! ||UIIH| — — — — — ,——__ — — — _ — — — FACTORS INFLUENCING ESCAPE FROM CORRECTIONAL INSTITUTIONS ' ”4-" ‘ - - ‘ o-ow”~_ov ~fiu-““p~-.‘ 00w. -- —L ‘ M‘ I“-"‘- “-"'O v.‘ ‘w. - ‘ --A-o-— Mummnmutmmnnnn”Human ' 293 01107 0962 MSU LIBRARIES m MAGIC 2 MAY 2 51939; RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES wiII be charged if book is returned after the date stamped below. FACTORS INFLUENCING ESCAPE FROM CORRECTIONAL INSTITUTIONS AN ABSTRACT SUBMITTED IN PARTIAL FULFILLMENT FOR MASTERS OF SOCIAL WORK DEGREE MICHIGAN STATE UNIVERSITY SOCIAL WORK 895 CAROL LINGENFELTER DR. ANDREW MICHAEL ORDING JUNE, 1969 MARY NASH 11-41mm ERW‘L lurfia “a”; «I e. F.” fifinf’q‘ ' " 'u g" '1‘ §i_ ':“":‘% wc’A :Efia 'N‘l‘ {fit-'7 it”, . ”.on H ‘5" :.'-Y3ld :2, ABSTRACT This was a preliminary, exploratory study designed to differentiate between escapees and non-escapees from correc- tional facilities. The study resulted in a prediction table which indicated the combinations of factors relating to both high and low rates of escape. The configuration for high escapism included prior adult commitment, relatively long time until parole hearing and IQ over 80. The configuration for low escapism included no prior adult commitment, relatively short time until pa- role hearing, and absence of crisis within the prison. With the completion of a validation study, these re- sults will be used in estimating the escape risk of candi- dates for minimum security placement. FACTORS INFLUENCING ESCAPE FROM CORRECTIONAL INSTITUTIONS A RESEARCH REPORT SUBMITTED IN PARTIAL FULFILLMENT FOR MASTERS OF SOCIAL WORK DEGREE MICHIGAN STATE. UNIVERSITY SOCIAL WORK 895 CAROL LINGENFELTER DR. ANDREN BJCHAEL ORDIN JUNE, 1969 MARY NASH ACKNOWLEDGEMENTS We wish to thank the personnel of the Bureau of Pro- gram and Planning of the Michigan State Department of Cor- rections for their assistance and cooperation in doing this study. William Kime, Acting Director of the Bureau, super- vised us in gathering the data and acted as a consultant in writing the study. William Madtes, formerly of the Research Division, assisted us in choosing our subjects for the study. Mr. Kime and Mr. Madtes gave us a general orientation to the Michigan prison system. We also wish to thank the administration and staff of the State Prison of Southern Michigan, located at Jackson, for assisting us in gathering data from the prison files. TABLE OF CONTENTS IntroduCtj-on 0.0.0.0....O00......0.00.0.0...OOOOOOOOCOCCOO 1 Method SUbjeCtS O00.0.0...0.00.00.00.0000.0.0.0...OOOOOOOCOCOCO 7 Apparatus'.OOOOOOOOOOCOOOOOO.O.OOOOOOOOOOOCOOOOOOOOOOCC 7 Procedure‘...’OCOO...OOOOOOOOOOOOOOOOOOOOOOO00.0.0000... 11 Results.O...OOOOOOOOOCOOCCOOOOOOCOOOOOOOOOOOOO0.0.0.000... 13 Table 1 0.0.00.0...0.0.00000000000000000COOOOOO00...... 16 DiSCUSSiOn 000000000o00000000000000.00000000000000.0000... 17 Appendix INTRODUCTION "Many years ago it was thought that the criminal is predestined to his criminal career on account of his in- herited traits and that he was therefore wholly irrespon- sible".l At that time, no attempt was made to study such an individual in any specific manner. Later, the theory of Lombroso became popular, and it was felt that there were certain physical anomalies which were characteristic only of the "criminal". However, no credible studies have shown conclusively that there is any correlation between physique and character. Therefore, it has been necessary to estab- lish new methods of studying the criminal (for the purposes of this study, "criminal" is an individual who has been in- carcerated because of some crime against society). Many studies of the incarcerated individual have been carried out. However, few which have been done have had the purpose of the present study, ie. to provide a practical guide to the prison classification personnel who assign prisoners to their custody levels (maximum, medium, or mini- mum). The need for such a guide has become obvious over the past years. In iichigan alone, the annual number of escapes has risen from 113 in 1962 to 179 in 1967; making an average l. Sol Levy, M.D., et.al., "The Outstanding Personality Factors Among the Population of a State Penitentiary", pp. 269- 276; drawn from the files of the Michigan Department of Cor- rections, Lansing, Michigan. -2- of 1h9 annual escapes over the last six years, from all Michigan Department of Corrections facilities, excluding Detroit House of Correction and Pre-Release Guidance Cen- ter. These increasing escapes could, and do, have very serious consequences. For example, public fear and appre- hension regarding escapes could endanger further "minimum custody" programs within the Michigan prison programs, which are needed because of the current overcrowded conditions of all Michigan prisons. In addition, "escape" itself is a crime in