 95 percent. The mean or average for this variable was 16.4 percent.

     Similarly, the survey asked "in your estimate, what percentage of all violence among inmates in your facility is caused by gangs or gang members". The results varied between a low of zero (29.6%) to a high of 98 percent. The mean or average for this variable was 20.4 percent.

     The survey also asked if "during the last 12 month period, have any inmates in your facility been diagnosed with tuberculosis". Some 62.9 percent indicated "yes".

       About half of the respondents (48.5%) reported that their facility is ACA accredited.

       The survey also inquired about the "set off" method for controlling inmate gang members. The "set off" method involves balancing the number of rival gang members in the same living unit, cell-house, wing, etc. When asked if in their opinion this is an effective way to control prison gangs, some 36.6 percent indicated "yes".

      The survey also inquired about the "circuit" method, sometimes called "diesel treatment" or "bus therapy", but simply means transferring gang leaders to other institutions. When asked if they felt this was an effective way to control prison gangs, some 59.9 percent of the respondents indicated "yes".

      The total inmate population being held in their facilities as reported by the respondents ranged from a low of 52 inmates to a high of 5800 inmates.

     The security level of the facilities, as reported by the respondents, was as follows: 23.5 percent minimum security; 42.4 percent medium security; and 34.1 percent maximum security.

      The survey asked "do you believe that providing tuition support for staff could help control the prison gang problem". Some 45.3 percent indicated support for this idea.

       When asked in what year the physical plant for their facility was constructed, the dates ranged from 1836 to 1990. Half of the respondents (50.6%) reported that their physical plant was constructed on or after 1966. Some 38.6 percent reported that their physical plant was constructed on or before 1950. The mean or average value of 1953.7 emerged for this variable.

     About a fifth of the respondents (20.7%) reported that their institution has any full-time staff employed in the role of an ombudsman for inmates.

      Some 83.5 percent felt that the Supreme Court has gone too far on ruling in favor of inmate rights.

      Some 85.7 percent felt that gang members tend to have a higher recidivism rate.

      Some 61.8 percent of the respondents reported that their institutional classification system takes gang membership into account.

       Using a standard "check-off" list, the respondents were asked to report what strategies their facility uses to control gangs. The percentage distribution for those using these various gang control strategies were as follows: 73 % transfers; 53.4% use of informers; 57.5% segregation; 50% isolating leaders; 35.1% lockdown; 29.9% prosecution; 39.1% interrupting communications; 50% case by case dealings; 4.6% ignoring their existence; 4% infiltration; 52.3% displacing members to different facilities; 5.7% coopting of prisoners to control gangs; 13.2% meeting with gang leaders on "as needed" basis; 5.2% joint meetings between various gang leaders; 19.5% balance the number of rival gang members living in the same unit (e.g., the set-off method); 45.4% Task Force to monitor and track gang members; 42.5% locking up gang leaders in high security facilities; and 8.6 percent "other".

      The survey asked "how many felony crimes committed by inmates in your facility were court prosecuted during the last year". The results ranged from a low of zero (33.5%) to a high of 100 such incidents. A mean or average value of 7.2 such felony crimes, a sum of 1170 emerged for this variable. A follow-up question that gets at the question of how many felony crimes occur "behind bars" each year that may not show up in the FBI's Uniform Crime Report, nor the NCS victimization surveys, asked "in your estimate, how many felony crimes committed by inmates in your facility COULD HAVE LEGALLY BEEN PROSECUTED UNDER STATUTE during the last year". The results ranged from a low of zero (21.2%) to a high of 556 such incidents. The mean or average value for this variable was 21.3, with a sum or total of 3352 such felonies.

      The self-reported "recidivism" variable comes from the question "in your best estimate, what percentage of the inmates confined in your institution have served time with you before". The results ranged from a low of zero percent (2.5%) to a high of 100 percent. The mean or average value of 35.7 was found for this variable.

     Some 41.4 percent of the survey respondents felt that prison gangs have significantly affected their correctional environment.

     The survey asked the state adult correctional administrators if they believe that the United States Department of Justice has provided effective leadership in suppressing the gang problem in American cities. Only 7.3 percent of the respondents said "yes"! With 92.7 percent of the respondents who deal with offenders on a daily basis answering "no" to this question, it would appear there is scant support for the idea that the federal government enjoys a reputation for effective aggressive pro-active initiatives in dealing with the national gang problem. This research finding is particularly critical in light of the fact that many of the crimes that gangs engage in are interstate and federal in nature.

