ANALYSIS OF RETURN RATES OF THE INMATE COLLEGE PROGRAM PARTICIPANTS
STATE OF NEW YORK
DEPARTMENT OF CORRECTIONAL SERVICES
THE STATE OFFICE BUILDING CAMPUS
ALBANY, N.Y. 12226
THOMAS A. COUGHLIN III
COMMISSIONER
CHESTER H. CLARK
ASSISTANT COMMISSIONER
DIVISION OF PROGRAM PLANNING, RESEARCH
AND EVALUATION
EXECUTIVE SUMMARY
The present report examines the return rate of participants in the Inmate College Program. The report was prepared at the request of the Director of Education[1] and was designed to generate statistical data that would address the question of whether completing a college degree during a period of incarceration reduced the likelihood of participants being returned to the Department's custody following release.
This survey sampled all males who participated in the Inmate College Program during the 1986-1987 academic year. Females were not included in this study because the number of women who participated in the Program was too small to support analysis.
The Department's computerized data files were utilized to determine the number of these program participants who had been released. A cut-off date of February 28, 1990 was selected to insure a minimum follow-up period of 12 months as of February 28, 1991. There were 986 male program participants who had been released by February 28, 1990. Of this total, 356 were successful program participants who had earned a college degree in 1987, and 630 were unsuccessful program participants who were administratively removed or had voluntarily withdrawn from the program. Neutral program participants who were still working toward the completion of degree requirements were not included in the study.
Inmate College Program participants in 1986-1987 who had earned a degree were found to return at a significantly lower rate than participants who did not earn a degree. Of those earning a degree, 94 (26.4%) had been returned to the Department's custody by February 28, 1991 whereas 281 (44.6%) of those participants who did not earn a degree were returned to custody (see Figure 1). This difference between the return rates of participants who earned a degree versus those who did not was found to be statistically significant at the .001 level. Degree earning participants also returned at a lower rate than would be expected when compared to the overall male return rate. These findings suggest that earning a college degree while incarcerated is positively related to successful post-release adjustment as measured by return to the Department's custody.
PREPARED BY:
DAVID D. CLARK
PROGRAM RESEARCH SPECIALIST III
AUGUST 1991
![inmate college program graph](inmatecollegeprograms.jpg)
ANALYSIS OF RETURN RATES OF INMATE COLLEGE PROGRAM PARTICIPANTS
The present report examines the return rate of participants in the Inmate College Program. The report was prepared at the request of the Director of Education and was designed to generate statistical data that would address the question of whether completing a college degree during a period of incarceration reduced the likelihood of participants being returned to the Department's custody following release.
PROGRAM OVERVIEW
A number of public and private colleges currently operate a college program for incarcerated offenders in cooperation with the New York State Department of Correctional Services. Eligible inmates may enroll in certificate, two- or four-year degree programs, or masters programs with majors in selected fields.
A recent Department survey obtained responses from colleges participating in the program during the 1986-1987 academic year (see Appendix A for a list of reporting colleges). These colleges reported that 635 degrees were awarded. Of the 635 inmates who were awarded degrees, 218 (or 34%) were awarded certificates, 320 (or 50%) were awarded associates degrees, 96 (or 15%) were awarded bachelors degrees, and one inmate was awarded a masters degree.
SAMPLE SELECTION
Earning a college degree is a time dependent process that typically requires study for a minimum of one year to obtain a certificate, two years to earn an associates degree, four years to earn a baccalaureate, and one year of graduate study to earn a masters. While these time frames are generally accepted standards for obtaining a degree, it should be noted that individual students (both inmate students and students at large) may take more time to complete degree requirements. In a given academic year, it is possible to classify inmate college participants into one of the following three mutually exclusive categories.
1.) Degree requirements completed and degree awarded. Inmates in this category are considered successful program participants.
2.) Degree requirements being pursued with continued program participation. Inmates in this category are considered neutral program participants.
3.) Degree requirements no longer being pursued due to withdrawal or administrative removal. Inmates in this category are considered unsuccessful program participants.
Since the process of acquiring a college degree occurs over time, the information provided by the colleges for the 1986-1987 academic year amounts to a snapshot of their academic standing at one point in the process. It is a snapshot in the sense that the aforementioned categories reflect the classifications of inmate participants at the end of the academic year. Therefore, the information provided by the colleges will not support a study that tracks individual inmate progress in the College Program throughout the period of incarceration. Rather, it provides a cross-sectional view of academic standing for inmate college participants as a group.