Michigan and is therefore detrimental to the welfare of the prisoner, because he must then serve an additional prison term (very few escapes are successful). It is unfortunate, but as David I. Morgan stated, "one of the most painful factors with which the inmate has to contend, is his loss of freedom".2 This includes both confinement in prison away from society and confinement within prison in a cell. Such confinement symbolizes most clearly the individual's rejection by society at large. However, individuals, because of personal and situational factors, respond and adjust differently to this confinement, and also to any reduction in this confinement, such as would be found in the minimum custody facilities. "The imprisoned criminal must somehow find constructive means 2. David I. Morgan, "Individual and Situational Factors Related to Prison Escape", American Journal of Correction, larch-April, 1967, PP.30. for handling these emotions if he is to adjust and endure psychologically. An alternative is escape".3 The problem thus becomes one of isolating those personal and situational factors which influence adjustment to confinement and reduced confinement. In this way, it would become possible to pro- vide classification officials with a guide to placing indi- viduals in minimum custody - which would be in the best interests of both the individual and society. "Once pri- son officials were alerted to the pattern, they could try to eliminate the possible causes for it, and prevent future escapes".h A few studies have begun to examine this problem. One such study, carried out by Levy, et.al., was concerned with the practical application of the MMPI (Minnesota Multiphasic Personality Inventory). This test is a "psychometric instru- ment designed ultimately to provide in a simple form, scores on all the more important phases of personality".5 The study itself was an attempt to establish outstanding person- ality factors among the populations of penal institutions. However, certain findings relate directly to the subject of escapism. The authors, after giving the test to 300 inmates at Washington State Penitentiary, divided the group into 3. Ibid., p. 32. h. Professor Albert Morris, Boston University Press Release; future publication, June 22, 1962, pp. 2 5. Levy, op. cit., p. 270. nonescapees (270) and escapees (30). They found that the escapees had significantly higher scores on the Pa (paranoid), Sc (schizophrenic), and Ma (hypomanic) scales. They con- cluded that escapes may be based on "projection", and an attempt to get even with society. In addition, the Ma score showed that the escapee tends to be more restless, hyper- active, and unable to settle down in any environment. This study began the exploration of the problem, but in a limited sense. Several other studies have brought out additional information concerning both individual and situational factors related to escape. Morgan isolated eight factors which showed significant differences between escapees and nonescapees. These factors were the following. Escapees: (1) had a sentence of five years or less, (2) were single, (3) had no dependents, (A) had at least a secondary education, (5) were 2h.5 years old or less, (6) had an 1.0. of 96.5 or more, (7) were first offenders, and (8) had served less than 1/2 of their present sentence. W.S. Loving et.al., did a similar study in Louisiana. These authors also isolated eight variables which were significantly higher for escapees. These were: 1. no dependents 2. one or more commitments elsewhere 3. property offense A. one or more juvenile commitments 5. young age at first arrest 6. residence in a town of less than 100,000 7. over 150 miles to home from prison 8. less than two years residence in Lousiana The authors then did additional statistical testing which yielded two clusters of four variables each. These clus- ters were "transient criminality", including factors 2, 6,7,8 and "early criminal history" (factors l,3,h,5). One additional study on escapism was performed by Professor Albert Morris. He found, as in the above.study, that "most of the escapers were those who had committed a prOperty offense".6 However, Morris felt that this might be because many "person" offenders are usually kept under maximum security. Professor Morris also found that the escapee "was apt to be younger than the non-absconder, less likely to be married, usually in an unskilled occupation, and had usually left school at the earliest permissible age despite an adequate intelligence".7 The escapee had some further outstanding characteristics. He had been subjected to needed disciplinary action, was withdrawn and uncoopera- tive, emotionally immature, and likely to escape within five months of sentencing. 