      The survey also asked the respondents to estimate what percentage of all illicit drugs brought into their facility was accomplished by prison gang members. The results ranged from a low of zero percent (28.5%) to a high of 100 percent. The mean or average value for this variable was 27.2 percent. This is the gang drug smuggling variable and basically measures the extent to which prison gangs control the illegal drug smuggling and importation inside the prison inmate economy. Future research would improve the value of this line of inquiry by including estimates of the economic value of such illegal drug operations.

     The survey did ask a separate question on this issue which asked the respondents to estimate "what percentage of the illicit drug trade in your facility is dominated by prison gang members". This is the gang drug sales variable and basically measures the extent to which prison gang members control the drug trade behind bars. The results ranged from a low of zero percent (29.9%) to a high of 100 percent. The mean or average value for this variable was 31.4 percent.

      Some 47.7 percent of the respondents reported that overcrowding is a problem in their facility.

     The survey include the question "please estimate how many violent assaults among inmates that resulted in serious or life-threatening injuries happened during the last twelve month period". The results ranged from a low of zero such incidents (47.9% of the respondents) to a high of 100 such incidents. The follow-up question to this was "please estimate what percentage of these inmate-against-inmate assaults involved gang members", and the results varied between a low of zero (57.2%) to a high of 100 percent (mean = 17.4%).

       Only 4.1 percent of the respondents reported that any firearms had been confiscated from inmates or have been found in their facility during the last twelve months.

      Some 18.1 percent of the respondents reported that gang members, generally, are considered high escape risks in their facility.

       Some 19 percent of the respondents reported that they had staff who belonged to the state or regional Gang Investigators Association.

       The zip code was analyzed from the survey to ascertain which states were represented in the sample. Nine states were not represented in this sample response at the time of the arbitrary data collection cut-off point for this Preliminary Report: Hawaii, Montana, New York, South Dakota, Tennessee, Utah, West Virginia, Wisconsin, and Wyoming. Thus, the sample for this analysis includes administrators of adult state correctional facilities from 41 states.

DOES ACA ACCREDITATION SIGNIFICANTLY DIFFERENTIATE BETWEEN FACILITIES WITH AND WITHOUT VARIOUS PROBLEMS?

     The variable reflecting whether or not the facilities managed by the respondents to this survey were accredited by the American Correctional Association (ACA) showed that about half of the sample enjoyed this certification recognition (48.5% were ACA accredited). But are ACA accredited facilities really different than those who lack such accreditation in terms of the various problems reported by the administrators of the adult state correctional institutions in this sample? The analysis that was undertaken here looked at a large number of variables.

      First, let us summarize those variables for which there is no significant difference between ACA accredited facilities and those that lack such accreditation. The following are those problem areas and variables examined in this survey where no significant difference exists:

       *** Whether gang members have been a problem in terms of assaults on staff OR threats against staff.

      *** Whether racial conflicts are a problem among the inmates in the facility.

      *** Whether the respondent's state has a separate correctional facility for confidential informants.

      *** Whether the administrator receives any pressure from state officials to "play down" gang activity.

      *** Whether any criminal prosecutions resulted from sexual assaults within their facility during the last twelve month period.

      *** Whether the respondent believed that correctional officers may face a greater actual day-to-day danger of physical confrontation than do police officers on the street.

      *** Whether the facility of the respondent provided any of the seven "rehabilitation" programs ranging from basic education (high school degree or GED) to graduate work . Nor did estimates of the effectiveness of such programs vary significantly by the ACA accreditation variable.

      *** Whether it is legally possible in their state to make inmates reimburse the state for the costs of their incarceration.