As stated above, the purpose of this study is to ascertain whether obtaining a college degree during a period of incarceration reduces the likelihood of being returned to the Department's custody. Since neutral program participants neither obtained a degree, withdrew, or were administratively removed from the program during their period of study and incarceration there is no definitive outcome to measure that can be attributed to program participation. Therefore, colleges were asked to provide information only for successful and unsuccessful program participants.
For the 1986-1987 academic year, colleges responding to a Department survey reported that 1,688 inmates were successful or unsuccessful program participants. Of the 1,688 inmates for whom colleges provided information, 635 obtained degrees and were classified as successful participants, whereas 1,053 were administratively removed or voluntarily withdrew and were classified as unsuccessful participants.
In order to permit an adequate follow-up period of at least 12 months, a cut-off release date of February 28, 1990 was established. This selection criteria produced a sample of 1,012 Inmate College Participants (see Table 1). Table 2 shows the number of college program participants who had been released each year between January 1, 1987 and February 28, L991 by gender. These data show that of the 1,012 college program participants 986 were male and 26 were female. The 26 females were not included in this study because small changes in the number of females returned in a given release cohort would produce large changes in estimates of the proportion returned. For example, if 6 female inmates were released in 1987 and 2 were returned to custody, the return rate would be 33%. However, if 3 of the 6 females were returned to custody, the return rate would be 50%. When such small changes in the numbe r of females returned to custody produce such large changes in return rates, the return rates are considered to be inherently unstable and potentially misleading.
Of the 986 males, 334 were released in 1987, 404 in 1988, 214 in 1989, and 34 in 1990. Three hundred fifty-six (356) of these college participants were awarded degrees in 1987 while 630 were administratively removed or voluntarily withdrew from the program (see Table 3).
CALCULATION OF RETURN RATES
The return rates used in this report were derived by dividing the number r of program participants in the category of interest who were returned to the Department by February 28, 1991 by the total number of program participants in that category, For example, according to the data presented in Table 4, there were 116 male college program participants who obtained a degree and who were released in 1987. Of these 116 male college program participants, 43 had been returned by February 28, 1991. By dividing the number of returns for this category (43) by the total number of college participants in this category (116), a return rate of 37.1% is derived for degree completers released in 1987.
The data in Table 4 indicate, in general, inmate College Program participants who were awarded degrees in 1987 have a lower return rate than Inmate College Program participants who voluntarily withdrew or were administratively removed. A comparison of the total return rates shows that 26.4% of the participants who had been awarded a college degree were returned to the Department's custody, whereas 44.6% of those who withdrew or were administratively removed had been returned. These samples of Inmate College Program participants who earned degrees and those who did not were in the community for comparable time periods.
The data presented in Table 4 indicate that Inmate College Program participants in 1987 who earned college degrees were less likely to be returned to the Department's custody than Inmate College Program participants who did not. These data, therefore, suggest that earning a college degree is positively related to successful post-release adjustment.
COMPARISON TO OVERALL RETURN RATE
For general comparison purposes, the return rate of all male releases has also been calculated and compared to the return rates of Inmate College Program participants. The return rates for all- male releases can be constructed in the same manner as the return rates for program participants. It is simply the quotient of the number of male returns to the Department's custody during a given period of time divided by the total number of males released during that same period of time.
The Bureau of Records and Statistical Analysis is tracks all Department releases for a five year period to generate return rate statistics. Theses return rates are derived by examining the time to return to the Department's custody for a cohort of releases in a given year. For example, there were 10,818 male releases in 1986. Of these 10,818 releases, 5,133 had been returned to the Department's custody within a 48 month time period, generating a return rate of 47.4% (5,133/10,818).
The data in Table 5 present the number of Inmate College Program participants released in each year from January 1, 1987 through February 28, 1990 The return rate for all males released during the corresponding release year as described above was then used to construct an estimate of the number of Inmate College Program returns that would be expected irrespective of program participation.
In Table 6, these estimated number of returns based on the Departments overall male release population are compared to the actual number of returns observed. These data show that the return rate of Inmate College Program participants who had earned degrees was lower than the return rate that would be expected had they been selected solely as part of the male release population for the corresponding release year. In contrast, Inmate College Program participants who had not earned a degree returned at a higher rate than did the male releases for corresponding release years.
FINDINGS
This study suggests the following two findings:
1.) The sample of Inmate College Program participants who earned degrees returned to the Departments custody at a significantly lower rate than did those Inmate College Program participants who did not earn degrees. The difference between the degree earners and those who did not earn a degree was statistically significant (see Table 7).
2.) The sample of Inmate College Program participants who earned degrees returned to the Department's custody at a lower rate than would be expected had they been selected from the overall male release cohort corresponding to their release year. The sample of Inmate College Program participants who did not earn a degree returned to the Department's custody at a higher rate than would be expected had they been selected from the overall male release cohort corresponding to their release year.
DISCUSSION
These findings suggest that successful completion of a college degree may be positively related to post-release community adjustment. As mentioned in Finding 1 above, the difference in return rates between Inmate College Program participants who earned degrees and those who did not was statistically significant. This means that it is unlikely that the difference in return rates between degree earners and non-degree earners occurred by chance alone. This conclusion was derived in the following manner.
The chi square test statistic was used to test the hypothesis that there is no relationship between earning a college degree and being returned to the Department's custody. The chi square value obtained was 31.96 and is statistically significant at the .001 level. It can be concluded then, that there is a tendency for Inmate College Program participants who do not earn degrees to be returned to the Department's custody at a higher rate than participants who earn degrees and that this tendency is not likely to occur by chance alone.
All data sets have their limitation and this one is not an exception. The Inmate College Program data do not contain the depth of information necessary to examine the possible reasons why degree earners return to the Department at a lower rate than participants who do not earn degrees.
For example, it should be noted that the Inmate College Program data set did not contain information regarding participation in other prison programs. Therefore, if post-release adjustment is attributable to some degree to participation in prison programs, it is impossible to disaggregate the effects of multiple programs on post-release adjustment from available data.
Finally, these data do not contain the depth of information necessary to make inferences concerning the reasons why degree earners in the Inmate College Program return to the Department at a significantly lower rate than participants who do not earn degrees. Factors such as an individual's motivation to enter the Inmate College Program, changes in an individual's level of commitment to obtain a degree, and the level of support from family and friends in the community may affect an inmate's post-release community adjustment and the likelihood of being returned to the Department's custody.
CONCLUSION
In closing, the reader is cautioned against any definitive conclusion concerning the Department's Inmate College Program based on this follow-up study. However, the findings of this research as well as prior Department research support the position that earning a college degree during a period of incarceration is positively related to post-release adjustment.
1. In 1983, the Department conducted a follow-up study of a sample of inmates from selected college programs who had earned college degrees while incarcerated. This study found that inmates who earned a degree returned to the Department's custody at a lower rate than inmates who withdrew or were administratively removed. Recent legislative, academic, and Department queries regarding the Inmate College Program prompted the Director of Education to request a replication of the 1983 study with a more recent cohort of Program participants from a broader spectrum of colleges.
TABLE 1 NUMBER OF COLLEGE PROGRAM PARTICIPANTS ELIGIBLE FOR FOLLOW-UP NUMBER OF COLLEGE PROGRAM PARTICIPANTS RELEASED ON OR BEFORE 2/28/90 1,012 RELEASED AFTER 2/28/90 120 UNDER CUSTODY AS OF 2/28/91 556 TOTAL 1,688
TABLE 2 NUMBER OF COLLEGE PROGRAM PARTICIPANTS ELIGIBLE FOR FOLLOW-UP BY YEAR OF RELEASE AND GENDER NUMBER OF COLLEGE PROGRAM PARTICIPANTS GENDER FEMALE MALE TOTAL YEAR OF RELEASE 1987 8 334 342 1988 7 404 411 1989 9 214 223 1990 2 34 36 TOTAL 26 986 1,012
TABLE 3 NUMBER OF MALE COLLEGE PROGRAM PARTICIPANTS BY TYPE OF PARTICIPATION TYPE OF PARTICIPATION FREQUENCY PERCENT OF TOTAL WITHDRAWAL OR ADMINISTRATIVE REMOVAL 630 64% COMPLETED DEGREE 356 36% TOTAL 986 100%
TABLE 4 RATES OF RETURN FOR MALE COLLEGE PROGRAM PARTICIPANTS BY TYPE OF PARTICIPATION AND YEAR OF RELEASE WITHDRAWAL OR ADMINISTRATIVE REMOVAL PERCENT YEAR OF NUMBER NUMBER RETURNED RELEASE RELEASED RETURNED (RETURN RATE)* 1987 218 124 56.9% 1988 250 124 49.6% 1989 146 29 19.9% 1990 16 4 25.0% TOTAL 630 281 44.6% COMPLETED DEGREE PERCENT YEAR OF NUMBER NUMBER RETURNED RELEASE RELEASED RETURNED (RETURN RATE)* 1987 116 43 37.1% 1988 154 38 24.7% 1989 68 12 17.6% 1990 18 1 5.6% TOTAL 356 94 26.4% *PERCENT RETURNED = NUMBER RETURNED / NUMBER RELEASED
TABLE 5 ESTIMATION OF THE NUMBER OF RETURNS EXPECTED IF MALE COLLEGE PROGRAM PARTICIPANTS WERE SELECTED FROM THE GENERAL MALE RELEASE POPULATION WITHDRAWAL OR ADMINISTRATIVE REMOVAL RATE OF ESTIMATED YEAR OF NUMBER RETURN FOR NUMBER TO BE RELEASE RELEASED ALL MALE RETURNED BY RELEASES 2/28/91 1987 218 X 47.0% = 102 1988 250 X 43.9% = 110 1989 146 X 31.4% = 46 1990 16 X 18.1% = 3 TOTAL 630 X 41.0% = 261 COMPLETED DEGREE RATE OF ESTIMATED YEAR OF NUMBER RETURN FOR NUMBER TO BE RELEASE RELEASED ALL MALE RETURNED BY RELEASES 2/28/91 1987 116 x 47.0% = 54 1988 154 x 43.9% = 68 1989 68 x 31.4% = 21 1990 18 x 18.1% = 3 TOTAL 356 x 41.0% = 146
TABLE 6 COMPARISON OF EXPECTED AND ACTUAL RETURNS TO DEPARTMENT CUSTODY FOR MALE COLLEGE PROGRAM PARTICIPANTS PROGRAM PARTICIPATION WITHDRAWAL OR ADMINISTRATIVE COMPLETED REMOVAL DEGREE ESTIMATED: NUMBER: 261 146 RETURN RATE: 41.0% 41.0% ACTUAL: NUMBER: 281 94 RETURN RATE: 44.6% 26.4%
TABLE 7 NUMBER OF MALE COLLEGE PROGRAM PARTICIPANTS RETURNED TO DOCS CUSTODY DURING THE FOLLOW-UP PERIOD BY TYPE OF PARTICIPATION WITHDRAWAL OR ADINISTRATIVE COMPLETED REMOVAL DEGREE TOTAL NOT RETURNED DURING FOLLOW-UP 349 262 611 PERIOD RETURNED DURING FOLLOW-UP 281 94 375 PERIOD TOTAL 630 356 986 x2 = 31.96 p= .001
Note: The chi square test statistic was used to test the hypothesis that there is no relationship between completing/not completing a college degree and being returned to the Department's custody. The chi square value obtained was 31.96 and is statistically significant at the .001 level. Therefore, the hypothesis that there is no relationship between completing/not completing a college degree and being returned to the Department's custody is rejected.
APPENDIX A COLLEGES PARTICIPATING IN THE DEPARTMENT SURVEY BRONX COMMUNITY COLLEGE CAYUGA COM14UNITY COLLEGE CLINTON COMMUNITY COLLEGE COLUMBIA-GREENE COMMUNITY COLLEGE CONSORTIUM OF THE NIAGARA FRONTIER CORNING COMMUNITY COLLEGE DUTCHESS COMMUNITY COLLEGE GENESEE COMMUNITY COLLEGE JEFFERSON COMMUNITY COLLEGE MARIST COLLEGE MATER DEI COLLEGE MEDAILLE COLLEGE MERCY COLLEGE MOHAWK VALLEY COMMUNITY COLLEGE NORTH COUNTRY COMMUNITY COLLEGE SAGE -JUNIOR COLLEGE OF ALBANY SKIDMORE COLLEGE SULLIVAN COUNTY COMMUNITY COLLEGE S.U.N.Y. COLLEGE AT NEW PALTZ SYRACUSE UNIVERSITY ULSTER COUNTY COMMUNITY COLLEGE