6. Morris, op. cit., p. l. 7. Ibid, p. 1. -6- The above studies isolate certain general character- istics which can be attributed to escapees from penal insti- tutions. However, as with the eighteen factors now used by the Michigan Outside Placement Director, these factors are not weighted by their comparative importance. That is, it has not yet been established scientifically which of all the above-mentioned factors differentiate most clearly be- tween escapees and nonescapees. To provide classification officials with a practical guide, it is important to deline- ate such factors, since all prisoners will probably have one or more of the mentioned characteristics. Therefore, it is the purpose of this study to (1) isolate the factors which differentiate between escapees and nonescapees, (2) establish a "priority scale" for these factors, and (3) show interrelationships between the significant factors. METHOD Subjects: A group of 1th inmates at Ionia Reformatory, (young adult offenders age 17-23) placed in a minimum security dormitory, were selected for this study. Fifty-two of these subjects had escaped from the dormitory during a four year period between August, 1963 and September, 1967. Each of the escapees was matched with two non-escapees with respect to race, date of reception at the dormitory, and work assignment. Except for the matching variables the non-escapees were cho- sen at random. Apparatus: The selection of variables with which to compare escapism with non-escapism was made within the constraint of information available regarding the subjects. The intent to set up a prediction table of factors which significantly con- cur with escape behavior, made it desirable to compile the most comprehensive list of variables possible from the information available. Seventy-five factors were selected which could be coded from information given in the case file of each of the subjects. Outline of Factors I. Personal description of subjects A. Age at Reception Diagnostic Center admittance B. Height II. (31 U O o e e 0:3 I QHFQ L. -8- Significant physical defect present Tatoo Intelligence quotient Minnesota Multiphasic Personality Inventory profile-- normal or abnormal Hiatory of mental illness Impulsiveness (social worker's evaluation) Immaturity (social worker's evaluation) Prognosis for community adjustment (social worker's evaluation) Cooperation during interview (social worker's evaluation) Remorse for crime Personal Background A. Q '12) F?) C o 0 Educational level 1. A. G. R. level 2. School years completed 3. Age at which subject left school Religious preference Work background 1. Number of jobs in three years prior to first adult commitment 2. Average weekly income Geographic mobility Alcoholism and/or drug addiction Military history AWOL type behavior in the past--escape from Boys' Training School, Army, probation camp, etc. III. IV. Family History A. B C D E :x c. #4 :2 c) *a 0 Marital status Number of dependents Occurence of twins in family Marital status of parents Age difference between parent or parent-surrogate and subject Arrest history of family members Age when subject left home for more than six months Communications with family while in prison Distance from home to prison Urban or rural residence Years of residence in Michigan Criminal record A. B. Type of offense--person, property or other Sentence status 1. 2. 3. h. Length of sentence--minimum Percent of term served at time of placement in dormitory Amount of time to parole hearing at time of place- ment in dormitory Detainers on file Prior record 1. 2. 3. Adult commitments in Michigan Adult commitments elsewhere Juvenile commitments E. -10- A. Age at first arrest 5. Prior escapes and attempts to escape 6. Number of contacts with police on rap sheet 7. Age at first adult commitment Parole history 1. Time from parole to parole violation 2. Type of violation Premeditated or impulsive offense Prison Record ZZHNQHFQ’UWUOwtb O 0 Type of work assignment during placement in dormitory Request for minimum security placement. Academic training in prison Academic improvement Vocational training in prison Completion of vocational training Number of work assignment changes Number of prison changes Number of disciplinary reports Attitude toward staff--good, fair, or poor Relationship with other inmates--good, fair, or poor Number of letters to prison authorities Denial of parole within last year Recent crises within prison during last year (homosexual pressure, sudden change in work assignments, severe disciplinary measures, etc.) Recent crisis outside prison during last year (divorce, death of family member, etc.) -11.. P. Location from which escape occurred. Q. Time of day escape occurred R. Escaped alone or with companion. Only those factors are listed which were available in the case file at the time of the subject's placement in the minimum custody dormitory. Procedure: The folders of all subjects were read by the re- searchers to determine the individual's status on each of the seventy-five factors. Two researchers made independent ratings of the factors for each file and any discrepancies were resolved mutually by all the researchers. It was decided that the study would be most useful if it could be used to predict the characteristics and combinations of characteristics which indicate whether a person is escape- prone. Therefore the data was organized into a configuration table, patterned8on the one used by Glaser in predicting post- release success. The first step was identification of the one factor which most significantly differentiated escapees from non-escapees. This was done by comparing the incidence of each of the seventy- five factors for those who escaped with the incidence within 8. Daniel Glaser, "Prediction Tables as Accounting Devices," Crime and Delinquency, Vol. 8, No. 3, July, 1962, pp. 250-2550 -12- the total group. The resulting percentage gave a rate of escapism for those with this factor. Following Glaser's method, only those variables were considered which discriminated some group with an escape rate exceeding limits set arbitrarily at l/3 the distance between the escape rate for the total sample and either 0% or 100%. Within our sample of 14A, 52 were escapees, resulting in a percentage of 36.16%. Thus in our first division we looked for factors which occurred in less than 24.11% of the cases (one-third the distance from 36.16% to 0%), or in more than 57.hh% of the cases (one-third the distance from 36.16% to 100%). The selection among variables meeting this criteria was based on maximizing the number of cases which this factor included. We also considered the ease of operationalizing the the factor. -13- RESULTS The most significant factor in the first division was prior adult institutional commitment. Thirty-two of the cases (22.2% of the entire sample) had a history of prior adult com- mitment. Sixty-two percent of this group escaped. This ex- ceeded the 57.AA% set as a criterion. The other 112 cases with no history of adult commitment showed a 28.57% rate of escape. Each of these two groups was then recounted on all varia- bles as to the incidence of the other factors. The arbitrary selection criteria for this second level were set at below 16.07% (one-third the distance from 2A.ll% to 0%) or above 71.63% (one-third the distance from 57.h4% to 100%). Again the choice of factors was also influenced by the number of cases which were discriminated. Within the group with prior commitments, the factor most significant in discriminating between escapees and non-escapees was the length of time to parole hearing. Those with over six months to their hearing had a 78.06% escape rate, whereas those with under six months showed a 22.02% escape rate. This latter group was not further subdivided. Those with over six months to their hearing were again counted on all remaining variables to determine if another factor could differentiate among them as to escape probability. -14- On this third level, the levels of significance were set at below 10.71% (one-third the distance from 16.07% to 0%) and over 81.09% (one-third the distance from 71.63% to 100%). The factor breakdown which showed the highest escape rate and included the most number of people was intelligence quotient. There were 13 cases with an IQ over 80 who had an escape rate of 86.67%. Those with an IQ under 80 had an escape rate of 62.50%. The distribution of IQ within the prison population is somewhat skewed. Therefore there are a significant percen- tage who fall into the category below 80, and this is a mean- ingful division. The other side of the table now developing (Table l), i.e. those with no prior adult commitments, was also analyzed in the same way. On the second level, the factor which differen- tiated the best was again length of time to parole hearing. Those with less than three months to parole hearing had an escape probability of 1h.10%. Those with over three months had an escape rate of 35.05%. The group with less than three months to hearing was further subdivided. At this third level, the most significant factor was whether there had been a crisis in prison within the last year. Those who had no record of a crisis had an escape probability of 3.h5%. Those who had a record of crisis had a 66.67% rate of escape. 13-1- -15- This completed the analysis of the data, and the result was Table l which can be used, subject to validation, to place individuals in escape risk categories. abwbm H noszmdwbaHOZ prbm wow mmobwmmm 926 zozlmmnbwmmm mocmmb GZUWW ZHZHZGZ mWodWHH& PH Hmm ZHOIHGDZ wmwow2>aow& PH HOZHD mswou Pthn noaawfiamnd w W No mm m 5 $me 20 wswos bthd ooBBHflamnw w W m H mm 5 HHm mm.mux ommd I‘VI- ‘I I —.». . .L in Hr. - ' '- 30 H. .L m I? l‘ ’5 I' .1 0 Mil f... - _—~/...‘.. ...f.'.J-T.b -l7- DISCUSSION These results show that factors influencing escapism are prior adult commitment, length of sentence to be served, intelligence, and crises in prison. However, before these factors can be used as indicators of escape risk, they should be validated in further studies, first among a sample from the same correctional institution, and then among a random sample drawn from general prison population placed in minimum security. This study cannot be directly compared to previous stu- dies on escapism because of the method of analysis which was employed. The prediction table was used in this case because simple correlation studies have the weakness of isolating fac- tors without indicating the relative influence they have in determining escape risk, or the way in which they interact. This study was not intended to analyze the reasons why a particular factor is relevant to escapism. However, we would like to suggest that escapism results from frustration because of repeated incarcerations coupled with a long waiting period until parole. We feel that the person with high intel- ligence is more susceptible to frustration and less able to adjust to the prison regime. Crises inside the prison appar- ently increase this frustration. APPENDIX ALL SIGNIFICANT FACTORS * A. First Level (18 factors) 1. 2 3: a. 5. 6. Prior adult commitments (62.16% — 20/32) Three or more juvenile commitments (66.67% - 6/9) One or two contacts with police (22.10% ~10/A5) Age at first adult commitment a. 17 years or less (57.2% - 23/L0) b. 20 years or more (21% - 7/32) Length of sentence a. 12 months or less (22.h% - h/18) b. A years or more (70% - 7/10) Time to parole hearing at dorm placement a. less than 6 months (18.3% - 12/65) b. more than 12 months (57.h% - 16/28) Percent of term served at placement ' a. over 80% served (21.h5% - 12/55) IQ (less than 70) (7.9% - l/lO) AGR, less than A.O (21% - 5/23) Age difference between father and child a. less than 22 years (14% - A/27) b. less than 25 years (21% - lO/A?) AWOL behavior; more than one (83% - 10/12) Age difference between mother and child a. over 32 years (57% - 11/19) Not rated impulsive (1A% - 5/35) Not rated immature (16% - 4/32) Academic improvement in prison (22% - lO/h5) Assignment changes; over 6 (61% - 8/13) Recent crisis in prison (67% - 25/37) Recent crisis outside prison (69% - 16/23) B. Second Level - Left Side (7 factors) 1. 2. 3. A. Over one juvenile commitment (87.5% - 7/8) Four or more contacts with police (80% - 17/21) Sentence of over 2 years (13.3% - 2/15) AWOL behavior (73.5% - L/11) * All othhese factors were significant but some of them were not used in the configuration table because of the small number of cases or the difficulty in operationalizing them. 5. 6. 7. C. Third I. 2. 3. h. 5. Age difference between mother and child a. over 23 years (76% - 4/13) Crisis outside prison (85.7% - 6/7) Over six months to parole hearing (78% - 18/23) Level - Left Side (5 factors) IQ over 80 (80% - 13/15) One or more juvenile commitments (92% - 12/13) Height over 5'9" (92% - 12/13) AGR over 6.1 (91% - 11/12) ANOL behavior (81% - 9/11) D. Second Level - Right Side (3 factors) 3: Less than 3 months to hearin (14% - 5/35) Not rated impulsive (13% - 4 30) Not rated immature (7% - 2/26) Level - Right Side (4 factors) No crisis within prison (3% - 1/29) One or less assignment change (0% - 0/15) Vocational training (7.6% - 1/13) Over two year sentence (10% - l/lO) BIBLIOGRAPHY Glaser, Daniel, "Prediction Tables as Accounting Devices," Crime and Delincuency, Vol. 8, No. 3, July, 1962, pp. 250-255. Levy, Sol, M. D. R. H. Southcombe, M. D., John R. Cranor, and R. A. Freeman, "The Outstanding Personality Factors Among the Population of a State Penitentiary," pp. 269-276; drawn from the files of the Michigan Department of Corrections, Lansing, Michigan. Loving, fl. 3., R. E. Stockwell, and D. A. Dobbins, "Factors Associated With Escape Behavior of Prison Inmates " Federal Probation, Vol. 23, Sept., 1949, pp. h9-5 . iorgan, David 1. "Individual and Situational Factors Related to Prison Escape," American Journal gf Correction, March-April, 1967, pp. 30-32. Morris, Professor Albert, Boston University Press Release; future publication, June 22, 1962, pp. 1-3. ‘7‘" ‘a . . .1 I v . ‘I .l . 4 lax. L _ h w I . . r I . . I Q . . ... - . . In \1 U. I Y. r" ) 1 x p I. I: x. r. . I o \ \ ol. ‘ . ‘ w I _ 1. .- ’- ’ - U m. .1 . . .I 1; l ’ Is. I. r r. \ ‘ . I e 113MB? mchigsn €5th '- University 355 @65 DD? 3P5 BUS 3Y5 ACCO‘9PRESS' S 8507 2507 2507 2507 250‘! 2507 8507 COHEN 33 l- AJ L Y El. LUV." fire-USO I I\'(;;; 6"“.er &"r‘ “it 7- L‘Afif". $5.:FN : / ‘i.l'~f-'G'I¢.*Nt 3E»; vac-'7 scum. ELL)! Mint-'C‘WTED beam-Aw O.- ‘\