      *** Whether any of the inmates in their facility have been diagnosed with tuberculosis during the last 12 months.

      *** Whether or not their staff receive formalized training in dealing with the gang problem.

      *** Whether or not gang members have been a problem in terms of assaults on staff AND threats against staff.

      *** Whether or not white inmates have a separate gang.

      *** Whether or not during the last 12 months there have been any disturbances related to gang members in their facility.

      *** Whether or not during the last 12 months there have been any racial disturbances in their facility.

      *** Whether or not prison gangs have tended to result in more improvised weapons production in their facility.

      *** Whether or not their Department has an effective Affirmative Action program.

      *** Regardless of the security level of the institution of the respondent (minimum, medium, maximum).

     *** Whether or not the facility employs full-time staff as ombudsmen for inmates.

     *** Whether or not their institutional classification system takes gang membership into account.

     *** Whether or not prison gangs have significantly affected the correctional environment.

     *** Whether or not prison overcrowding is reported as a problem.

      Thus, for most variables examined there was no statistically significant difference using the Chi-square test (p < .05) in examining the effects of ACA accreditation on these various factors. Only two factors differed significantly in terms of this ACA accreditation variable.

     Table 1 shows the relationship between those facilities where "stress reduction" means are available in relationship to ACA accreditation. Clearly, ACA facilities are significantly more likely to report having this capability (Chi-square = 5.98, p = .01). The Chi-square test compares the "expected frequency" in a normal random distribution with the "observed frequency" in a table. The larger the value of the Chi-square test statistic, the stronger the association exists between the two variables. Thus, when it is significant, it means they are not independent, they are in fact related: one variables significantly differentiates the other.

TABLE 1

FREQUENCY DISTRIBUTION OF WHETHER ANY MEANS ARE

AVAILABLE IN THE FACILITY FOR STRESS REDUCTION

IN RELATIONSHIP TO ACA ACCREDITATION

                        Is the facility ACA Accredited?

                            No Yes

Are there any means

available to you in

your work place for you

to reduce stress?

                    NO 46 28

                    YES 42 55

                        Chi-square = 5.98, p = .01

      As seen in Table 1, some 47.7 percent of the non-ACA accredited facilities had the "stress reduction" in the work place capability compared to 66.2 percent of the ACA accredited facilities.

     The other factor on which a significant difference existed by the ACA accreditation variable was that of whether or not the respondent would be the chief negotiator in a hostage situation. This finding is shown in Table 2 below. This shows that the respondent (typically the warden or superintendent) of the non-ACA accredited facilities were more likely to be the chief hostage negotiator in such an incident. Some 22.7 percent of the non-ACA accredited facilities indicated the respondent would be the chief negotiator, compared to 10.8 percent among ACA-accredited facilities. Future research would improve our knowledge about this issue by focusing on the extent to which actual specialized training has been provided to correctional officers in hostage negotiation.

TABLE 2

FREQUENCY DISTRIBUTION OF WHETHER THE RESPONDENT

WOULD BE THE CHIEF NEGOTIATOR IN A HOSTAGE SITUATION

BY WHETHER OR NOT THE FACILITY IS ACA ACCREDITED

                           Is the Facility ACA Accredited?

                                NO YES

Would you be the chief

negotiator in a hostage

situation? NO 68 74

                      YES 20 9

                             Chi-square = 4.28, p = .03

      So the answer to the question: does ACA accreditation really distinguish between these facilities in terms of a number of common problems and situations, is for the most part, and at least with regard to the factors examined here, NO. Two such differences did emerge, however, as explained above that are at least consistent with the recognition for meeting professional standards that is implied by ACA certification. The fact, though, that this ACA accreditation variable did not differentiate such facilities in terms of a number of other variables tends to suggest the value for future research in actually comparing certain organizational capabilities and "standards" compliance with regard to having achieved certification, for it is often remarked by officials that the only reason they lack such certification is the old age of their physical plant itself.

HOW EXTENSIVE IS THE GANG PROBLEM IN ADULT STATE CORRECTIONAL INSTITUTIONS TODAY?

     Vermont is about the only state in the United States that has not surfaced in the research literature as showing or admitting a gang problem to some extent at the various levels of the criminal justice system. It is valuable here to take a closer look at the data from the 1993 Prison Warden Survey to answer a basic question: what percentage of such adult state correctional institutions in America currently have a gang problem? And how serious is it?

      Recall from the descriptive summary discussed earlier in this report, that several different variables can help us answer this basic question. First, we get a rough approximation of the answer from the question that measured "in what year did gangs first become recognized as a problem in your facility? 19____". From that variable, the fact that N = 104 of the respondents provided an actual year in response to the question, would suggest that at least 59.7 percent of the institutions have such a gang problem.

     Recall, as well, that when asked "do you believe that the prison gangs have significantly affected your correctional environment" that some 41.4 percent said "yes".

      It is true --- and it could be nothing more than the problem of "denial" itself --- that a number of correctional institutions in the present survey claimed they "had no gangs". Because of this a more detailed analysis was made here. A subsample was identified as the "no gang problem" group consisting of those not providing an actual year in response to the question "what year did the gang problem first arise in your facility". Before summarizing some of the other findings from this "no gang problem" group, it is useful to remark that it is at least plausible that the respondent simply did not know the exact year in which gangs first surfaced as a problem in that particular facility. That is, simply lacking this information may not mean "denial" per se.

      Characteristics of the Group of Respondents Who Claimed "No Gang Problem" or who "left a blank answer" to the Question Regarding the Actual Year in Which Gangs First Became Recognized as a problem in their facility. What this subgroup represents is truly not a completely "non-gang" problem is what the findings of the present analysis reveals. This subset analysis looks at those who did not provide data or a quantitative response to the question about the year in which gangs were recognized as a problem in their correctional facility. By looking at their responses to other gang problem variables, what emerges is obviously not a complete absence of the problem.

     Perhaps the real "problem" here simply linguistics. It is possible to have gang members in a facility and still --- technically --- not have a "gang problem". They may simply be dormant, inactive, neutralized, not of sufficient density or critical mass in terms of membership to be formidable, or alternatively they may simply be truly "underground" or "secret". A lot of possibilities exist.

       Consider these basic findings from the survey among those who did not respond to the question about the year in which prison gangs were first recognized as a problem:

      *** On the question about the percentage of male inmates who are gang members, the range was between zero (27.6%) to a high of 75 percent.

     *** On the same question about the percentage of female inmates who are gang members, the range was between zero (65.6%) to a high of 10 percent.

     *** Some 24.3 percent indicated their staff receive formalized training in dealing with the gang problem for which they were not sure what year first arose.

     *** Some 4.2 percent report that gang members have been a problem in terms of assaults on staff; and 6.5 percent report gang members have been a problem in terms of threats on staff.

     *** Some 15.9 percent report their white inmates have separate gangs.

     *** Some 1.4 percent report they receive pressure from state officials to "play down" gang activity.

     *** Some 2.8 percent Report that there have been disturbances related to gang members in the last twelve months; and 12.7 percent report racial disturbances during the same period. Racial conflict is often a surrogate measure of the gang problem as it is one of the factors that fuels the fire of the gang problem.

     *** Some 13.3 percent report that prison gangs have tended to result in more improvised weapons production.

     *** Some 31.7 percent report that gangs or gang members account for at least some percentage of all institutional management problems in their facility.

     *** Some 27.9 percent report that gangs or gang members account for at least some percentage of all violence among inmates in their facility.

     *** Some 78.2 percent felt gang members tend to have a higher recidivism rate.

     *** Some 47.1 percent report that their institutional classification system takes gang membership into account.

     *** A lot still use various gang control strategies even though no date was provided for the year in which gangs first became recognized as a problem:

            --- 47.9 percent use Transfers

            --- 29.6 percent use Informers

            --- 35.2 percent use Segregation

            --- 26.8 percent use Isolating Leaders

            --- 18.3 percent use Lockdown

            --- 16.9 percent use Prosecution

        --- 16.9 percent use Interrupting Communications

            --- 36.6 percent use Case by Case Dealings

            --- 1.4 percent use Infiltration

            --- 31 percent use Displacing members to

                 different facilities

            --- 1.4 percent use Coopting

--- 4.2 percent use Meeting with gang leaders on "as needed" basis

            --- 1.4 percent use Joint meetings between

                 various gang leaders

            --- 9.9 percent use Balance the number of rival gang members living in the same unit

            --- 25.4 percent use a Task Force to monitor and track gang members

            --- 21.1 percent use Locking up gang leaders in high security facilities

    *** Some 10.4 percent report that gang members have significantly affected their correctional environment.

    *** Some 26.9 percent report that gang members account for at least some percentage of all illicit drugs brought into their facility AND for gang control of the inmate drug market.

     *** Some 6.5 percent report that gang members account for at least some percentage of inmate-against-inmate assaults that resulted in serious or life-threatening injuries.

      *** Some 14.5 percent consider gang members high escape risks.

      *** Some 4.5 percent report they have staff who belong to the state or regional Gang Investigators Association.

      The conclusion that must be reached here is that by simply lacking the exact year in which gangs first became recognized as a problem in the facility of the respondents to this survey is not sufficient to represent a "no gang problem" condition. Thus, in order to answer the question about what percentage of the facilities really have a "gang problem" will require looking at the inter-relationship between at least two or more variables. Because this one variable (year in which gangs first became recognized as a problem) is not sufficient to truly separate those facilities that do or do not report gang problems.

       An Multi-Dimensional Approach To Answering the Question. Let us begin with the logic that there is value in using the variable of "in what year did gangs first become recognized as a problem" because it is doubtful that any of these problems have really disappeared or discontinued. The only theoretical condition in open-systems theory where the problem could be quickly eradicated would be that of pure displacement: shipping out all gang members to some other facility, and ensuring no new gang members entered the facility. It is helpful, then, to look at the central tendencies of the data from this sample as an actuarial approach to ascertaining how much more than the assumed base rate of 59.7 percent (the percentage who provided a specific year in response to the question on when the gang problem was first recognized in their facility) a preferred estimate may be.

     The statistical approach used here is not dissimilar from that of the actuarial method. The sample is divided into successively smaller subsamples; looking at factors of risk. Each new subsample eliminates those cases previously examined as "risks" and thus the subsample sizes become progressively smaller as new risk factors are examined. Figure 1 below summarizes the results of this line of inquiry. As seen in Figure 1, looking at entirely different variables which measure different aspects of the "gang problem" and its deleterious impact on correctional institutions, it may clearly be stated that under these more specifically defined conditions that were in fact measured in the survey, that the percentage of adult state correctional institutions reporting some such important ingredient of the gang problem may be as high as 76.4 percent using only nine additional factors of analysis.

      Factor 1 reflects the fact that among those in the sample who did not provide data for the question about "in what year did gangs first become recognized as a problem in your facility", seven cases were found where the respondents did in fact also state that gangs had significantly affected their correctional environment. In the sample analyzed next, those seven cases were deleted for purposes of analysis. Thus, in analyzing factor 2 the sample surviving cases subject to analysis decreased by seven. Factor 2 examined the percentage of male inmates who were gang members, and there were 7 cases where the estimate was 15 percent or higher; thus, with Factor 2, the cumulative percentage of respondents meeting the "gang problem" criteria increased to 67.8 percent as seen in Figure 1.

     Factor 3 was the variable for whether or not white inmates had a separate gang. Obviously, if white inmates have a separate gang, it might be said that the same facilities must have a "gang problem". Thus, this increased the cumulative "gang problem" percentage to 71.2 percent because there were six cases where the remaining sample answered yes to the question whether or not white inmates had a separate gang in their facilities.

FIGURE 1

PERCENTAGE OF ADULT STATE CORRECTIONAL INSTITUTIONS WITH A “GANG PROBLEM” USING THE MULTI-DIMENSIONAL METHOD

                                         "Gang Problem"

                 INCREMENTAL EQUATIONS Rate (%)

                 Base = 59.7

                 Base = Base + Factor 1 62.5

                 Base = Base + Factor 2 67.8

                 Base = Base + Factor 3 71.2

                 Base = Base + Factor 4 71.8

                 Base = Base + Factor 5 72.4

                 Base = Base + Factor 6 74.7

                 Base = Base + Factor 7 75.2

                 Base = Base + Factor 8 75.8

                 Base = Base + Factor 9 76.4

     Factor 4, whether gang members have been a problem in terms of assaults on their staff, increased the "gang problem" estimate to 71.8 percent (one case where this "gang assault on staff" factor was found in the remaining sample).

     Factor 5, whether during the last twelve month period there have been any disturbances by gang members in the facility, increased the "gang problem" estimate to 72.4 percent (again, only one case where this additional risk condition was found).

     Factor 6, reflecting four cases where the respondent expressed the opinion that prison gangs have tended to result in more improvised weapons production (e.g., shanks, etc) among inmates in their facility, increased the "gang problem" estimate to 74.7 percent.

     Factor 7 captured one more case in the remaining sample, increasing the "gang problem" estimate to 75.2 percent. Factor 7 included a case where 15 percent of all institutional management problems in the facility were estimated to be caused by gangs.

     Factor 8 also involved only one case, and reflected the situation where the respondent estimated that ten percent of all illicit drugs brought into the facility was done so by prison gang members. Factor 8 therefore increased the "gang problem" estimate to 75.8 percent.

     Finally, Factor 9 also involved only one case, and this was where the respondent estimated that at least five percent of the illicit drug trade in the correctional facility was dominated by prison gang members. Thus, with the addition of Factor 9, this increased the "gang problem" estimate to 76.4 percent.

      A number of other possible factors could have been examined in addition to those included in Figure 1. The point of this analysis should be clear at this juncture. The "gang problem" is multi-dimensional in nature. It involves many different aspects of risk: in terms of crime, violence, management problems, conflicts, etc. To summarize what we have learned here, from the present survey our best or preferred estimate for: "what percentage of adult state correctional institutions" have a "gang problem" is somewhere near or above 75 percent. This would be more in line with the recent national findings on city and county law enforcement agencies in terms of what proportion of those law enforcement administrators report a "gang problem" in their jurisdiction .

SUMMARY, CONCLUSION, AND RECOMMENDATIONS

       This preliminary report described the basic raw data and provided a glimpse into some other detailed issues from the 1993 Prison Warden Survey. This mail questionnaire project was undertaken and reported in a very short period of time, but still had use of a response rate of slightly over a fourth of the sample universe. Which is not uncommon for mail questionnaire projects of this type. Obviously, the response rate could have been increased with sufficient resources, but this was an unfunded project which along with the time constraints did not permit "follow-up" reminders.

      This was the third national survey of the wardens and superintendents of state adult correctional institutions carried out and reported by the National Gang Crime Research Center. This survey included several new questions, reflecting the input of the research associates of this project. Some such questions like the one dealing with correctional officers who have recently tested positive for the PPD (tuberculosis) test proved to be valuable contributions and improvements to our knowledge of the problems faced by modern adult state correctional institutions. Obviously, the gang problem is not the only problem facing correctional administrators today. It is, however, from the present analysis; a major problem that would appear to be affecting more facilities than its administrators claim.

     Why this should be is in part an analytical problem, and in part a semantic and perhaps contextual problem. To use an analogy, it might not be uncommon for an inmate who is a drug addict, when asked by the prison psychiatrist "do you have a problem with drugs" to reply "No...I don't have a problem with drugs....I enjoy drugs." When a prison administrator is asked a simplistic question about a complex issue such as gangs, it may not be unusual to find only 59.7 percent say "gangs are a problem". But when different components of the problem are measured and examined separately, as aspects of the multi-dimensional nature of what is meant by the "gang problem", it may not be unreasonable to estimate that the rate of those facilities with a "gang problem" is substantially higher.

       This research also uncovered the need for improvements in future research efforts in assessing the scope and extent of the gang problem in corrections. An urgent need exists for providing correctional institutions with research assistance to profiles of gang members and provide threat analyses of prison gangs.

29 November 1993

Dear Respondent to the 1993 Survey:

      As promised, here is a free copy of the "Preliminary Report". Nearly 150 correctional facilities throughout the USA are receiving this report today. We want to thank you for your help in making this research possible. If we had the time and money, then I am sure the report would have been much longer: because there are a lot of issues that need more analysis. This data is very current and reflects a strong national representation. We will try at the National Gang Crime Research Center to provide further analysis in the near future.

       Enclosed, please find also a list of our other research and training products and information about the Gang Journal.

       One of our 1994 research and service priorities is to provide more localized in-depth analysis. We are looking for host sites that are willing to collaborate for such projects. From our research in other contexts, we can now put together "gang member profiles", and actually predict gang membership with a high level of accuracy on simple screening questions. There is a need to be able to apply this locally, using measures that may be more suitable to the correctional environment. There is also a need to use the same type of data for threat analysis purposes using the gang organization itself as the unit of analysis. Both can be accomplished through anonymous surveys administered to inmates. The idea is to survey an entire facility all at once. The idea is to have your input as well, collecting and analyzing factors that you think are important on your own population.

       If you think you would be interested in having such local knowledge developed for your own use in planning, management, and operations, please let us know. Or if there is any other way in which we may be of assistance, then please contact us.

       Once again, thank you for your vital assistance in making this research report possible.

       Cordially,

       George W. Knox, Ph.D.

       Director, National Gang Crime Research Center

GWK/yc

ENC: