Academically Speaking... Criminal Justice Research by Florida's Doctoral Candidates 1991
Diane 1. Zahm, Ph.D. Editor
j\lCJRS
SEP 16 1994
ACQUISITIONS
150081- U.S. Department of J~'':Jtlce 150085 National Institute of Justice
This document has been reproduced exactly as received from the p~rson or organization originating it. Points of view or opinions stated in this document are those of the authors and do not necessarily represent the official position or policies of the National Institute of Justice. Permission to reproduce this -....-riii!ll; ~TI~~ic Domain/BJS U.s. Department of to the National Criminal Justice Reference Service (NCJRS). Further reproduction outside of the NCJRS system requires permission of the ~ owner.
Florida Statistical Analysis Center Florida Criminal Justice Executive Institute Florida Department of Law Enforcement P.O. Box 1489 Tallahassee, Florida 32302 (904)487-4808
This document was produced under cooperative agreement 89-BJ-CX-K020 with the United States Department of Justice, Bureau of jtlstice Statistics. FORWARD II II
When the Florida Legislature crea~ed the Florida Criminal Justice Executive Institute in 1990, it also encouraged on-going cooperation between the Institute and Florida's community colleges and unive.rsities. Our relationship is still in its formative stage, and we are working to define the most appropriate role the State University System can and will play at FCJEI. Academically Speaking .. .is the first in what we hope will be a long line of cooperative ventures between the State University System arId the Florida Criminal Justice Executive Institute. Academically Speaking .. .is an important document for several reasons. It is a clear demonstration of our return on investment in higher education in Flmida. It offers recent doctoral candidates a unique opportunity to publish, because the results of their hard work will be read by such a broad audience. Publication is an essential part of their professional development. Through this compendium, we hope you will understand the contributions they already have made to criminal justice in Florida, and the contributions they will make to criminal justice here and elsewhere in the future. One of the most valuable things Academically Speaking ... will do is to highlight the. vpportunity we have to advance one another's point of view. Each of us considers our work important. Those of you in law enforcement, corrections, prosecution and other areas make necessary and important contributions to the criminal justice system. Your peers recognize the value of your work. Those of us in academic institutions place great value and emphasis on research and evaluation. Our work provides data and information which other academics indicate is meaningful, but which also has great value for practitioners. Too often each of us fails to understand the other's contribution -- we do not recognize the degree to which research and practice overlap, or how much one supports and directs the other. Fortunately, in Florida we have a unique opportunity to cultivate this overlap through the Florida Criminal Justice Executive Institute. FCJEI was created to fill Florida's need for an innovative, multifaceted approach to the professional development of criminal justice executives. FCJEI carries out its mission through an integrated program of education, training and research. While the staff of FCJEI might carry out this mission alone, they have elected to take advantage of the knowledge and skills available at Florida's colleges and universities. This relationship is not new, nor is it unique. Florida's community colleges historically have participated in criminal justice training and career development as the State's regional training centers. Several of them are actively involved in the development of new or experimental programs in criminal justice education. Santa Fe Community College, for example, is working with IDLE and the Criminal Justice Standards and Training Commission on the Enhanced Criminal Justice Training System. The Allstate Center at St. Petersburg Junior College is an outstanding example of the working relationship between an educational institution and the
i business and criminal justice communities. Our State University System offers criminology, criminal justice and related programs which form the basis of many careers in Florida. Unlike the training centers, however, our universities may not come to mind when considering alternatives for "continuing" education in the criminal justice field. Our work with the Florida Criminal Justice Executive Institute has the potential to clarify that role. Already, SUS faculty and staff are serving as instructors at FC]EI. Publications like Academically Speaking... and other similar projects afford us the opportunity to become more involved in professional development. Increased interaction between practitioners and academics will help us to understand one another's views. Then, together, we can work to improve criminal justice in Florida.
Charles B. Reed Chancellor, State University System of Florida
ii Academically Speaking... Criminal Justice-Related Research by Florida's Doctoral Candidates 1991
TABLE OF CONTENTS
Forward ...... , ...... 1 Charles B. Reed, Chancellor State UniversitJ) System of Florida
Table of Contents ...... iii
Introduction ...... 1 James D. Se-wel/, Ph.D., Director, Florida Criminal Justice Executive Institute
The Effect of Execution on Homicides: Florida, 1979-1987 .. .I ~~Q()S~ ...... , 3 Carol Bullock, Ph.D. Florida State, Crimin%ll)
Geographic Spillover Effects of Drug Law Enforcement: Evidence and Implications for Local Government in Florida ...... 16 David Sollars, Ph.D. Florida State, Economics
Moral Panics and the Vocabularies of Motives: Content Analysis of the Cyclical Nature of the Drug Crisis, 1970-1985 ...... 33 Myrna Cintron, Ph.D. Florida State, Criminology
Correlates of Police and Correctional Officer Performance .... 1.5;t:?~fJ.~. . . .. 44 Tracy Griffith, Ph.D. Florida State, Counseling Psycholall)
iii An Analysis of Differences in Role Perceptions Among Senior Patrol Officers and University Judicial Officers in Selected Public Higher Education Institutions in the United States ...... 62 Diane Waryold, Ph. D. Florida State, Higher Education
An AnalYSIS. of Factors Influencmg., ProbatIOn Outcome. . . . . 15""00...... 8-bf ...... 69 Km) Morgan, Ph.D. Florida State, Criminology
Did I Cause This and Can I Change It: The Role of Perceived Responsibility in Treatment of Adolescent Offenders ...... 85 Cynthia Swenson, Ph.D. Florida State, Psyc1lOlogt)
DUI Offenders' Perceptions of Recovering and Nonalcoholic Counselors ...... 98 Shirley Spooner, Ph.D. University of Florida, Agency, Correctional and Developmental Counseling
The Fight for Victim's Rights in Florida: The Realities of Social Reform I5t?P8?ill Susan Sayles, Ph.D. Florida State University, Criminology
Appendices A. :M1vIPI x IPI Pearson Bivariate Correlation Matrix B. Criminal Justice-Related Doctoral Research in Florida, 1991
IV INTRODUCTION II II
Academically Speaking ... 1991 is the first in an annual series planned by the Florida Criminal Justice Executive Institute (FCJEI). This volume contains nine articles on criminal justice-related topics, each based on the dissertation research of a doctoral candidate from Florida's State University System. These works were selected after a review of dissertation titles provided by state universities offering advanced degrees. Each candidate with a criminal justice-related title was contacted regarding possible participation in the compendium, and alJ positive replies were accepted for publication. The final product spans two universities, six disciplines and a wide range of topics. Here, in one place, is current research on homicide, drugs, and officer selection and performance; on juvenile and adult offenders; and on victims. Each article also includes a short biographical statement provided by the author. They are as varied in their careers as they are in their contributions. For example, the news media has offered extensive coverage regarding new limitations the United State Supreme Court has placed on habeas corpus petitions, particularly since several offenders have been scheduled for execution in recent months. A more critical issue for law enforcement may be whether ex~cutions have any impact on homicide rates, something not discussed in the news. Carol Bullock provides just such an examination here. Since 1987 the United States has spent billions of dollars in its war against drugs. David Sollars and Myrna Cintron discuss two unusual aspects of the drug problem: the geographic spillovers resulting from drug enforcement, and the cycle of "crisisll we have experienced when combatting drugs. Tracy Griffith adds an article of interest to anyone reviewing pre-employment psychological screening information. "Correlates of Police and Correctional Officer Performance" describes the testing instruments often employed, as well as possible strengths and weaknesses. His paper is followed by Diane Waryold's evaluation of the role of campus police/law enforcement as perceived by university law enforcement officers. The success of probation, counseling or other rehabilitation programs is a function of a number of factors. Three articles, by Kay Morgan, Cynthia Swenson and Shirley Spooner, discuss these factors and their outcomes. Clearly, what works and what doesn't are essential questions if we expect to be successful in our work. Academically Speaking... closes with an historical review of the creation of the Florida Network of Victim Witness Services, written by Susan Sayles. Florida is a model for other states when it comes to the rights of victims. Becoming a model, however, was not as easy as one might expect. It is our hope that Academically Speaking ... will demonstrate the value that university research has for the broader criminal justice community. FCJEI was created to provide an integrated program of training, education and research. One of our
1 goals, especially given our statutory affiliation with the State University System, is to integrate the academic and the professional community through a broad range of publications. This compendium serves as one of the vehicles to this necessary integration.
James D. Sewell, Ph.D. Director Florida Criminal Justice Executive Institute
2 The Effect of Executions on Homicides1 ii Florida, 1979 - 1987
Carol A. Bullock, Ph.D.
Overview homicide data for the state during that period are utilized, as well as a While some death penalty deterrability index dividing the deterrence studies have reported marders in that data set into three varying evidence of an inverse categories: those most likely to relationship between the death penalty decrease ("deterrable"), those most and homicides, suggesting deterrence, likely to increase ("brutalizable"), and (Ehrlich, 1975, 1977; Yunker, 1976, 1982; those most likely to remain unchanged McFarland, 1978; Cloninger, 1979; ("unaffected") by executions. A Phillips, 1980; Stack, 1987), most have number of interrelated hypotheses are reported little or no evidence of tested: deterrence (Schuessler, 1952; Sellin, (1) There is no relationship 1955, 1967a, 1967b; Savitz, 1958; between executions and the total Reckless, 1969; Bowers ai.1d Salem, homicide rate. 1972; Bowers, 1984; King, 1978; (2) There is a negative McFarland, 1978; Peterson and Bailey, relationship between executions and 1988; Bailey and Peterson, 1989). Many the "deterrable" homicide rate. studies have indicated a positive (3) There is a positive relationship, suggestive of relationship between executions and counterdeterrence (Dann, 1935; Graves, the '1Jrutalizable" homicide rate. 1956; Schuessler, 1952; Savitz, 1~58; (4) There is no relationship Sellin, 1967a, 1967b; Samuelson, 1969; between executions and the Forst, 1977, 1983; Bailey, 1974, 1975; "unaffected" homicide rate. King, 1978; Phillips, 1980; Bowers and Included also is a publicity Pierce, 1980; Peterson and Bailey, 1988). measure to test the hypothesis that, (5) Both an interrupted time series TherE' is a negative relationship design and an ARIMA analysis are between the amount of execution used to investigate the effect of publicity and the total homicide rate. executions on homicides in the State of Florida for the years 1979-1.987.2 Daily
This paper is from Dr. Bullock's dissertation, "The Effect of Florida's Post Funrl1ln Executions on Homicide Rates, " Gordon Waldo, Major Professor. Carol Bullock has a bachelor's degree in business, and a master's degree in Slavic and East European Studies/Comparative POliCl) Studies. Her research at the master's level dealt with the criminal justice Sl)stem of Yugoslavia. Current research interests include female crime and delinquency and victimologJ), as well as deterrence. ~======~
3 Effect of Executions Bullock
Variables Table 1 lists names of inmates executed during 1979-87, and the dates of their Executions. "Execution" in the executions, along with notes of those study refers to the date of execution. excluded. Seventeen inmates were executed Deterrability Index. The between 1979 and 1987, but not all are "deterrability index" is based on the included in the study. In several judgments of 144 undergraduate instances one execution occurred students in the School of Criminology within a very short time of another. and Criminal Justice at Florida State That close proximity could contaminate University.3 To develop the index, the data in such a way that it would be each student was given a list of 51 difficult to determine whether changes items from four categories of in frequencies of murders are the result information included on the SHRs of the execution being studied, or a [method used, relationship of offenders cumulative effect of more than one and victims, number of victims and execution in the same time period. For offenders in a murder, circumstances, that reason, this article includes e.g., quarrel, robbery] and asked analyses of only those executions with whether each of the 51 factors would 14-day pre-execution and 14-day post be associated with a type of murder execution periods during which there likely to (a) increase, (b) be unaffected, were no other executions in the state. or (c) decrease as a result of state executions.4
Table 1 Inmates Executed in Florida, 1979 - 1987
1. OS-25-79 John Spenkelink 10. 11-08-84 Timothy Palmes
2. 11-30-83 Robert Sullivan 11. 01-30-85 James Raulerson
3. 01-26-84 Anthony Antone 12. 03-06-85 Johnny Paul Witt
4. 04-05-84 Arthur Goode 13. 05-29-85 Marvin Francois
5. OS-10-84 James Adams 14. 04-15-86 Daniel Thomas .. 6. 06-20-84 Carl Shriner 15. 04-22-86 David Funchess 7. 07-13-84 David Washington.. 16. 05-20-86 Ronald Straight
8. 09-07-84 Ernest Dobbert' 17. 08-28-87 Beauford White
9 . 09-20-84 James Henry.
• These executions were excluded from the study because of other executions within a 14-day period before or after. .. These executions were excluded from the first interrupted time series analysis, but included (with a reduction in days examined) in the analysis looking at the effect of execution publicity on murder.
4 ------~- ~------
Effect of Executions Bullock
A mean for each of the 51 items are analyzed for consistency with the was calculated after assigning a value hypothesis tested as well as for any of 3 to each "increased II response; 2 to other pattern that might emerge. each "unaffected"; and 1 to each Analyses are conducted for total IIdecreased. 1I A computer program homicides, and then for each of the analyzed each homicide in the data set three categories into which the total and, where an element of the homicide homicides have been divided. matched one of the 51 items, the mean Reported results of t-tests measuring value for that item was assigned to the the differences in pre- and post matching element of the homicide. execution homicides are the t-values Because the data set contained up to for the 14-day period. While there is four circumstances, up to four no reason to think the effect would be methods, and up to five relationships specific to the 14-day period, there is for each homicide, a subindex value no theoretical reason to perform the (the mean) for each of the categories test for any other particular time (circumstance, relationship, method, period -- nor did any pattern emerge in victim:perpetrator ratio) was calculated. the analyses that would suggest the An index value (the mean of the choice of an alternate time period.6 subindex values) was thereby Autoregressive Integrated calculated for each homicide.s Moving Averages Analysis. In Publicity. Publicity is a measure addition to an interrupted time series of newspaper space allotted in The design, the data were subjected to an Miami Herald (the state's most widely econometric analysis, the circulated newspaper) and The Autoregressive Integrated Moving Jal1ahassee Democrat (the only daily Averages (ARIMA) design; such newspaper in Florida's capital city) to designs have not been widely used in day-after-execution stories for each criminological research. ThE;! ARIMA execution. Story text, headlines, and structure used in the study was photographs are included in the identified on the basis of the entire measurement, but only information data set (homicide data for the years directly related to the execution 1976-1987), and the impact of studied is measured. On the basis of executions then analyzed by entering these measurements, the executions are their dates into the program? divided into four categories: high McDowall et al. (1988:14-15) publicity, medium high publicity, have identified three "sources of noise" medium low publicity, and low found in social science data, that could publicity. "obscure" a time series intervention component: (1) trends (systematic Data Analyses changes in the process, usually thought of as in an upward or_ a downward Interrupted Time Series. For direction) and drifts (nonsystematic each execution included in the study, changes in the time series process - both daily and cumulative frequencies could shift up and down, whereas
5 Effect of Executions Bullock trends are usually considered to be lowest daily frequency is on day 13 in changes in either an upward or a the post-execution period. While this downward direction); (2) seasonality change appears to be in the direction of (changes in the time series process at deterrence, it should be noted that regular intervals); and (3) random error cumulative frequencies for the post (fluctuations in the data that are execution period, with the exceptions present even if trend and seasonality of those reported on days I, 13, and 14, are "removed"). ARIMA time series exceed those for the same periods prior models help to identify and reduce (or to executions. This would suggest a "whiten") the effects of all these brutalization effect. At the end of the obscuring factors so that the true 14-day period, however, there were impact of an intervention can be two fewer homicides following observed. It is also possible to executions than in the 14-day period determine from the ARIMA analysis prior to executions. Overall, there is whether there is any true significance no effect; the two-tail probability of in any impact of the intervention; this such a finding is 0.937. is not always discernible from simple Frequencies of brutalizable time series analyses. homicides indicated that the executions had no effect on the number of homicides to follow. The probability of Findings 0.783 is not statistically significant. For deterrable homicides, Interrupted Time Series. As although daily frequencies of post indicated earlier, there are eleven execution homicides level off at one executions in the initial analysis of point and then drop to zero at the end murders for a period of 14 days before of the 14 days, the cumulative and 14 days after executions. Table 2 frequencies show that at no point in shows the frequencies of homicides for the 14-day period is there a decline. 14 days prior to and following all While there is essentially no effect, the executions in the analysis, then for direction is suggestive of brutalization, each of the three subfiles. The day not deterrence. Again, the finding is number refers to the number of days' not statistically significant, with a distance from the event. For example, probability of 0.520. Day No.1 in the Before column refers The pattern of unaffected to the day immediately preceding homicides shows a slight deterrent execution; in the After column, the day effect at day I, and consecutively at immediately following. days 3-8. Overall, however, there is no Daily frequencies in the total effect. While there were four fewer section show peak daily homicide homicides in the 14 days following frequencies on days 10-12 prior to executions than during the 14 days execution. At no time in the post before executions, the difference is not execution period did the daily statistically significant, with a frequencies reach these numbers. The probability of 0.872.
6 tr:1 Table 2 ct:: ro Effects of Executions on Homicides .....() Frequencies for a 14-Day Period Prior to and Following Executions of Spenkelink, Sullivan, Antone, Francois, ,..,.,o White, Goode, Adams, Palmes, Raulerson, Witt, Straight ro~ Total" Brutalizable" Deterrablec Unaffededd 8p: I o Before After Before After Before After Before After ~ [J) Day No. f cf f cf f cf f cf f cf f cf f cf f cf
1 25 25 23 3 4 4 2 2 4 4 7 7 15 15 14 14 2 20 45 31 54 2 6 5 7 3 7 5 12 17 32 21 35 3 28 73 27 81 2 8 4 11 4 11 3 15 26 58 20 55 4 30 103 30 111 3 11 1 12 5 16 6 21 25 83 23 78 5 27 130 26 137 5 16 8 20 4 20 3 24 30 113 15 93 -....:J 6 25 155 27 164 3 19 3 23 7 27 7 31 11 124 17 110 7 23 178 24 188 5 24 3 26 2 29 3 34 17 141 18 128 8 25 203 29 217 3 27 5 31 5 34 3 37 16 157 21 149 9 21 224 32 249 7 34 2 33 3 37 3 40 15 172 27 176 10 35 259 31 280 5 39 8 41 0 37 3 43 18 190 20 196 J1 34 293 25 305 4 43 2 43 5 42 4 47 22 212 19 215 12 36 329 30 335 8 51 6 49 2 44 5 52 22 234 19 234
13 25 354 16 351 5 56 2 51 3 47 3 55 15 249 11 245
14 22 376 23 374 4 60 6 57 3 50 0 551 17 266 17 262 • _ ~- _____ L-..-- ___ .__ .. _ ., 21 murders were committed on the days these inmates were executed. h 3 "BmtaJizable" murders were committed on the darls these inmates were exeCllted. c 5 "Deterrable" murders were cOl/lmitted on the days' these inmates were executed. d 13 "Unaffected" murders were committed on the days these inmates were exeClited. tP ~ -o ~ tr1 Table 3 ~ro (') Effects of Execution Publicity on Homicides n- Frequencies for a 14-Day Period Prior to and Following Execution Stories, According to Amount of Publicity ....,O tr:i d d High' Medium Highhe Medium Low Low >
1 3 3 4 4 7 7 9 9 5 5 11 11 7 7 12 12 2 3 6 5 9 4 11 8 17 7 12 8 19 12 29 10 22 3 4 10 1 10 6 ]7 6 23 2 14 8 27 13 32 20 42 4 ] 11 1 11 8 25 7 30 5 19 9 36 23 55 15 57 5 6 17 2 13 8 33 6 36 8 27 10 46 19 74 14 71 6 6 23 4 17 9 42 6 42 15 42 5 51 18 92 13 84 00 7 3 26 4 21 4 46 11 53 5 47 5 56 14 106 15 99 8 2 28 5 26 8 54 10 63 5 52 8 64 11 117 12 111 9 5 33 4 30 5 59 13 76 6 58 4 68 13 130 17 128
10 2 35 . 3 33 12 71 12 88 10 68 3 71 8 138 11 139 11 4 39 5 38 16 154 11 150 12 4 43 3 41 17 171 7 157 13 5 48 6 47 16 187 13 170
14 2 50 5 52 14 201 11 181
• Includes Spenkelink and Sullivan; 3 murders were committed on the days these stories were published. h Includes Antone, Shriner, Adams; 6 murders were cOlllmitted on the days these stories were published. < Includes Goode, Raulerson, Washington; 7 murders were committed on the days these stories were published. d Includes Palmes, White, Francois, Straight, and Witt; 10 murders were committed on the days these stories were pllblishea. • Shriner was added to the analysis to include 1110re executions with Medium High publicity; the time period was necessarily reduced to ten dalJs to avoid contamination OJ effect of other executiolls within the period of analysis. S o r See note (e) regarding the shorter period of analysis resulting from the addition of Washington's e.:recution. -- ~ Effect of Executions Bullock
Table 3 gives the same 0.152). information for the publicity analysis.s One would expect less impact Organization of the table is the same as from Low Publicity executions than that for Table 2. Only two of the from those in the other categories; but executions (Spenkelink and Sullivan) the frequencies show a greater were in the "High Publicity" category. red uction at the end of the 14-day If Hypothesis 5 is correct, there shoulq period than did any of the other be a statistically significant decrease in publicity categories. There were the murders following executions in increases in the cumulative frequencies this analysis. The daily and only at days I, 3, and 4 (a movement cumulative frequencies show similar in the direction of brutalization), while patterns. Cumulative frequencies at the movement was in the direction of days 5 through 13 decrease following deterrence for the remainder of the executions; the cumulative frequencies period. Again, the probability (.326) at days 2 and 3 are equal; and at days shows the finding to be not statistically I, 2, and 14 the after-execution significant. frequencies are greater than those of ARIMA. An ARIMA analysis the same days prior to execution. was applied to each of the data sets by Although the change is somewhat modeling the entire series, then adding consistent with the hypothesis, there is the interventions (as dummy variables) a changing pattern within, and the to that overall structure. Interventions significance of the overall change at the were first added for all the executions, end of the period is 0.889. to determine the overall effect of the If findings were consistent with seventeen, then for each execution. the hypothesis, the second greatest The same pattern of analysis was used amount of deterrence in the publicity for each of the "indexed" data files. analysis would be following executions Results of these analyses are reported with Medium High Publicity. in Table 4, where e is the trend However, all the cumulative parameter and (0 is a measure of the frequencies in that table move in the intervention effect. For example, a (0 direction of brutalization with the value of -0.876 (if it is statistically exception of day 6, when the significant and has a small standard cumulative frequencies are equal. With error) indicates that for that particular a probability of 0.835, the results are intervention the level of the series not statistically significant. drops by 0.876 homicides. The ARIMA Cumulative frequencies of analysis for the total homicide file homicides following executions with shows some of the (0 estimates to be Medium Low Publicity were greater on positive and some to be negative; none each of the ten days following are statistically significant. The tables executions than was the case for the indicate that each of the (0 values is a ten days immediately prior to point estimate. Listed below those executions. The results are not estimates and t-values in the tables is statistically significant (probability = the standard error. In some of the (0
9 Table 4 ~n ARIMA(O,l,l) Analysis of Effects of Executions on Homicides (d.f. =: 623 for each) ..... Q, Total Brutalizable Deterrable Unaffected ~ OO(d) OO(d) ro e e fl 00(") e 00(4) Execution (I.v.duf') n (t-value) (t-value) (t-value) (t-value) (t-value) (t.v."u~) s:: (t-value) 0'. [sLerror] [st.error] [st.error) [st.error) [st.error) [st.error) [st.error) [st. error) g (Jl All .9232 -2.427 .9293 -1.204 .9699 -.8311 .9370 -.5610 (63.76) (-1.72) (67.22) (-1.38) (120.70) (-1.28) (72.04) (-.69) [.0145] (1.4151] [.0138] [.8757) [.0080] [.6513] [.0130] [.8098]
Spenkelink /J227 2.046 .9285 4.123 .9701 -1.945 .9368 2945 (63.48) (.37) (66.71) (1.69) (122.60) (-.75) (71.94) (.09) [.0145J [5.6050] [.0139] [2.4458) [.0079) [2.5921] [.0130] [3.2012]
Sullivan .9236 -7.686 .9294 1.193 .9878 -1.245 .9371 -3.621 (63.96) (-1.37) (67.30) (.48) (410.48) (-.48) (72.11) (-1.13) [.0144) [5.5946] [.0138] [2.4640] [.0024) [2.5832] [.0130] [3.1994]
Antone .9232 -4.152 .9294 1.474 .9699 2.786 .9367 -4.811 I-l o (63.66) (-.74) (67.22) (.60) (433.98) (1.08) (71.89) (-1.50) [.0145) [5.6062] [.0138] (2.4749) [.0023] [2.5810) [.0130] [3.1967)
Goode .9231 -.5207 .9293 5.337 .9698 -5.273 .9368 -.7188 (63.71) (-.09) (67.23) (1.54) (120.62) «?04)* (71.98) (-.22) [.0145j (5.6034) [.0138] [3.4598) [.0080J (2.5M2] [.0130) [3.2010]
Adams .9231 2118 .9295 .9882 .9699 1.830 .9369 -2.773 (63.73) (.04) (67.30) (.29) (121.27) (.71) (72.00) (-.87) [.0145J [5.6028] [67.28] [.01] [.0080) [2.5919) [.0130) [3.1993]
Shriner .9237 -.9073 .9293 -2.164 .9699 1.840 .9368 -.7611 (64.18) (-.16) (67.19) (-.62) (121.29) (.71) (71.97) (-.24) [.0144) [5.6032) [.0138] [3.4661j [.0080] [2.5920) [.0130J [3.2016]
Washington .9231 -8.151 .9291 -5.279 .9878 -2.225 .9584 -.8547 (63.68) (-1.46) (67.07) (-1.53) (409.82) (-.86) (95.79) (-.27) [.0145] [5.5935J [.0139J [3.4598] [.0024] (2.5819J [.0100] [3.1955]
Dobberl .9231 -1.251 .9294 -.1742 .9878 -.2153 .9368 -.8528 (63.72) (-.22) (67.28) (-.05) (412.79) (-.08) (71.98) (-.27) o:f [.0145J [5.6043] [.0138J [3.4662J [.0024J [2.5833] (.0130] [3.2013] ...... E. --- o ~ tr:I Table 4, con't. t:+:ro () Total Brutalizable Deterrable Unaffected .... Q.. 6 wC.) 6 wCa ) 6 wCa ) 6 wCa) Execution (I.v.duc) (I-value) ~ (t-value) (t-value) (t-value) (t-value) (t-value) (t-value) ro [st.error] [st. error] () [st.error] [st.error] [sLerror] [st.error] [st.error] [st.error] ...... J;:: Henry .9230 -5.639 .9294 -1.316 .9700 -.2377 .9368 o -3.986 ::J (63.62) (-1.01) (67.26) (-.30) (121.50) (-.09) (71.94) (-125) fJl [.0145] [5.6008] [.0138] [3.4670] [.0080J [2.5936] [.0130] [3.1973J
Palmes .9226 -3.682 .9293 -.8010 .9698 -3286 .9369 .7870 (63.42) (-.66) (67.23) (-.23) (120.29) (-1.27) (72.01) (.25) [.0145J [5.5995] [.0138) [3.4657J [.0081] [2.5902J [.0130] [3.2017J
Raulerson .9234 7.949 .9297 3.075 .9700 -1.444 .9367 4.9520 (63.86) (1.42) (67.47) (.89) (121.85) (-.06) (71.91) (1.55) [.0145] [5.5961J [.0138] [3.4648) [.0080) [2.5934] [.0130] [3.1955]
Witt .9231 1.116 .9295 1.207 .9700 -.1118 .9368 -.1606 I--l. (63.62) (.20) (67.23) I--l. (.35) (121.85) (-.04) (71.94) (-.05) [.0145] [5.6105] [.0138] [3.4707] [.0080] [2.5934] [.0130] [3.2022]
Francois .9235 2.626 .9294 -1.634 .9699 .9604 .9368 2.887 (64.10) (.47) (67.29) (-.47) (121.47) (.37) (71.98) (.90) [.0144] [5.6024] [.0138] [3.4666] [.0080] [2.5927] [.0130J [3.1991]
Thomas .9231 -.4383 2.9294 -.0038 .9699 -2.231 .9368 1.837 (63.74) . (-.08) (67.28) (.01) (121.31) (-.86) (71.91) (.57) [.0145) [5.6023] [.0138J [3.4663) [.0080] [2.5918] [.0130] [3.2015]
Funchess .9229 -8.812 .9293 -6.161 .9700 -.2067 .9367 -2.330 (63.59) (-1.58) (67.23) (-1.78) (121.66) (-.08) (71.89) (-.73) , [.0145J [5.5911] [.0138J [3.4571J [.0080J [2.5936) [.0130] [3.2006) Straight .9227 -6.169 .9291 -6.258 .9699 -1.246 .9368 .5784 (63.48) (-1.10) (67.09) (-1.81) (121.74) (-.48) (71.99) (.18) [.0145) [5.5978] [.0138) [3.4589) [.0080) [2.5927) [.0130] [3.2017]
White .9238 -4.434 .9294 1.474 .9699 -2.219 .9367 .6985 (64.25) (-.79) (67.22) (.60) (406.52) (-.86) (.7190) (.22) 0:1 [.0144] [5.6077] [.0138] [3.4598] [.0024] [2.5870] [.0130] [3.2083] J;:: - ---_._- --- ...... o C·) The intervention coefficients are point estimates. * p < .05 ~ Effect of Executions Bullock estimates shown in these four tables, increase; and (4) Unaffected, would the standard error is so large that it is neither increase nor decrease following possible the sign of the OJ estimate is executions. The fifth hypothesis was actually the opposite of that shown in that there would be a negative the table, owing to error. relationship between amount of The brutalizable file again publicity and total homicides. resulted in no statistically significant (I) Hypothesis 1 was not disproved estimates and in very large standard with the interrupted time series nor the errors. ARIMA analysis. Once the effects of There is a statistically significant trends, autoregression, and moving
OJ estimate in the deterrable file. The averages were "removed II from the data standard error is very large, but so is with the ARIMA model, the impact of the (I) estimate. It is unclear why this executions was negligible, as evidenced occurred for only one of the executions by the OJ estimates. The interrupted in the file; it is the only statistically time series showed an erratic pattern. significant (I) in any of the files. The While there were clearly peaks and implicati on is that there is an valleys throughout, there appeared to intervention effect at the point of of be no general trena in the homicides Goode's execution that drops the level immediately preceding and following of the series by 5.273 homicides. While executions for this data set. this possibility is not discounted, there Hypothesis 1 was accepted since it was is also a possibility that this is a chance supported by the analysis. occurrence. A test of Hypothesis 2 revealed There are no statistically that the pattern and frequencies of significant (I)'s in the unaffected file. deterrable homicides increased following executions, although not significantly so. Change was actually Conclusions in the direction of brutalization rather than deterrence, but the amount was The findings produced mixed not statistically significant. This results in that they provide support for hypothesis cannot be accepted. some of the hypotheses being tested, ARIMA analysis resulted in one but fail to support some of the other statistically significant (I) estimate, at hypotheses and the main thesis of the the intervention of Goode's execution. study. The main thesis consisted of It is not clear what this large estimate five interrelated research hypotheses. means. All of the (I) estimates for the The first research hypothesis was that Goode execution are quite large. This executions would have neither a intervention effect was also larger thaIl brutalization effect nor a deterrence any of the other (I) estimates in the effect on the total homicide rate. deterrable file, and the large standard Hypotheses 2-4 were that homicides in error arouses suspicion about the point the subfiles (2) Deterrable, would estimate. Hypothesis 2 cannot be decrease; (3) Brutalizable, would accepted, but it must be rejected with
12 Effect of Executions Bullock
qualification because of the (J) estimate pattern of homicides following at the intervention of the Goode executions is that it is an erratic and execution. fluctuating pattern that could NOT be Hypothesis 3 was not supported detected if only weekly (7-day) or other by the interrupted time series analyses. periodic frequencies are reported, as is Rather than an effect of brutalization the case in most death penalty resulting from executions, the pattern deterrence studies. seems to suggest a deterrent effect. This effect is most pronounced at days 5 and 7, but several spikes in the days that follow diminish the magnitude of Endnotes the decrease. Not only is the decrease not significant, but, importantly, the 1. The words "homicide" and "murder" are change in the number of homicides is used synonymously in this report. in the wrong direction to support the 2. Homicide data for 1976-1987 were collected hypothesis. This would appear to from Supplemental Homicide Reports (SHRs) suggest more of a deterrent effect with filed with Florida Department of Law this "brutalizable" data set than for the Enforcement (FDLE) by each law enforcement "deterrable" data set. Hypothesis 3 is agency in the state. The analysis begins with rejected. The evidence presented 1979, the year of the first post-Funnan appears to support the deterrence execution in Florida; SHRs were available for consecutive years only through 1987. (FDLE hypothesis better than it does the estimates that, because of changes in the brutalization hypothesis. reporting system and changeover to a new Hypothesis 4 is accepted; computer system in 1988, some 30% of the homicide frequencies show an erratic total crime data for the state for that year were pattern,and he ARIMA analysis shows not reported to the Department. It should be no effect in the Unaffected file. noted that although 1987 is the last year included in the study, Florida has continued to Frequencies of various levels of execute inmates since that date.) execution publicity, used to test Hypothesis 5, showed that at the end 3. Although potential murderers, not murders, of the 14 day period of High, Medium are deterred, brutalized, or unaffected, it is felt High, and Medium Low, there is that the index is a reasonable proxy measure of indication of a brutalization effect; and those effects. of Medium High, an effect suggestive 4. Modal responses to the 51 judges show only of brutalization. rather than deterrence. nine (18%) of the 51 items associated with The only suggestion of a deterrent brutalization; nineteen (37%) with deterrence. effect is for Low Publicity. Hypothesis Of the remaining 23 items, two tie 5 is rejected. (brutalization/no effect, no effect/ deterrence) In general, this study confirms and 21 (41%) are judged to be unaffected by executions. the findings of the overwhelming majority of other deterrence research: 5. Means, rather than totals, were used executions have no effect on homicides. throughout to avoid assigning low index What the study does reveal about the values to those cases for which less
13 Effect of Executions Bullock infonnation or less relevant infonnation is Bowers, William J. and Glenn Pierce available. 1975 The illusion of deterrence in Isaac Ehrlich's research on capital 6. DBaseIII was used for data entry; SPSS was punishment. Yale Law Journal 85:187- used for t-tests, recoding frequencies, and 208. deterrability indexing. 1980 Deterrence or brutalization: what is the effect of execution? Crime and 7. Both BMDP and RATS statistical packages Delinquency 26:453-84. were used in the AruMA analysis. Bye, Raymond T. 8. Publicity analysis includes executions of 13 1919 Capital Punishment in the United of the 17 inmates. While those additions States. Philadelphia: The Committee of increased the number of executions in the Philanthropic Labor of Philadelphia analysis, they necessitated a reduction in the Yearly Meeting of Friends. numbers of days analyzed, to avoid contamination of the data by other executions Dan!1, R.H. in the same period. 1935 The deterrent effect of capital punishment. Friends Social Services 29:1-20.
REFERENCES Doan, Thomas A. 1988 RATS (Regression Analysis of Time Bailey, William C. Series), Software. Evanston, IL: V AR 1974 Murder and the death penalty. Journal Econometrics, Inc. of Criminal Law, Criminology, and Police Science 65:416-423. Ehrlich, Isaac 1975 Murder and capital punishment: Some 1975 The deter-rent effect of capital further evidence. American Journal of punishment: A question of life and Orthopsychiatry 45:669-688. death. American Economic Review 65:397-4]7. Bailey, William C. and Ruth D. Peterson ] 977 Capital punishment and deterrence: 1987 Police killings and capital punishment: Some further thoughts and additional The post-Furman period. Criminology evidence. Journal of Political Economy 25:1-25. 85:741-88. 1989 Murder and capital punishment: A monthly time-series analysis of Forst, Brian E. execution publicity. American 1977 The deterrent effect of capital Sociological Review 54:722-43. punishment: A cross-state analysis of the 1960s. Minnesota Law Review Blac~ T. and T. Orsagh 61:743-67. 1978 New evidence on the efficacy of 1983 Capital punishment and deterrence: sanctions as a deterrent to homicide. Conflicting evidence? Journal of Social Science Quarterly 58:616-3]. Criminal Law and Criminology 74:927- 42. Bowers, William J. 1974 Executions in America. Lexington, Graves, William F. MA: Lexington Books. 1956 A doctor looks at capital punishment. 1984 Legal Homicide: Death as Punishment Journal of the Loma Linda University in America 1864-1982. Northeastern School of Medicine 10:137-41. University Press.
14 Effect of Executions Bullock
King, David R. Sellin, Thorsten 1978 The brutalization effect: Execution 1955 The Royal Commission on Capital publicity and the incidence of homicide Punishment (1949-1953). Report, Great in South Carolina. Social Forces 57:683- Britain Parliament, 17-24. Papers by 87. Command 8932, H.M. Stationery Office. McDowall, David, Richard McCleary, Errol E. 1967a Homicides in retentionist and Meidinger, and Richard A. Hay, Jr. abolitionist states, in Thorsten Sellin 1980 Interrupted Time Series Analysis. (ed.), Capital Punishment. New York: Beverly Hills, CA: Sage Publications. Harper and Row, 135-38. 1967b The death penalty and police safety" in McFarland, Sam G. Thorsten Sellin (ed.), Capital 1983 Is capital punishment a short-term Punishment. New York: Harper and deterrent to homicide? A study of the Row, 138-54. effect of four recent American executions. Journal of Criminal Law Stack, Steven and Criminology 74:1014-30. 1987 Publicized executions and homicide, 1950-1980. American Sociological Peterson, Ruth D. and William C. Bailey Review 52:532-40. 1988 Murder and capital punishment in the evolving context of the post-Funnan Yunker, James era. Social Forces 66:774-807. 1976 Is the death penalty a deterrent to homicide? Some time-series evidence. Phillips, .David P. Journal of Behavioral Economics 5:1-32. 1980 The deterrent effect of capital 1982 Testing the deterrent effect of capital punishment: New evidence on an old punishment: A reduced form approach. controversy. American Journal of Criminology 19:626-44. Sociology 86:139-48.
Reckless, Walter 1969 The use of the death penalty. Crime and Delinquency 15:43-56.
Samuelson, Glenn W. 1969 Why was capital punishment restored in Delaware? Journal of Criminal Law, Criminology and Police Science 60:148- 51.
Savitz, Leonard D. 1958 A study in capital punishment. Journal of Criminal Law, Criminology, and Police Science 49:338-41.
Schuessler, Karl 1952 The deterrent influence of the death penalty. The Annals of the American Academy 284:54-62.
15 Geographic Spillover Effects of Drug Law Enforcement: Evidence and Implications for Local Government in Florida
David 1. Sollars, Ph.D.
1. Introduction property crime in community j? Questions of this type and models The economics of crime which purport to explain this literature has explored many phenomena ar~ found in the crime theoretical and empirical issues spillover literature. The central involved in modeling the behavior of problem to be examined here is the criminals, victims, law enforcement effect of law enforcement activity agencies, and the response of the aimed at the illicit drug market and public. One area that has received less how this emphasis partially determines emphasis from economists is the illicit the level of property crime. In drug market and how public policy particular, we seek to determine influences this market.1 During the whether the level of police reSources in 1980s, a large share of law enforcement one jurisdiction and that jurisdiction's resources was focused on controlling emphasis on the illicit drug market, the illicit drug trade. This paper affect the level property crime which examines one of the dimensions of the occurs in neighboring jurisdictions. If crackdown in the State of Florida, substantial spillover effects do exist, namely, the interjurisdictional behavior then greater cooperation among of drug market participants and neighboring law enforcement agencies property crime perpetrators. in developing enforcement strategies is Specifically, the focus is on how required. enforcement efforts directed at the Section 2 is a review of the illicit drug market affect the level and relevant economics of crime literature, geographic distribution of property with a particular focus on studies that crime. deal with the crime spillover How do drug and property phenomena. Theoretical considerations crime arrests in community i affect and presentation of the general model
This paper is from Dr. Sollars' dissertation, "Allocation of Law Enforcement Resources, Spillovers, and the Illicit Drug Market, " committee members Bruce Benson (Chair), David Rasmussen, Thomas Zuehlke, Glenn Parker and Laurin Wollan. Partial funding for the dissertation was provided by the Economics and Demographic Research Division, Joint Legislative Management Committee, Florida Legislature, and the Policy Sciences Program, Florida State University. Dr. Sollars is now Assistant Professor of Economics, Auburn University at Montgomery, Montgomenj, AL He holds degrees in economics (B.B.A., 1986; M.A., 1987) from Ohio University and a Ph.D. in economics from the Florida State University (1991). He has published in Public Choice and EQlif1!. Studies Review. His current research interests include the economics of law enforcement and crime, legislature compensation, and other public finance topics.
16 Geographic Spillover Effects Sollars are presented in Section 3. Section 4 data, and Carr-Hill and Stem (1973) includes a discussion of the data to be use similar data from uses used in estimating the empirical municipality data for 53 isolated cities. relationships. Section 5 contains the Phillips and Votey (1975) use California statistical results. The concluding county level data. Thaler (1977), Section (6) summarizes the major Mehay (1977), Mathieson and Passel findings of this study and explores the (1976) use a single metropolitan area implications of the findings. with many precincts or divisions. Buck et al. (1983) utilize a data set of 230 2_ Review of the Literature communities in New Jersey. Witte (1980), Schmidt and Witte (1985), use The economics of crime individual data, and Trumbull (1988) literature is an application of uses both county level and individual micro economic theory to the problem data. of crime. The application of modem Geographic spillovers occur economic theory to criminal behavior is 'when the criminal moves from location usually credited to Becker (1968) and to location in response to increased Ehrlich (1973), and pioneering work crime deterrence activity and / or was also done by Sjoquist (1973), Carr changes in the potential payoff Hill and Stem (1973), and Blocke and sh-ucture in one geographic area Heineke (1975). The literature has relative to another (Reppetto 1976, grown substantially, evidenced by the Hakim and Rengert 1981). The nature numerous literature reviews such as of the spillover problem involves a Cloninger (1975), Chaiken (1976), production externality across Greenberg (1977), Blumstein, et al jurisdictions where different levels of (1978), Palmer (1977), Brier and law enforcement inputs (and therefore Feinberg (1980), and Cameron (1988). outputs) result in an "exporting" of Empirical testing of the crime to other jurisdictions. Spillovers economic model of crime has also are important in models which seek to produced voluminous literature. In measure deterrence effects since a what has become the standard reduction in the crime rate is the goal approach, a multi-equation system is of law enforcement. If the criminal specified (Taylor 1978, Cameron 1988). activity is simply exported to other The first equation specified is an communities, then overall social offenses function, or a crime supply welfare is not enhanced. equation. Other equations include a A stylized model of the spillover probability of arrests and / or conviction process might be represented as function, a punishment severity follows: for simplicity, assume that function, and a police resources only one type of crime exists, for function. Testing has used many types example, burglary, and that there are of data and different levels of two jurisdictions i ana j. Also assume aggregation. For example, Ehrlich that criminal activity can be deterred (1973) uses cross-section state level by law enforcement activity and that
17 Geographic Spillover Effects Sollars criminals are free to move from ito j. crime in the two jurisdictions would be Initially, burglars do not cross achieved. Measurement of the crimes jurisdictional boundaries. To begin the in j committed by those who enter spillover process, let the deterrent from jurisdiction i minus those activity increase (as a result of the criminals who either dropped out of demand and supply interaction in the the market or moved to jurisdiction i market for crime prevention) in one would thus give the "net spillover" jurisdiction relative to the others in effect of an increase in jurisdiction i's order to create a state of police activity. Thus from a disequilibrium. We would then expect measurement point of view, the net to see two phenomena. First, as the spillover effect would give the true cost probability of being caught for imposed on j as the result of policy burglary increases in area i, we would changes in i. expect the amount of burglary offenses Directl y measuring spillover to decline as potential burglars view effects is difficult given the paucity of the expected net payoff from burglary location specific data. Three types of to be less than that involved in methodologies have been employed in alternative activities. Secondly, facing an attempt to indirectly measure the a lower relative probability of being geographic spillover phenomena: 1) caught in jurisdiction j, we would Experimental (Press 1972, Chaiken et expect some criminals from i to move al. 1974, Knapp et al. 1980); 2) Partial to j in order to commit burglaries. equilibrium (Mehay 1977, Fabrikant Thus, the amount of burglary offenses 1980); and 3) General equilibrium would rise in j, as burglars from i enter (Hakim, Ovadia, Sagi, and Weinblatt j. The increase in burglaries in j might 1979, Mathieson and Pass ell 1976, Avio be termed the tlgross spillover effect" of and Clark 1978, Kennett 1982; Mehay an increase in law enforcement 1981, McPheters and Stronge 1981, activities in i (Hakim et aI. 1979). Katzman 1981). However, if jurisdiction j Given the wide variety of types responded to the increase in burglaries of models used and the different data within its jurisdiction by increasing sets, it is somewhat remarkable that deterrence activities, then the same most of the studies in the geographic results that occurred in jurisdiction i spillover literature come to the same would also occur in j. Some burglars conclusion: when spillovers are found would find the probability of being to exist, they are small. In general, it is arrested for burglary increasing and found that a majority of the crimes would thus opt out of the burglary investigated (property and violent profession, while others might tend to crime) occur in close proximity to the redirect their efforts to jurisdiction i criminal's home residence, ranging since the probability of being arrested from a few blocks to three miles in i relative to j has now fallen. (McIver 1981, 24). This might explain Eventually, a new equilibrium in terms the small levels of measured spillover.2 of police expenditures and the level of Nonetheless, the evidence is
18 Geographic Spillover Effects Sollars
sufficiently strong enough to suggest of the community. In order to reduce that an effort to fully specify a model any bias which would result from of property crime determination should simply analyzing one sector of the account for potential geographic model, a multi-equation a structural spillovers between nearby jurisdictions. model is suggested. The model used to evaluate the 3. Model Specification potential property crime spillover effects is specified as follows (expected Cameron's (1988) review of the signs are also included): crime literature suggests that most econometric models of property crime (4) OTP = f{ AOTP, ASPP, PV ALUE, use a simultaneous systems approach AVWAGES, UR, ABLKPCT, in order to model criminal and law TP A1524, DENS, SWDIFR, enforcement activity. According to DRDIFR, u} Cameron, the basic structure is given by where: OTP = property crime rate (1) C = f{ CR, S ... } AOTP = arrest/offense ratio, fAoTP < 0 ASPP = stolen property arrests rate, fASPP < 0 PV ALUE = assessed property value, (2) CR = g{ C, I ... } fpVALUE <> 0 A VW AGES = average wage or salary, (3) 1= h{ C .. , } fAVWAGES < 0 UR = unemployment rate, fUR> 0 where: ABLKPCT = percent of population which is C = property crime rate black, fABLKPcT > 0 CR = clearance rate (probability of arrest TP A1524 = percent of population aged 15 to and / or conviction) 24, fTPA1524 > 0 S = severity of punishment DENS := population per square mile land area, I = police resources fDENS > 0 DIFR = police resources differential ratio, This type of formulation is fDIFR > 0 appropriate because of the DRDIFR = drug arrests differential ratio, fDRDIFR <> 0 interrelationships between the different U = error term components of the model. For example, law enforcement activity (4) might be termed a crime should reduce crime rates, but it is also supply equation. In modelling the expected that crime rates help to crime supply function, we assume that determine the level of law enforcement individuals who engage in crime are activity. The effectiveness of police rational economic agents faced with will also, in part, be determined by the both legal and illegal income earning caseload they face (reported crimes) possibilities.3 The purpose of the crime and the level of inputs. The level of supply function is to determine those inputs will in turn be influenced by the factors which affect the decisions of amount of crime and the fiscal ability rational economic agents to commit
19 Geographic Spillover Effects Sollars crimes (as opposed to other forms of therefore expect that wealthier areas economic activity). These factors can will attract more potential criminals.s be divided into three categories: On the other hand, wealthier areas will Potential direct costs and benefits from also tend to be populated by those crime; opportunity costs; and other with much higher opportunity costs for relevant socioeconomic or criminal behavior, thereby resulting in environmental characteristics. less criminal activity. Wealthy The major disincentive for a individuals will also probably spend potential criminal is the possibility of more on protection from private sector detection, arrest, and conviction, both markets for such goods as guards, in a psychic as well as in a pecuniary alarms, etc., which also deter crime. sense. Ideally, we would also want to The type of wealth effect which include a severity of punishment dominates is not clear. variable such as probability of sentence The second category of factors if convicted or the length of prison which affect crime rates are term. These data, however, do not yet opportunity cost variables. Since a exist at the county or municipal level . property crime is committed to either of aggregation (the data type used in have command over the good which is this study). Moreover, given the stolen or derive income from the sale sentencing guidelines in Florida, the of the good, this second category will variation in punishment may not be as include those factors which determine large as it would otherwise. We can the possible returns from legal activity. proxy for the subjective probability of The first of these is the return from arrest by the arrests! offenses ratio.4 legal employment. We would expect The expectation is that increases in the higher legal wage rates to have a arrests! offenses ratio will deter negative influence on a potential property crime. criminal's decision to commit crimes. On the other hand, the potential However, the ability to earn legal return to criminals from committing incomes will be determined by general property crimes is a function of many labor market conditions. Areas with factors. Since stolen goods are higher levels of unemployment will generally sold to generate income, offer fewer legal income earning police efforts to reduce the size of the opportunities, suggesting that illegal stolen goods or "fence" market will also activity may be more prevalent in discourage property crime. Secondly, those areas. Differences in reported the gains from property crime will also legal income may also be a function of be determined by the type and value of the racial makeup of the population. potential stolen goods. Wealthier areas This is not meant to suggest the will have, holding everything else existence of an inherent predilection constant, a variety and abundance of towards crime by any racial group; high value goods Gewelry, electronics, rather it is viewed as a result of the etc.) relative to that found in poorer lack of economic opportunity (possibly areas. For this reason, we would due to discrimination) not picked up
20 Geographic Spillover Effects Sollars
by the other variables. spillover effect, we include a police . The third major category of resource differential variable, DIFR, m factors reflect other general the crime supply equation. DIFR is socioeconomic or environmental calculated as follows:6 variables. The first of these is the age , (5) DIFR = "L/ =1 (I/N) / n distribution within a given community. : (IJN;) According to Wilson and Herrnstein . (1985), uCriminal behavior depends as where: much or more on age than on any I = measure of police inputs other demographic characteristics - N = population sex, social status, race,' family . n = number of contiguous or "nearby" configuration, etc. -- yet exammed by jurisdictions criminologists." (Chapter 5, 126) i = home jurisdiction Community size and population j = other jurisdictions density may be an important . The emphasis on drug arrests determinant of the property cnme rate may also have an effect on the (Hoch 1974). Densely populated urban incidence of property crime across areas are expected to have higher crime communities. Other studies have rates than rural areas. This may be due explored the possibility that drug to the fact that densely populated areas arrests influence the level of property offer larger net returns to criminal crime (Benson and Rasmussen 1991, activity since traveling distance Sollars et al. 1991, Benson et al. 1992). between targets is low, or because of It may also be the case that this the relative anonymity offered by an influence might extend into other urban setting. jurisdictions. For example, suppose Measuring Spillovers. that as police crack down on drugs Measurement of the crimes in j within one jurisdiction, both drug committed by those who enter from dealers and drug users turn toward jurisdiction i, minus those criminals surrounding areas as locations where who either dropped out of the market they can either sell or purchase drugs or moved to jurisdiction j, would thus and commit property crimes. give the "net spillover" effect of an Residents of j are not limited to increase in jurisdiction i's police committing crimes only in j. If they activity. Thus, from a measurement cross over to community i, then point of view, the net spillover effect reported crime in i will increase. would give the true cost of the On the other hand, a rise in externality imposed on j as the result drug arrests in j may result in drug of policy changes in i. Unfortunately, suppliers in j moving to to continue this can not be measured directly. i their trade. If drug supplies in Instead, we assume that differen~es in community i increase, then drug prices the level of police resources act, m in i will decrease, resulting in a essence, as relative prices to potential reduction of the need on the part of criminals. In order to capture this
21 Geographic Spillover Effects Sollars drug users to finance their drug Florida for the y:!ars 1984 to 1987.9 Of consumption via property crime. the 357 jurisdictions, there are 66 Hence, the effects of a relative county sheriffs offices and 291 crackdown on drugs across municipal police agencies.lO We communities i ar,d j are ambiguous. assume that the sheriff's jurisdiction is Drug suppliers can be spilled into the area within the county not covered community i if police crackdown in j by a municipal police department. (which should reduce the level of Due to missing data for some of the property crime in i), but drug users in observations for some of the years, less community j, now facing higher drug than 350 observations are available. prices, may also move to community i The offense and arrest statistics to commit property crimes or in search are from the Uniform Crime Reports of a more abundant supply of drugs. series collected by the Florida To measure the influence of the relative Department of Law Enforcement emphasis on the drug trade by police (FOLE, various years). There are three across jurisdictions, a variable intended measures of the level of police to capture the relative drug emphasis resources. Sworn officers per capita, effects, DRDIFR, is included in the SWPP, is a measure of the labor crime supply function. DRDIF is component used by police agencies, calculated in the same fashion as OIFR and is dominant in terms of total with the number of drug arrests budget allocation (approximately 75 replacing police resources: percent). This data is provided by FOLE. Alternative measures of the (6) DRDIFR = '2./=1 (O/Nj)/n levc.] of law enforcement activity are (D;lNj) municipal or county noncapital and total expenditures for law enforcement. "Nearby jurisdictions" in both (5) This data is provided by the and (6) are those jurisdictions within a Comptroller, Bureau of Local five mile radius of the home Government Finance, Florida jurisdiction? The other equations Department of Banking and Finance (arrests / offenses and police resources) (various years). Given the lumpy specified are variants of (2) and (3). nature of the capital expenditures, we Given space limitations, these equations are not explored in detail. focus only on noncapital expenditures. Total municipal revenues, Further discussion of the empirical population, and assessed property model, with respect to these equations value data are also provided by the and the empirical results, can be found in Sollars (1991).8 Comptroller, Bureau of Local Government Finance (various years). Measures of other types of 4. Data and Estimation socioeconomic data, such as per capita income, age and racial distributions, The data set consists of the unemployment, and average wages are political jurisdictions in the state of only available at the county level for
22 Geographic Spillover Effects Sollars
the years 1984-1987. Since municipality met--that is, the number of exogenous data of this type is not available for the variables excluded from the ith time period under consideration, all of equation must be at least as large as these variables are calculated at the the number of included endogenous county level so that each of the variables. Since only one endogenous communities within the county are variable is included on the right hand assumed to have the county level side of both the crime supply and characteristics. For variables such as arrest / offenses, this is obviously not a DR and AVW AGES, this is not problem. Below, we report the necessarily a problem as labor markets regression results for the OLS and two are geographically large and labor stage least squares (TSLS) estimations. quite mobile within these markets. The variables were converted to pey, UR, and AVW AGES are provided natural logarithms, so the estimated by Bureau of Economic and Business coefficients reported below are Research, University of Florida (various interpreted as elasticities. years). The age distribution variables are provided by the Population 5. Empirical Results Program, Bureau of Economic and Business Research (various years) Table 1 reports the OLS results based on their intracensus of the crime supply equation (4) where interpolations and other standard aggregate property crime (OTP) is the demography techniques. The black dependent variable and noncapital population percent variable ABLKPCT expenditures is used to calculate the is also from the BEBR. Again, these differential variable (NCDIFR).l1 As are county-level variables. hypothesized, AOTP has a negative Estimation of the empirical coefficient and is statistically significant model is dependent upon the for each year. The coefficients on specification chosen. The three AOTP range from -0.178 in 1986 to - equation model suggests that 00407 in 1985. The interpretation given simultaneous equation bias could this outcome is that potential criminals occur. If simultaneity is not assumed, are indeed rational--higher probability then the parameter estimates can be of arrest leads to less property crime, estimated by ordinary least squares ceteris paribus; PV ALUE has a (OLS). On the other hand, assumed positive, significant coefficient (except simultaneity of the endogenous for 1987), implying wealthier areas variables calls for the use of suffer from more property crime (or at simultaneous equations techniques. least report it more) than do poorer When using a simultaneous equations areas, holding other factors constant. estimation procedure, identification is Areas which are more densely logically prior to estimation (Hoenack populated, and those with a larger and Weiler, 1980). In the three black population also -suffer from more equation model discussed above, property crime, holding everything else identification by the order condition is constant. The other labor market and
23 C) ro Table 1. Property Crime Detenninants, Noncapital Table 2. Property Crime Detenninants, Sworn Officers 0 ~ Expenditures as the Measure of Police Resources, OLS. as the Measure of Police Resources, OLS Pl ~ Regression No. (01) (02) (03) (04) Regression No. (01) (02) (03) (04) 2: () Year 1984 1985 1986 1987 Year 1984 1985 1986 1987 Cf) ~ ...... ~ CONSTANT -0.439 2.691 4.314 3.900 CONSTANT -0.573 2.835 4.083 4.114 0 (0.17) (1.84) (1.77) (1.52) (0.22) (1.92) (1.69) (1.57) PVALUE ..... 0.174 0.173 0.274 0.059 PVALUE ..... 0.199 0.201 0.295 0.077 (4.13) (3.17) (5.09) (1.33) (4.86) (3.75) (5.78) (1.74) AVWAGES 0.444 0.208 -0.060 0.252 AVWAGES 0.435 0.151 -0.052 0.208 (1.58) (1.30) (0.21) (0.89) (1.52) (0.95) (0.19) (0.74) ~ II UR ...... 0.328 -0.017 0.061 0.063 UR ...... 0.302 -0.021 0.072 0.046 (3.31) (0.17) (0.59) (0.56) (3.00) (0.20) (0.70) (0.40) ABLKPCT .... 0.118 0.118 0.128 0.067 ABLKPCT .... 0.126 0.126 0.137 0.073 (6.20) (5.66) (6.65) (3.37) (6.52) (5.96) (7.15) (3.64) TPA1524 0.040 -0.087 -0.101 -0.205 TPA1524 0.049 -0.058 -0.125 -0.193 (0.24) (0.46) (0.60) (1.26) (0.29) (0.31) (0.75) (1.19) DENS ...... 0.224 0.221 0.233 0.189 DENS ...... 0.222 0.224 0.235 0.194 (11.09) (10.72) (10.80) (8.92) (10.75) (10.67) (10.96) (9.07) NCDIFR 0.224 0.166 0.119 0.106 SWDIFR 0.175 0.126 0.155 0.066 (3.92) (2.56) (2.25) (1.54) (3.44) (2.25) (3.03) (1.19) DRDIFR 0.093 0.158 0.135 0.196 DRDIFR 0.100 0.16] 0.119 0.]98 (2.85) (4.65) (3.84) (5.]3) (2.99) (4.56) (3.42) (5.35) R,2 ...... 0.6]8 0.6]0 0.600 0.525 R,2 ...... 0.620 0.610 0.6]4 0.527 F-statistic ..... 34.011 34.252 32.493 26.7]2 F-statistic ..... 33.612 33.344 34.039 26.274 '[ Cf)2- N ...... 205 214 214 234 N ...... 201 208 209 228 IfffFl t-statistics are in parentheses t-statistics are in parentheses Table 3. Property Crime Determinants, Noncapital Table 4. Property Crime Determinants, Sworn Officers Expenditures as the Measure of Police Resources, TSLS. as the Measure of Police Resources, TSLS C) ro Regression No. (01) (02) (03) (04) Regression No. (01) (02) (03) (04) 0 ~ Year 1984 1985 1986 1987 Year 1984 1985 1986 1987 ~ '"d CONSTANT 4.767 2.682 4.084 3.337 CONSTANT 4.214 ::r..... 2.576 2.893 4.354 n (1.29) (2.09) (1.57) (1.21) (1.24) (1.73) (1.03) (1.57) CJ) '"d..... AOTP ...... -0.918 0.032 0.132 -0.224 AOTP ...... -0.750 0.342 0.355 0.432 ..... (4.18) (0.17) (0.62) (0.98) (3.63) (1.46) (1.52) (1.84) 0 AVWAGES -0.248 0.250 -0.048 0.340 AVWAGES -0.170 0.230 0.067 0.176 (0.59) (1.73) (0.16) (1.08) (0.44) (1.39) (0.21) (0.56) UR ...... 0.190 0.088 0.007 0.028 UR ...... 0.197 0.102 -0.001 0.047 (1.34) (0.95) (0.06) (0.24) (1.55) (0.94) (0.01) (0.39) ~ II ABLKPCT .... 0.142 0.110 0.129 0.068 ABLKPCT .... 0.142 0.127 0.144 0.081 (5.23) (5.79) (6.39) (3.33) (5.89) (5.80) (6.65) (3.84) TPA1524 0.267 -0.264 -0.126 -0.248 TPA1524 0.232 -0.329 -0.193 -0.161 (1.09) (1.46) (0.72) (1.35) (1.06) (1.59) (1.04) (0.88) DENS ...... 0.226 0.234 0.250 0.189 DENS ...... 0.227 0.252 0.257 0.191 (8.00) (11.83) (10.57) (8.86) (8.87) (10.80) (10.15) (8.84) NCDIFR 0.123 0.189 0.250 0.189 SWDIFR 0.082 0.116 0.175 0.980 (1.43) (3.17) (2.15) (1.83) (1.21) (1.89) (2.97) (1.71) DRDIFR 0.146 0.114 0.090 0.170 DRDIFR 0.145 0.108 0.048 0.181 (3.03) (3.70) (2.30) {4.15) (3.37) (2.91) (1.11) (4.63) R,2 ...... R,2 ...... F-statistk ..... F-statistic ..... N ...... 199 210 209 229 N ...... 199 204 204 223 t-statistics are in parentheses t-statistics are in parentheses I II £ Pl >-ten Geographic Spillover Effects Sollars age variable coefficients (A VW AGES, expendi tures as the measure of the UR, TPA1524) are generally not police resources differential variable statistically significant.12 ASPP, the (SWDIFR) does not markedly change stolen property market control variable, any of the outcomes of the coefficients has a positive sign (not as anticipated) in the crime supply equation (see Table but low levels of statistical significance. 2). Once again, the AOTP coefficient is The hypothesized sign was negative, statistically significant and negative, the inference being that a crackdown and both spillover variables (SWDIFR on the fence market would reduce the and DRDIFR) have the expected returns to property crime thereby positive coefficient (although the resulting in a lower crime rate. On the SWDIFR coefficient for 1987 is not other hand, it may simply be the case statistically significant), indicating the that the positive correlation betw"een presence of spillovers. As in the OTP and ASPP reflects the fact that previous case, the coefficient on areas with higher levels of property DRDIFR for 1987 is twice the size of crime give police more opportunities to the coefficient for 1984. In general, the make stolen property arrests. explanatory power of the model is The two spillover variables, fairly good as evidenced by the NCDIFR and DRDIFR, have adjusted (for degrees of freedom) R coefficients which are strongly positive. squared and F statistic. The adjusted This is interpreted as a measure of the R-squared is approximately 0.60 over spillover of property crime resul ting the four years. from the differential in police spending In order to control for the and drug arrests respectively across possibility of simultaneity bias which communities which are in close might result, the system is estimated proximity to one another. In via a ISLS / recursive technique. This magnitude, the individual spillover entails the following. First, the effects are small (the elasticities are not predicted values of the police resource large), a finding consistent with much equation are used in the of the literature. However, the arrest/offenses equations. Next, two combined coefficients are often as large stage-least-squares (ISLS) regressions as some of the other coefficients, such are run on (4) and the arrests / offenses as AOTP. Moreover, the existence of equation with the fitted values of the drug arrests spillover (police SWPP and NONCAP. Use of ISLS crackdowns on drugs in surrounding gives results which are biased but jurisdictions increases the property consistept, and purged of simultaneity crime rate in the home jurisdiction) is bias. important given the fact that the The results for the TSLS coefficient on DRDIFR increases over equations are given in Tables 3 and 4. the four year period (0.098 in 1984 to Most strikingly, the use of this 0.196 in 1987). estimation model renders the Replicating regressions 1-4 using deterrence variab!e, AOTP, sworn officers rather than noncapital insignificant in more than one-half of 26 Geographic Spillover Effects Sollars the years. The rest of the crime supply making drug arrests. The data equation results are similar to those employed in this study to capture the reported using the other estimation possibility of the spillover effects were techniques. ASPP is not significant, somewhat unique to the crime spillover and PV ALUE's coefficient is positive literature. Rather than focus on a and generally statistically significant. central city or metropolitan area, the ABLKPCT and DENS have coefficient~ spotlight was instead on a varjety of which are positive. The coefficients on types of communities located within a the spillover variables are generally single stateY The results reported here significant and positive. The TSLS suggest that spillovers do exist, and results suggest, again,. that spillovers that these spillovers may be larger than may indeed a problem. The some of the previous studies in this coefficients remain roughly the same area suggest. More importantly, size. spillovers might occur over distances Of course, the estimation larger than a city precinct or techniques used are determined by the metropolitan area. actual specification of the model. Economic theory offers a Unfortunately, the "true" specification number of solutions for spillover of the model is never known. In problems arising from private activity. general, the much debated deterrence For example, smokestack pollution hypothesis is generally supported by creates an efficiency problem since the the evidence (a result u.sually found in owners of the plant do not bear the full the economics of crime literature and cost of the resources used in much debated by other social production (clean air is used up by the scientists). Secondly, a good case can smoke). Hence, the firm produces be made that geographic property more than the socially optimal amount. crime '~~pillovers do exist as a result of Neighbors adjacent to the factory bear relative differences in the level of the cost of the pollution. To increase police resources and drug-law efficiency, economic theory suggest enforcement. partial solutions ranging from taxation to merger to payment of damages. The 6. Implications and Conclusions goal of these schemes is to force the spillover generator to bear the full The hypothesis that geographic costs of their activities, thereby spillovers of property crime could bringing the optimal private decision result if there were location-based making calculus in line with the differentials in the level of police maximization of social welfare. resources and drug arrests has An analogy to the smokestack evidence in the Florida data. Criminals problem can be found in the case of appear to respond ·to differentials in law enforcement activity. From a police activity by moving to areas policy perspective, the evidence where police are less abundant or presented above suggests that an where there was less emphasis on individual police agency does not exist 27 Geographic Spillover Effects Sollars in a vacuum. The enforcement choice ENDNOTES of one agency within one jurisdiction affects not only individuals in that 1. For a review of this literature, see Sollars jurisdiction but others in surrounding (Chapter 2, 1991). jurisdictions as well. While a "get_ 2. For example, of a police department toughII policy within a given increased the number of officers assigned to a jurisdiction may achieve desired goals shopping mall in order to increase protection for the jurisdiction, surrounding for the mall, it may be the case that the jurisdictions must cope with the new criminals simply move down the street to the influx of crime which is analogous to next mall. If both malls are in the same jurisdiction (or unit of observation), then the smokestack pollution. What is spillovers would be measured. desirable from the viewpoint of a single jurisdiction may not be desirable 3. This does not imply that all crime for the surrounding jurisdictions. perpetrators are always the rational, The spillover phenomena cost/benefit calculator of economic theory. suggests that decisions concerning the What it does imply is that criminal behavior can be influenced in a predictable way be allocation of police resources cannot be changes in the incentive structure. made independently; coordination of pol!ce activity across jurisdictions may 4. The arrests/ offenses ratio is a proxy for the be required if true crime deterrence is subjective probability of being arrested. In the to take place. This is not to suggest crime literature, this variable has been called that local law enforcement must be the probability of arrest as well as the clearance rate. It would seem, however, that controlled by state and Federal these definitions imply a precision that is not officials, since this type of dominanLc necessarily justified. For example, AOTP can creates a whole other range of exceed unity, which violates the notion of problems. It is possible, however, for probability that the variable is supposed to local governments and law represent. With respect to the term "clearance enforcement agencies to share ratio," this term has a precise definition in police tenninology -- the number of cases information and resources, coordinate closed by arrest, or by some other means. policing strategies, and plan for future corrections facilities. How this 5. Many previous studies use the value of coordination would work is beyond the reported stolen goods as a proxy. Note, scope of this study, but is an issue that however, that this statistic is, in part, must be considered by policy makers. determined by reporting behavior which may be a function of police activity. 6. This fonnulation is similar to the one used by Mehay (1977). 7. Estimates of land area and the maps used to construct the variables are from Purdum and Anderson (1988). . 8. The arrests/offenses equation contains the following explanatory variables: property 28 Geographic Spillover Effects Sollars crime rate, measures of police resources, REFERENCES percentage of arrests which are drug arrests, and population density. The police resources Avio, Kenneth 1., and s.c. Clark. "The Supply equation contains both the property and of Property Offenses in Ontario: violent crime rates, property value and income, Evidence on the Deterrent Effect of drug arrests, and the number of "nearby" Punishment." Canadian TournaI of police agencies. Economics 11(1) (1978): 1-19. 9. The time frame is in part determined by the Becker, Gary S. "Crime and Punishment: An data. Comparable police expenditure data for Economic Approach." TournaI of earlier years are not available, and due to Political Economy 78 (Marchi April changes in Uniform Crime Reports for 1988, 1968): 169-217. the 1988 arrest and offense data are not consistent with earlier data. Benson, Bruce L., and David W. Rasmussen. "The Relationship Between Illicit Drug 10. Other law enforcement agencies include Enforcement Policy and Property state agencies, such as the Florida Highway Crimes." Contemporary Policy Issues 9 Patrol, university police, and airport police (October 1991): 106-115. departments. Both the number of offenses reported to and the number arrested by these ____, IIjoong Kim, David W. Rasmussen, agencies is small. Thus, they are not included Thomas W. Zuehlke. "Is Property in the data set. Crime Caused by Drug Use of Drug Enforcement Policy?" .6E.P.li.ed 11. OTP is the sum of robbery, burglary and Economics (1992) Forthcoming. larceny crime rates. Block, Michael K. and John M. Heineke. "A 12. TP A1544, another measure of the crime Labor Theoretic Analysis of Criminal prone population (ages 15-44) performs just as Choice." American Economic Review poorly as TP A1524. 65(3) (1975): 314-25. 13. Interestingly, Florida's geographic makeup Blumstein, A., J. Cohen, and D. Nagin, ed. as a peninsula and lack of large population Deterrence and Incapacitation: centers along Florida's northern border Estimating the Effects of Criminal reduced the need to control for interstate Sanctions on Crime Rates. Washington, influences. D.C.: National Academy of Sciences, 1971. Brier, S. and S.E. Feinberg "Recent Econometric Modelling of Crime and Punishment: Support for the Deterrence Hypothesis." Evaluation Review 4 (1980): 147-191. Buck, Andrew J., Meir Gross, Simon Hakim, and J. WeinblaU, J. "The Deterrence Hypothesis Revisited." Regional Science and Urban-Economics 13 (1983): 471-486. Bureau of Economic and Business Research. 29 Geographic Spillover Effects Sollars Florida Statistical Abstract. Florida Department of Law Enforcement. Gainesville: University of Florida, Crime in Florida, Annual Report. various years. Tallahassee: Florida Department of Law Enforcement, various years. ____. Florida Estimates of Population, April .1. 1989. Gainesville: University Greenberg, D.F. "Crime Deterrence Research of Florida, 1989. and Social Policy." In Modeling the Criminal Iustice System, ed. 5.5. Nagel, Cameron, Samuel. "The Economics of Crime 281-295. Beverly Hills: Sage Deterrence: A Survey of Theory and Publications, 1977. Evidence." KykIos 41 (1988): 301-323. Hakim, Simon, Arie Ovadia, Eli Sagi and J. Carr-Hill, R.A. and N.H. Stem. "An Weinbiatt. "Interjurisdictional Spillover Econometric Model of the Supply and of Crime and Police Expenditure." Control of Recorded Offenses in Land Economics 55 (May 1979): 200- England and Wales." TournaI of Public 212. Economics 2 (1973): 289-318. ____ and George F. Rengert, ed. Crime Chaiken, J.M. "What's Known about Deterrent Spillover. Beverly Hills: Sage Effects of Police Activities." Santa Publications, 1981. Monica, CA: Rand Corporation, 1976. Hoch, Irving. "Factors in Urban Crime." ____, M. Lawless, and K. Stevenson. Journal of Urban Economics 11 (1974): The Impact of Police Activity on 184-229. Crime: Robberies on the New York City Subway System. New York: Rand Hoenack, Stephen A. and w.e. Weiler, "A Institute, 1974. Structural Model of Murder Behavior and the Criminal Justice System." Cloninger, Dale .0. "The Deterrent Effect of American Economic Review 70(3) Gune Law Enforcement: An Evaluation of 1980): 327-34]. Recent Findings and Some New Evidence." American TournaI of Katzman, Martin T. "The Supply of Criminals: Economics and Sociology 34 (1975): A Geo-Economic Examination" in 323-336. Simon Hakim and George F. Rengert, eds, Crime Spillover. Beverly Hills: Comptroller, Department (If Banking and Sage Publications, 1981. Finance. Local Government Financial Report. Tallahassee, FL: State of Kennett, David. "Standards and Procedures for Florida, Department of Banking and the Distribution of a Public Service: Finance, various years. Shoup Revisited." Public Finance 37(1) (1982): 80-97. Ehrlich, Isaac. "Participation in Illegitimate Activities: An Economic Analysis." Knapp, F., J.A. Sharkey and J.R. Metts. "The TournaI of Political Economy 81 Effects of an Increase in a County (May/June 1973): 521-565. Sheriff's Departments Patrol Capability." TournaI of Police Science Fabrikant, Richard. "Interjurisdictional and Administration 8 (March 1980): 5- Spillovers of Urban Police Services: 14. Comment." Southern Economic TournaI 46(3) (1980): 955-961. Mathieson, Donald and Passell, P. "Homicide 30 Geographic Spillover Effects Sollars and Robbery in New York City: An Institute of Science and Public Affairs, Economic Model." Journal of Legal Florida State University, 1988. Studies 6: (1976): 83-98. Reppetto, Thomas A. "Crime Prevention and McIver, John P. "Criminal Mobility: A Review the Displacement Phenomenon." of Empirical Shldies." In Crime Crime and Delinquency 22(2) (1976): Spillover, ed. Simon Hakim and 166-77. George F. Rengert, 48-53. Beverly Hills: Sage Publications, 1981. Rowe, A.R. and c.R. Tittle. "Life Cycle Changes and Criminal Propensity." McPheters, Lee R. and Stronge, William B. The Sociological Quarterly 18 (1977): "Crime Spillover in the Boston Area." 223-236 In Crime Spillover, ed. Simon Hakim and George F. Rengert, 83-95. Beverly Schmidt, Peter, and Ann D. Witte. An Hills: Sage Publications, 1981. Economic Analysis of Crime and Tustice: Theory, Methods, Applications. Mehay, Stephen 1. "Intetjurisdictional Orlando: Academic Press: 1985. Spillovers of Urban Police Services." Southern Economic TournaI 43 (1977): Sjoquist, David L. "Property Crime and 1352-1359. Economic Behavior: Some Empirical Results." American Economic Review ____. "Burglary Spillover in Los 63 Gune 1973): 439-446. Angeles." In Crime Spillover, ed. Simon Hakim and George F. Rengert, Sollars, David L. "Allocation of Law 67-82. Beverly Hills: Sage Publications, Enforcement Resources, Spillovers, and 1981. the Illicit Drug Market." Ph.D. diss., Florida State University, 1991. Office of National Drug Control Strategy. National Drug Control Strategy. Taylor, John B. "Econometric Models of Washington D.C.: Government Criminal Behavior: A Review." in J.M. Printing Office, 1990. Heineke, ed. Economic Models of Criminal Behavior. Amsterdam: Palmer, Jan. "Economic Analysis of the North-Holland Publishing, 1978. Deterrent Effect of Punishment; A Review." TournaI of Research in Crime Thaler, Richard. "An Econometric Analysis of and Delinguency 14 (1977): 4-21. Property Crime: Interaction Between Police and Criminals." Journal of Population Program, Bureau of Economic and Public Economics 8(1) (1977): 37-52. Business Research. Florida Estimates of Population. Gainesville: University Trumbull, William N. "Estimations of the of Florida, various years. Economic Model of Crime Using Aggregate and Individual Level Data." Press, James S. Some Effects of ill!. Increase in Southern Economic TournaI 56(2) Police Manpower in the 20th Precinct (October 1989): 423-439. of New York City. New York City Rand Institute, 1971. Wilson, James Q. and Richard J. Hermstein. Crime and Human Nature. New York: Purdum, Elizabeth D. and James R. Anderson, Simon and Shuster, 1985. ed. Florida County Atlas and Municipal Fact Book. Tallahassee, FL.: Witte, Ann D. "Estimating the Economic Model 31 Geographic Spillover Effects Sollars of Crime with Individual Data." Quarterly Journal of Economics 94 (1980): 57-84. 32 Moral Panics and the Vocabularies of Motives: A Content Analysis of the Cyclical Nature of the Drug Crisis, 1970-1985 11yrna Cintron, PhD. ======~ INTRODUCTION they set standards of moral conduct upon the population. Assessments of The recently reported rise in drug these crusades have raised questions abuse in the United States has led to a about the wisdom of enforcing laws national perception of a "drug crisis" that are directed at victimless crimes that requires immediate and far (Schur, 1965), which in turn may reaching control efforts. Anti-drug precipitate secondary deviance (Lemert, efforts have received major media 1951). coverage and presidential endorsement. Despite past experiences with A national campaign has been directed drug control, the current drug crisis is toward ed1:cating the public to "just occurring at a time of unprecedented say no" to drugs. A drug-free America adversarial relationships between the by 1995 has become the stated policy high demand for a single drug of the United States government (Anti --cocaine, crack-- and the national and Drug Abuse Act of 1988). international efforts to thwart the From the early 1900s to the supply of narcotics. present, most observers argue that The theoretical assumptions of the various United States anti-drug labeling and critical perspectives laid strategies have been ineffective in the ground for empirical inquiry to controlling the so-called menace or clarify the interests that are embodied crisis. On the one hand, the efforts are in legislative or moral reforms. thought to have amplified the drug Previous studies have attempted to problem while failing to curb their single out the significant elements illicit availability. On the other hand, underlying the development of reforms legal responses have stimulated a series to control illegal drugs in the United of what Gusfield (1963) has termed States. These studies tend to give symbolic crusades. Skolnick (1967) particular weight to the historical refers to the enforcement of symbolic context in which control takes place. crusades as "coercion to virtue II because Overall, the development of this This paper is from Dr. Cintron's dissertation, "An Assessment of and Evolving Control Strategy: Cocaine, the United States and I.i1tin America," Tom Blomberg, Major Professor. Myrna Cintron holds B.A. and M.A. degrees from the University of Puerto Rico, and an Ph.D. in criminology from the Florida State University. She is now Assistant Professor in the Department of Sociology at West Georgia College, where she teaches criminolo8lJ and sociology. Her cllrrent research interests include historical analysis of crime, and drug use and abuse. In AU8llst 1993, Dr. Cintron will join the faculty in the Criminal Justice Center at Sam HOllston State University. 33 Moral Panics Cintron literature can be divided into three eliminate. They also shape the public's categories (Le. the symbolic moral attitudes, beliefs, and reactions toward crusader, organizational legitimacy or the labeled behavior. survival and the policy decision More recent drug control studies making process). Underlying these give particular attention to the policy studies is the importance of a single de~ision making processes in which factor as being responsible for the control policies are developed adoption of particular anti-drug (Blumberg, 1973; Himmelstein, 1978; policies (for example, the Marijuana Morgan, 1983; Reasons, 1974; Tax Act of 1937). Reinarman, 1983). Researchers assume By 1970 a growing number of that the nature of drug use; its threats; studies had been developed supporting ideological, political, and economic the symbolic nature of drug controls factors; and interest woups all bear (Be:'7ker, 1963; Dickson, 1973, 1968; upon the federal government's control Duster, 1970; Helmer, 1975; Morgan, efforts (Reinarman, 1983). Moreover, 1978a, 1978b; Musto, 1973; Norland these studies suggest that some of and Wright, 1984; Reasons, 1974). these elements serve to facilitate The organizational legitimacy control efforts, while others impede the theme assumes that criminal laws are process. determined by powerful groups. These groups, social control agencies in PURPOSE AND METHOD particular, use criminal laws to advance their interests and professional The mass media (newspaper, preferences. Agents of social control, radio, television, magazines) is "may not be interested in the content of significant in, generating public concern the rule itself, but only in the fact that about what becomes defined as a social the existence of the rule provides problem. The news stories content not [them] with a job, a profession, and a only gives visibility to an issue, it is raison d' etre" (Becker, 1963, p. 156). also in a position to influence what Social control organizations have a society thinks about and how society vested interest in the amplification of reacts about a problem. Through the the problem. Dickson (1973, 1968) has mass media representation of a social suggested that drug legislation is a problem the public becomes vicarious bureaucratic response to environmental participants in the definition of the pressures. Deviance outcomes reflect problem. It has been documented how the organization's day-to-day working public campaigns have been generated environment. Its legitimization efforts through the mass media to bring about to gain public and political support the support and convinced the public and bureaucratic needs for survival that a specific problem (Le., crime) is a when faced with decreasing budget serious and growing tp.reat to society appropriations. Under these terms, (Bortner, 1984; Braden, 1973; Cohen, then, control strategies create the 1973a, 1973b; Davis, 1973; Fishman, conditions they are formally charged to 1984, 1978; Hall, et al., 1978; Lewis, 34 Moral Panics Cintron 1984; Linsky, 1973; Nunnally, 1973; RESULTS Pearce, 1973; Phillips, 1973; Sheley and Ashkins, 1984; Terry, 1984; Young, Of the two hundred and twenty 1973a, 1973b). five articles sampled, 28.0 percent The appearance of newspaper (sixty-three articles) were published stories regarding drug abuse is during the first period of study (1970- heightening the official and public 1974), 33.8 percent (seventy-six articles) sensitivities about the existence and were published during the second nature of the drug crisis. For this period of study (1975-1979), and 38.2 stud y newspaper stories will be the percent (eighty-six articles) in the last channel through which the current period (1980-1985). Nineteen Seventy drug crisis will be interpreted and Nine was the peak year for articles analyzed. In this capacity the study's sampled with thirty-one drug related purpose is twofold. First, it will stories; 1973 had only four stories describe the natural history of a social sampled. On the average, fourteen problem. This study will cover the news articles were sampled for each period between 1970 and 1985. year. Secondly, it will quantify the Over fifty percent of the sampled newspaper coverage of the drug crisis articles were categorized as news of and the social context as it is expressed general interest. These included, in the press as they promote a social within the same article, topics such as problem that is considered drug-related crimes, descriptions of newsworthy. The study is important users, federal and local agencies work, because it critically assesses the seizures, task forces, and technological interplay of issues and interests that help. have merged in the current war on Almost thirteen percent of the drugs being waged by the United drug-related stories dealt with survey States. results, either at the state or national A simple random sample was level. Over nine percent of the articles selected for the study period between discussed legislation, either bills that January of 1970 and December of 1985. were proposed or were being discussed To decide how many cases were at the state or federal levels. Only needed form each one of these years, fifteen articles were classified as having the total sample was distributed human interest. That is, only 6.7 according to the proportion of news percent of the articles dealt with an stories within each of these years. That emotional or personal coverage of the is, from each year a sample devastating consequences of drugs in proportional to its size was selected. terms of human toll. Eleven editorials This sample was selected based on a (4.9 percent) were sampled, and only table of 4,000 random days for a 366- eight news articles dealt with drug day year published by Rosander (1977, abuse prevention (3.6percent). pp. 409-417). The format of presentation sets the pace by which an artic:le will be of 35 Moral Panics Cintron interest to the reader and from which developed within these. When the kr"1owledge will be gained. Since most three categories were collapsed of the articles were classified under the (national, state, and international), the general interest category, the readers state was the subject of 71.5 percent of might get a blurred picture of the drug the drug-related storie~, the nation was problem. The more general an article, the main topic in 32.7 percent of the the least prepared the readers might be articles, and international subjects were to understand the problem at hand. featured in 3.1 percent of the drug Ninety-five of the sampled news related articles. Once more the articles had drug-related stories that newspapers' tendency was toward pertained to the state of Florida (42.2 emphasizing the local and state drug percent). Sixty-nine articles (30.7 crisis, isolating or making no rercent) dealt specifically with the connection with the national or dl'lg crisis in the Miami -- Dade international scene. County -- area, and only 1.3 percent of The Webster's dictionary defines the articles made reference to regional a crisis as "a serious or decisive stage connections (other southern states); of things, or a turning point when an however, even when such a reference affair must soon terminate or suffer a was made, the emphasis was on the material change ... a situation whose state of Florida, either as the entry port outcome decides whether possible bad for drugs or levels of drug use in the consequences will follow" (1983, p. Dade county area. When the state and 432). In the sociological literature, local categories are collapsed, 72.9 social crises, such as the drug crisis, percent of the articles provided have generally been referred to as information on Florida's drug problem, "m oral panics." For example, according and very little information was offered to Cohen it is "a condition [that on the regional, national and becomes] ... defined as a threat to international implications of the drug societal values and ip~erests: Its nature problem. It is interesting to note that is presented in a stylized and the public was informed only about stereotypical fashion by the mass Florida's efforts and problems while media and ... experts pronounce their the articles tended to ignore the foreign diagnoses and solutions; ways of nation's own drug problems, for coping are evolved, or (more often) example, the economic impact of drug resorted to; the condition then trafficking for foreign countries. As a disappears, submerges or deteriorates result, we can say that the media and becomes more visible" (1973a, p. portrays a nativist orientation toward 9). the drug crisis. The development of a crisis or a Code categories were created for moral panic depends on a few factors. the subjects or topics that were It implies the notion of a process or discussed in the drug-related news stages, whereby the drug problem articles. Three code categories were becomes identified, develops and is created, and subcategories were reacted to. The research question, 36 Moral Panics Cintron "How was the crisis described?" dealt down" everything, from martinis to with the first stage, where a condition - pep-pills. - drugs -- became identified "as By 1973, it was reported that the ptoblematic, posing a danger or reason why drugs "had gotten out of containing the seeds of future hand" was due to the laxity of laws difficulties" (Becker, 1966). In the and the difficulty of enforcing these content analysis of the drug crisis, two laws. Law enforcement efforts were vocabularies ran through the sampled blamed for not getting at the man who articles: the first identified drugs as sells the drugs. While in 1971 heroin reaching epidemic proportions in wholesalers were described as the society, the second described Florida as biggest menace the country had ever the world's drug capital. faced, marijuana continued to be a Society: the drug epidemic. In major peril to the fabric of the October 1970, five articles described the American society in 1975, but by 1977 abuse of drugs as having reached the flow of cocaine crossing the United epidemic proportions in society. The States' borders had reached crisis articles reported that while man had proportions. By 1973, the "wave of used drugs for centuries, never before terror" that was sweeping the nation in history had society been faced with was being spread by non-addicts who the massive misuse of so wide a tried to hook young people and adults spectrum of drugs at all levels of its on drugs. Non-addicted drug population. But drugs, the articles wholesalers were described as cold, reported, were not the problem. Drugs calculating and ruthless criminals who, are a symptom of decay in society, and when arrested and convicted, should people use drugs as escape receive long, mandatory prison mechanisms. If it were not illicit sentences. But the size of the drugs, people would use something government's problem became else. It was reported that drug abuse apparent long before the wave of terror was not a benign disease; it was a swept the nation. In November 1970, cancer sapping vital energies and $30,000 could buy a half-ton of eroding precious human and social marijuana th~t could bring $150,000 on resources. the United States illicit marijuana Another reason for the epidemic market. For Florida, the problem was was a society well oriented toward the pressing. Law enforcement officials drug scene in one form or another. were watching Florida's overcrowded The drug scene orientation had its sky for airborne smugglers, while some roots on the impact television 2,200 miles of ragged, almost advertisements had on its audience, unprotected coastline, enticed the sea particularly for children, who spend traffic in cocaine. four to six hours a day watching the Florida: The world's drug capital. commercials. The effects of the By the end of the 1970s, it became commercials were reinforced when apparent that drug abuse had not only children watched their parents "put reached epidemic proportions in 37 Ivforal Panics Cintron society, but that some states had more had become the major center for at stake than others. From February organized crime in America. 1975 to February 1985, twenty-five According to the governor, other states sampled articles used in the headings had been more aggressive than Florida vocabularies describing Miami as the in their fight against drugs. That "world's drug capital." As a major port month the governor declared a war on of importation of illegal drugs, the drug smuggling. result was drugs from every direction, Law enforcement officers reacted in every way imaginable, on the water, to the declaration of war arguing that on the land, and as some headings narcotic smuggling had become described it throughout the years (in Miami's largest and most lucrative 1975, 1976, and 1977), drugs even ind ustry, as measured by the amount rained from heavens. of drugs law enforcement agents seized In March of 1979, the every month. Officials estimated that environment in South Florida was they were five years behind the drug described as a natural disaster area. smugglers, as far as financial resources, The same vocabulary ran again in May sophisticated equipment and of 1980 and June 1981, but this time manpower were concerned. For South Florida had become a law example, drug smugglers could afford enforcement disaster area. If the to buy a $200,000 "stash" house for federal government had a crime cash and walk away from it as law disaster law, it was argued that Miami enforcement officers were getting close would have probably gotten relief to making an arrest. However, because assista.nce. Drugs were blamed as the of bureaucratic impediments, number one contributing factor for the undercover agents ran into increase in violent crime in Dade appropriations problems just trying to County. It was also reported that by rent a one-hundred dollar a month 1970, more than eighty percent of the apartment to observe a drug operation. cocaine and marijuana entering the It was also reported that not since the country came through Florida. By repeal of prohibition had so much October 1979, South Florida was money been made in an illegal categorized as the combat zone in the enterprise. United States' drug war. By 1980, Between September 1979 and ninety percent of all the drug December 1983, Florida was fighting smuggling involved Florida either as the "cocaine war." Not only had destination point, as center of Florida developed a reputation as a international banking (money free-trade zone for drugs where laundering), or as site for making federal, state, and local officers made narcotics deals. It was reported that daily seizures of illegai narcotics, but drug smugglers had entrenched every also violent crimes attributable to dlUg aspect of life in the area. trafficking were common. The cocaine In March of 1970, Governor war was no longer made up of Graham publicly admitted that Florida skirmishes carried out in isolated areas 38 Moral Panics Cintron or in the hangouts of the Colombians movement. The second category, a who, by now, dominated the multi short period in 1973, called for the billion dollar cocaine smuggling involuntary commitment of addicted operations. The war was being carried persons. The last period, 1979 out in plain daylight on the streets throughout 1983, called for mandatory crowded with innocent people. minimum sentences for drug During this time, newspapers trafficking. During this period it was recounted the failures of the war on found that the control measures being drugs. In spite of the periodic implemented or being proposed at the multimillion-dollar drug seizures, state level ran parallel to the drugs South Florida continu~d to be flooded being abused in society. with high-grade cocaine. In August, Coordination of Resources and 1983 the South Florida market had so Efforts. In an effort to achieve some much cocaine that it seemed that cohesion on the drug abuse fight, the sellers were cutting their prices down. Dade County Metro Commission Drug enforcement agents were created the Drug Abuse Advisory reported as saying that drug Board in February of 1970. By county wholesalers were cutting their prices so ordinance the Ad visory Board was much that it looked like an end of the given the task of the coordination of all year clearance sale. By the end of the county programs dealing with drug year the flood had not leveled off in abuse, such as screening these South Florida, but had spread to other programs to avoid the duplication of areas of the Southeast. The spread led efforts. Additionally, the Board was to drug enforcement officers to suggest make recommendations on policy that while seizures had gone up matters after the programs had been through the year, the demand for evaluated. In October 1972, two years cocaine continued to be strong. after the Advisory Board had been in Ninety-three of the sampled news existence, it was reported that the articles discussed, to some extent, the Board had been unable to accomplish measures that needed to be taken to its mission. It had been almost control the drug epidemic. In the impossible to identify all of the local analysis procedure, it was found that drug abuse programs, let alone the control measures reported in the evaluate any of them. newspapers fell into two categories: While local control efforts were legal changes and the coordination of described as uncoordinated and resources and efforts. ineffective, federal law enforcement Legal changes. From 1971 efforts were reported as having through 1983, it was found that the positive results. After the legal changes being implemented or consolidation of several departments being proposed at the state level fell into the Drug Enforcement Agency, the into three categories. The first category federal government increased the runs from 1971 throughout 1975 and budget for drug abuse programs to could be called the decriminalization $419 million by 1973. From 1975 39 ------ Moral Panics Cintron through 1978, Florida representatives in the use of drugs in society -- must be Washington urged the federal interpreted as a prod uct of the government to augment the Drug vocabularies that motivate what to do Enforcement Agency's South Florida about the problem. This tradition forces by moving agents and implies that drug crises are not just equipment into the area from about the drug problem. Rather, it elsewhere in the nation. There were embraces the notion that the problem also plans to form a new anti is the product of a number of factors smuggling task force for Florida, that together have created the composed of the Drug Enforcement particular social context from which Agency, the Coast Guard, Customs, the the drug problem is defined and from State Department, the Justice which strategies of social control are Department, even the Air Force and generated. Navy patrols. By 1980 a Presidential Several findings are consistent Task Force had been created. Under wi th the research Ii terature. Over fifty the command of then Vice-President percent of the sampled articles were Bush the South Florida Task Force categorized as news of general interest, received wide publicity. It was the media reports only on certain reported the Task Force's anti-drug issues associated with the problem. campaign was responsible for the drop Since the format of presentation sets in crime rates in South Florida. Drug the pace by which the article will be of related homicides in Dade County interest to the reader and from which were down fifty percent, largely knowledge is gained, the readers might because the Task Force steered drug get a blurred picture of what the drug traffickers away from the area. By crisis is all about. 1983 drug seizures had increased fifty Another finding that is consistent two percent, and it was estimated that with the literature is that the articles' drug smugglers in South Florida had tendencies is toward isolating the lost drugs worth $4.4 billion because of problem. Over seventy percent of the the federal government efforts. sampled articles emphasized the drug Despite the above efforts, there problem form the Florida perspective, was no doubt that the law enforcement ignoring the national, regional, authorities were making only a small international and foreign connections dent in the illicit drug smuggling that exist. operations; estimates of drug seizures The state and local impact level of were never higher than ten percent. the articles is also reflected in the topics/subjects covered in the drug SUMMARY AND CONCLUSIONS related news articles. When the state, national, and international categories By using the periodic cycle of were collapsed, the drug problem in crisis approach, this work followed a Florida was the subject of over seventy sociological tradition whereby what percent of the articles. Since the becomes defined as a social problem -- newspapers' tendencies were toward 40 - Moral Panics Cintron an emphasis of the local/state drug employment in several government problem, it could be said that the agencies, serious concerns about drugs media portrays a nativist orientation and their consequences is on the toward the drug crisis. increase. However, the international, Two vocabularies motivated the deterrence and economic implications description of the drug problem as a of the drug problem continue to be crisis. The first identified drugs a ignored at the federal and state levels. reaching epidemic proportions in society; the second identified Florida as the world's drug capital. The drug crisis came BS a result of the impact and consequences that illegal substances had for society in general REFERENCES and Florida in particular. Another stage in the process or Anti-Drug Abuse Act, 21 U.S.CA., 1501 (1988). development of the drug problem is the recognition that the drug crisis Becker, H. S. (1963). Outsiders: Studies in the Sociologv of deviance.New York: The must be controlled. Proposals toward Free Press. the control of the drug crisis fell into two categories. The first category ran ___-,-._. (Ed.). (1966). Social problems: A from 1971 to 1983 and was identified modern approach. Introduction, (pp. as the need for legal reforms. The 1-31). New York: John Wiley & Sons, second group of proposals called for Inc. the coordination of resources and Blumberg, A. S. (1973). The politics of efforts. Local and state control efforts deviance: The case of drugs. TournaI of were described as uncoordinated and Drug Issues, 2. (2), 105-113. ineffective, while federal law enforcement efforts were reported as Botner, M. A. (1984). Media images and public having positive results. Federal efforts, attitudes toward crime and justice. In R. Surette (Ed.), Justice and the media (pp. in particular those of the South Florida 15-30). Springfield, Illinois: Charles C task force, were credited as being Thomas. responsible for a decrease in the criminal and drug trafficking rates in Braden, W. (1973). LSD and the press. In S. the area. Cohen and J. Young (Eds.), The It could be concluded that the manufacture of news (pp. 195-209). Beverly Hills, California: Sage vocabularies and motives that fuelled Publications. the drug crisis between 1970 and 1985 led the way to the control strategies Cohen, S. (1973a). Mods and rockers: The that were proposed after 1985. For inventory as manufactured news. In S. example, old stereotypical notions of Cohen and J. Young (Eds.), The the drug users are being challenge manufacture of news (pp. 226-241). Beverly Hills, California: Sage today; drug testing has been legally Publications. approved as a condition for 41 Moral Panics Cintron _____. (1973b). Sensitization: The case Issues, ~ (1), 37-52. of the mods and rockers. In S. Cohen and J. Young (Eds.), The manufacture of Lemert, E. M. (1951). Social pathology. New news (pp. 360-368). Beverly Hills, York: McGraw-Hill. California: Sage Publications. Lewis, R. G. (1984). The media and violent Davis, F. J. (1973). Crime news in Colorado and criminal behavior. In R. Surette newspapers. In S. Cohen and J. Young (ed.), Justice and the Media (pp. 51-69). (Eds.), The manufacture of news Springfield, Illinois: Charles C. Thomas (127-135). Beverly Hills, California: Sage Publisher. Publications. Linsky, A. S. (1973). Theories of behavior and Dickson, D. T. (1968). Bureaucracy and the image of the alcoholic in popular morality: An organizational perspective magazines, 1900-1966. In S. Cohen and on a moral crusade. Social Problems,1.§. J. Young (Eds.), The manufacture of (2), 143-156. news (pp. 146-155). Beverly Hills, California: Sage Publications. _____. (1973). Marijuana and the law: Organizational factors in the legislative Morgan, P. (1978a). The political use of moral process. lournal of Drug Issues, 2. (2), reform: California and federal dru"g 115-122. policy, 1910-1960. Unpublished doctoral dissertation, University of California, Duster, T. (1970). The legislation of morality: Santa Barbara. Law, drugs, and moral judgement. New York: The Free Press. ____. (1978b). The legislation of drug law: Economic crisis and social control. Fishman, M. (1978). Crime waves as ideology. Journal of Drug Issues, ~ (1), 53-62. Social Problems, 25 (5), 531-543. __-::-_. (1983). The political economy of __-::--::-_ .. (1984). Crime waves as drugs, and alcohol. TournaI of Drug ideology. In R. Surette (Ed.), Iustice and Issues, 13 (1), 1-7. the media (pp. 159-180). Springfield, Illinois: Charles C. Thomas Publisher. Musto, D. F. (1973). The American disease: Origins of narcotic control. New Haven: Gusfield, J. (1963). Symbolic crusade: Status Yale University Press. politics and the American temperance movement. Urbana, Illinois: University Norland, S. & Wright, J. (1984). Bureaucratic of Illinois Press. legitimacy and the drug menace: Notes on the Marijuana Tax Act. Deviant Hall, S., Critcher, c., Jefferson, T., Clarke, J., & Behavior, .2, 239-254. Roberts B. (1978). Policing the crisis: Mugging, the state, and law and order. Nunnally, J. C. (1973). Mental illness: What London, England: McMillian Publishing the media presents. In S. Cohen and J. Company. Young (Eds.), The manufacture of news (pp. 136-145). Beverly Hills, California: Helmer, J. (1975). Drugs and minority Sage Publications. oppression. New York: Seabury Press. Pearce, F. (1973). How to be immoral and ill, Himmelstein, J. (1978). Drug politics theory: pathetic and dangerous, all at the same Analysis and critique. Journal of Drug time: Mass media and the homosexual. 42 Moral Panics Cinh'on In S. Cohen and J. Young (Eds.), The ____. (1973b). The amplification of manufacture of news (pp. 284-301). drug use. In S. Cohen and J. Young Beverly Hills, Califomia: Sage (Eds.), The manufacture of new!i (pp. Publications. 350-359). Beverly rHUs, California: Sage Publica Hons. Phillips, D. (1973). The press and pop festivals: Stereotypes of youthful leisure. In S. Cohen and J. Young (Eds.), The manufacture of news (pp. 323-333). Beverly Hills, California: Sage Publications. Reasons, C. (1974). The politics of drugs: An inquiry in the Sociology of social problems. The Sociological Quarterly, lli 381-404. Reinarman, C. (1983). Constraint, autonomy, and state policy: Notes toward a theory of controls on consciousness alteration. TournaI of Drug Issues, 13 (1),9-30. Rosander, A. C. (1977). Case studies in sample design. New York: Marcel Decker, Inc. Schur, E. M. (1965). Crimes without victims: Deviant behavior and public policy. Englewood Cliffs, N.J.: Prentice-Hall. Sheley, J. F. and Ashkins, C. D. (1984). Crime, crime news, and crime views. In R. Surette (Ed.), Iustice and the media (pp. 12A-140). Springfield, Illinois: Charles C. Thomas Publisher. Skolnick, J. H. (1967). Coercion to virtue: The enforcement of morals. Southern California Law Review, £, 588-640. Terry, W. C. (1984). Crime and the news: Gatekeeping and beyond. In R. Surette (Ed.), Tustice and the media (pp. 31-50). Springfield, Illinois: Charles C. Thomas Publisher. Young, J. (1973a). The myth of the drug taker in the mass media. In S. Cohen and J. Young (Eds.), The manufacture of news (pp.314-322). Beverly Hills, California: Sage Publications. 43 Correlates of Police and Correctional Officer Performance Tracy L. Griffith, Ph.D. A growing number of police and measures. correctional agencies are utilizing In the current study, 590 road psychologists and objective personality patrol (police) and correctional assessment in the selection process of candidates completed both the MMPI officer candidates (Behrens, 1985). and IPI, and provided a variety of Although most professionals in the demographic information, during a field regard the addition of pre-hiring screening program utilizing psychological assessment favorably, a consulting psychologist. A large validation efforts have yielded mixed number of the candidates (n=375, 64%) results. Comprehensive reviews may were subsequently hired, allowing for be found in other sources (Spielberger, performance evaluations to be collected 1979; Burbeck & Furnham, 1985). and relationships analyzed. These reviews generally conclude that Subsequent performance measures of most inventories (ie., MMPI, CPt 16PF employment status and disciplinary and others) fail to indicate a significant infractions were gathered from official relationship between objective end-of-year evaluations. The length of personality assessment and job employment varied for each officer, performance, although some ranging from one (1) to 52 months. supporting results have been reported. Based on the performance evaluations, The Inwald Personality each officer was assigned to one of Inventory (IFI, 1980) was designed four performance groups for analyses. specifically to aid in the selection of Officers not hired were necessarily police and correctional officers, and has excluded from certain analyses due to shown promise in early validation a lack of follow-up data. efforts (Inwald & Shusman, 1984a, Results of the analyses indicated 1984b). However, little independent that the MMPI was not significantly validation research has focused on related to subsequent job performance long-term comparisons of the MMPI of either the road patrol or the (the most widely used instrument) and correctional officer samples. The IPI IFI using objective job performance was not found to be significantly This paper is from Dr. Griffith's dissertation by the same title, Dr. Harmon Burk, major professor. Dr. Griffith has degrees in criminal jllStice, criminologrJ and counseling psychology. He has worked with the Federal Dureau of Prisons and with a variety of in- and out-patient mental health agencies, providing individu!Zl, group and family treatment. His interests include pre-hiring screening, substance abuse and prevention, and general mental health treatment. Currently he is a Senior PSlJchologist with the Florida Department of Corrections, Jefferson C.l. Dr. Griffith extends special thanks to Dr. E.E. "Grif" Griffith, without whose help the current research would not have been possible. He also would like to thank the criminal justice agencies that provided subjects and follow-up infonnation. 44 Correlates of Performance Griffith related to performancE. for the road Court in a 1978 decision (Fabricatore, patrol sample, although it was found to in Spielberger, 1979). be significant in predicting Further, professional and performance for the correctional accrediting agencies have encouraged officers. Ten IPI scales were the addition of psychologists in the significantly different between the .~ pre-hiring decision, beginning in 1973 correctional officer groups. Stepwise when the National Advisory Discriminant Analysis revealed that Commission on Criminal Justice seven of the IPI scales were able to Standards and Goals included a correctly classify 53 percent of the psychological examination correctional officers. Implications for recommendation. More recently (1985), further research are discussed. the Commission on Accreditation for Law Enforcement Agencies published "Standards for Law Enforcement INTRODUCTION Agencies" requiring a psychological Behrens (1985) reports that a component for agencies seeking large number of criminal justice accreditation. agencies have been adding a Finally, widespread negative psychological screening component to publicity and the resultant loss of their pre-hiring evaluative process. In public trust have lead many agencies his nation-wide survey of state to include psychological testing in the agencies, 50 and 25 percent of police hiring process. At the time of this and correctional agencies were writing, national attention remains currently conducting psychological focused on the Los Angeles, California screenings of their officer candidates, case involving a brutal videotaped respectively. Eleven of the states beating of Rodney King, suffered at the mandated such screenings. Over one hands of four police officers. third of the agencies had implemented Allegations of excessive use of force their psychological component after are one of many ways in which public 1981, with most being introduced since trust is lost through acts of individual 1984. officers. It is clear that the use of Although the addition of this psychological screening is increasing in psychological component is generally spite of its expense, and is due in large recognized as a positive step by part to a series of legal cases in which psychologists and criminal justice agencies were held responsible for the administrators, the specific objective inappropriate behavior of their officers assessment instrument(s) to be used (i.e., Bonsignor v. City of New York, remains in question. Behrens notes 1981). This pOSition of holding that, "[t]he choice of objective municipalities responsible for the personality inventories to include in a wrongful actions of their employees pre-hiring psychological screening (under some circumstances) was program is typically at the discretion of upheld by the United States Supreme the agency and psychologist, with few 45 Correlates of Performance Griffith relating assessment and decisions of optimistic in the use of the IPI in the psychological suitability to subsequent area (Inwald & Sakales, 1982; Inwald & job performance are stressed (Burbeck Shusman, 1984a, 1984b; Shusman, & Furnham; Inwald, in More & Inwald & Landa, 1984), although Unsinger, 1987; Hiatt & Hargrave, clearly further research is needed to 1988). A major point focuses on the fact directly compare the effectiveness of that these instruments were simply not the MMPI and IPI using long-term and designed to be used in the area of objective on-the-job performance employee selection (Spielberger, 1979). criteria. Rather, they were typically developed to assist in the area of mental health PREVIOUS RESEARCH diagnosis, treatment and evaluation. Although a variety of objective A second deficiency in the personality instruments have been selection research focuses on the use of utilized in this area, the focus of the inconsistent and varied performance current stud y is on the MMPI and IF!. measures. Varying definitions of Reviews of related instruments may be "successful" versus "unsuccessful" found in other sources (Spielberger, officers have been used. It must be 1979; Burbeck & Furnham, 1985; noted that this criterion difficulty is Hargrave, 1984). based largely on the difficulty of Historically, the MIV1PI has been actually rating or evaluating officer the most widely used inventory in the performance. With no simple or area. Some of the earliest research consistent method of rating using the MMPI with police and / or performance, previous studies have correctional officer selection or used a variety of subjective and evaluation suggested that several scales objective performance criteria, leading were related to serious misconduct to methodologically inconsistent (Blum, 1964). Others (!v1arsh, 1962; studies and results that are difficult to Azen, Snibbe & Montgomery, 1973) interpret. reported modest, although statistically Related to this performance significant, relationships between discrepancy, a variety of settings have MMPI scales Ma (9) and De (2) and been used in validation efforts. Much automobile accidents. of the research has been cond ucted in Other MMPI research has found training academies or other short-term differences when comparing officers performance settings. Although such a who were employed for over three study minimizes the time and expense years with those not hired and those needed, it fails in providing long-term who resigned or were terminated (Saxe and direct "on-the-job" performance & Reiser, 1976). Training academy validation. Such studies have been performance was used as a criterion scarce. measure by Hargrave (reported in Finally, the Inwald Personality Yuille, ed., 1986) who reported that Inventory (1980) is a relatively young scales F, Hs (I), De (2), Pt (7) and Si (0) instrument. Early reports have been were able to differentiate between 46 Correlates of Performance Griffith cadets who completed the academy item true-false inventory designed, and those who did not. according to the author, to yield scores Further, Hiatt & Hargrave (1988) on a variety of personality, behavioral reported 1vfMPI scales Pa (6) and Ma and biographical characteristics of the (9) to be significantly higher among test taker. Designed specifically to officers who had more than one formal asse'3S criminal justice applicants, the unsatisfactory rating, had been IPI is composed of 25 original scales suspended or asked to resign or had and one validity scale. been involved in off-duty violations of A series of studies conducted by the law. Inwald and others have suggested that However, just as results the IPI may be more effective than the indicating that valid prediction may be MMPI in predicting job performance of found, studies failing to observe officer candidates. Inwald & Sakales significant differences between (1982) presented results showing that successful and unsuccessful officers are the performance of correctional officer also readily found in the literature. For candidates (using absences, latenesses example, Mandel (1970), Hess (1972), and formal disciplinary infractions as Balch (1977) Lester, et al., (1980), criterion) was significantly predicted by Schoenfeld, et al., (1980), Gettinger IPI scales TL (Trouble with Law), JD (1981) and others failed to find Gob Difficulties) and DG (Drugs). significant differences between their Next, Inwald & Shusman (1984a) groups of officers on any of the MMPI reported that IPI scales TL (Trouble scales. with Law), AS (Antisocial Attitudes), Contrasting results such as those AA (Absence Abuse) and RT (Rigid described above have lead several Type) were significantly higher for reviewers to conclude that male cor~"ectional officers who were relationships have been modest and terminated than those who were not. inconclusive, and that the practical Female correctional officers who were validity of the MMPI in predicting terminated were reported to differ success is limited (eg., Spielberger, from successful female officers on scale 1979; Burbeck & Furnham, 1985). LA (Lack of Assertiveness). A Overall, as suggested by Spielberger, it predictive component of the same is possible that some weak study indicated that the IPI was able to relationships may exist, although the correctly classify 72 and 83 percent of practical validity of the MMPI as a the officers (male and female, predictor of job performance for police respectively) on the termination officers is questionable. criterion; compared to 62 and 66 While the MMPI has been the percent predictions using the MMPI. inventory most widely used in the area The two inventories combined of criminal justice pre-hiring screening, provided the highest degree of IPI validation efforts have typically accuracy, correctly predicting 73 and 92 utilized both inventories in percent of the subjects. comparative studies. The IPI is a 310 Inwald & Shusman next 47 Correlates of Performance Griffith cond ucted a similar study (1984b) Dependent information was using the IPI and MMPI with police compiled by the employing criminal recruits in a training academy setting justice agencies and provided to the and reported a number of performance consulting psychologist and this writer diffe.rences using latenesses, absences, for use in the study. The dependent injuries on the job, formal disciplinary measures will be discussed further in a infractions, assignments to restricted later section. Briefly, however, duty, negative and positive reports, performance was collected on each overall supervisor ratings and subject annually, allowing for the supervisory ratings on "leadership assignment of subjects to one of four ability." Similar findings suggested the performance groups. Of the 590 IPI to be more effective than the MMPI candidates, 64 percent (375) were in predicting "success", with the most subsequently hired, allowing for the accuracy coming through the use of analysis of the relationships of both instruments. performance wi th the MMPI and IFI. Again, similar results were Because of the nature of such a design, reported by Shusman, Inwald & Landa the data currently being used was (1984) when male correctional officers compiled from a variety of time were analyzed on performance periods and sources. measures of retention/termination, Table] absences, latenesses and formal Subject Position and Gender disciplinary infractions. Road Patrol Corrections TOTAL THE CURRENT STUDY n % n % n % The current study used Male 200 33.9 304 51.5 504 85.4 independent data collected as an ongoing screening program from Female 16 2.7 70 11.9 86 14.6 October, 1985 to December, 1989. 590 Total 216 36.6 374 63.4 590 100 Road Patrol and Correctional officer applicants completed a pre-hiring Subjects. Both Deputy Sheriff psychological screening interview (also called Road Patrol) and during the 52 month period. The Correctional Officer applicants from screening evaluatio.:: ...... v~lsisted of the two county agencies in Southeastern completion of a questionnaire Florida, and Road Pah'ol officers from (providing a variety of demographic a nejghboring city were included in the information), completing both the current study. A total of 590 officer MMPI and IFI and meeting with the applicants were referred to the consulting psychologist. A brief consulting psychologist over a 52 intelligence measure was obtained month period. Each officer candidate during the interview as each applicant had completed several subsequent completed the Comprehension subtest stages of the hiring process and would of the Wechsler Adult Intelligence be processed further pending the Scale - Revised Edition (WAIS-R). psychological testing and a medica! 48 Correlates of Performance Griffith evaluation. Table 1 provides a (59%), typically in the police, breakdown of the subjects by position correctional or private security areas. and gender. Of the 590 subjects, 81 Thirty-nine percent had no previous percent (479) were caucasian, 16 experience in the criminal justice area, percent were black and two percent with thirteen subjects (2%) providing were of Hispanic origin. Three (.5%) no information. reported themselves as Asian/Oriental. Concerning the educational Regarding marital status, 189 levels of the subjects, 296 (50%) (32%) were single and never married. reported possessing a high school An additional 325 (55%) were either diploma or an equivalent. An married a first time or remarried. additional 256 (44%) reported attending Twenty (20) of the 325 (3% of the total) some post high school program, either were separated from their spouses at college or vocational-technical. Twenty the time of their evaluations. Seventy eight (5%) possessed a Bachelor's one (12%) were divorced from their degree, with three (.5%) having spouses and currently unmarried. completed a graduated degree. Four Spouses of two subjects WE're deceased (.7%) had not completed high school and three subjects failed to provide and three (.5%) failed to provide data. data on marital status. Additionally, a score on the Subjects ranged in age from 18 comprehension subtest from the to 65 years (See Table 2). A majority of Wechsler Ad ult Intelligence Scale the subjects had previously been (Revised edition) was obtained from employed in the criminal justice field each subject. This score is included as a raw score, with subjects ranging from 5 Table 2 to 15. Seventy two subjects (12%) had Subject Age no scores on this measure due to their Age (years) n % evaluations being completed prior to its implementation. See Table 3 for less than 21 52 9 Table 3 21-25 181 31 Subject IQ 26-30 114 19 WAIS-R n % 31-35 78 13 Comprehension Score 36-40 61 10 5 - 7 30 6 41-45 47 8 8 - 9 126 24 46-50 29 5 10 144 28 51-55 ]4 2 11 - 12 171 33 13 - 15 56-60 11 2 47 9 Total 61-65 3 1 518 100 Total 590 100 Subject IQ. 49 Correlates of Performance Griffith Length d Employment. As the were three validity scales (L, F & K), current study involved an on-going the ten major clinical scales and two screening program, length of additional scales (Iv1acAndrews employment varied by subject, ranging Alcoholism and a critical items score). from one (1) to 52 months. Table 4 . A large body of validity and reliability presents data on job performance research has accumulated on the MMPI groups (described more fully below) over the years (Greene, 1980; Buros, and the number of months since each 1982). Generally, the MMPI has been candidate had completed the found to possess satisfactory levels of evaluation process. Given the variance reliability and validity (Buros, 1982) in the length of emp19yment, it is and has been included in the current assumed that performance would be stud y due to its extensi ve use and significantly related to the number of stud y. As stated previous.ly, Behrens months employed; i.e., subjects who (1985) reports the MMPI as the most were f.!mployed for longer periods of widely used objective personality time are more likely to suffer some sort inventory in the criminal justice pre of misconduct. Conversely, subjects hiring screening field. Further, employed for brief periods of time (i.e., MMPI/IPI relationships are of interest less than one year) would be less likely in order to provide additional to suffer similar misconduct, due to the information on the IPI, discussed relative brevity of their employment. below. As may be seen in Table 4, The IPI is a 310 true-false item subjects employed less than 12 months inventory which yields scores on a typically performed without difficulty, variety of personality, behavioral and with only one (1) of 37 employed -- -- - Table 4 officers actually suffering any sort of Group Membership and Months Employed disciplinary infraction. It is likely that this exceptional performance during 2 3 4 Months Exemplary Acceptable Dropout Failure Total the first year is related to the brief period of employment and is not 1-11 36 o o 37 indicative of future performance in 10% general. For this reason, Exemplary 12-23 60 7 12 7 86 23% officers employed less than 12 months (n=36) were excluded from further 24-35 48 12 26 8 94 . 25% performance analyses. Simila:dy, Not 36-52 42 40 48 28 158 Hired officers were excluded from 42% further between-group comparisons, although descriptive statistics will be n 186 60 86 43 375 provided as information. % 50% 16% 23% 11% 100% Instruments. The MMPI is a 566 true-false item inventory yielding biographical characteristics of the test results on a variety of personality taker. Designed to "screen in" effective scales. Included in the current study candidates, the IPI is composed of 25 50 ------ Correlates of Performance Griffith candidates, the IPI is composed of 25 available. This lack of follow-up original scales and one validity scale. performance data is an inherent An additional score on the number of limitation in research of this type. For IPI critical items was included. The candidates subsequently employed, author describes the IPI as an agencies provided information on the inventory, "... designed specifically to current job status of each candidate aid law enforcement agencies in and the number and nature of formal selecting new officers who satisfy disciplinary infractions. This data was psychological fitness requirements compiled from annual performance (Inwald, 1982:1)." Further, the author evaluations and allowed the researcher states that the IP! attempts to to group each officer into one of four "... document combinations and patterns performance groupings, described of historical life events which studies below, with a fifth group not being suggest correlate significantly with hired. occupational failure in law enforcement (p.1).tI Additional information on the PERFORMANCE CRITERION development, reliability and validity GROUPINGS issues and norming of the IPI is (1) EXEMPLARY officers: Those provided in the IPI manual (Inwald, candidates who had been hired by the 1982). Despite its relative youth, it agency, had no formal diSCiplinary appears that the IPI possesses sufficient infractions during their employment reliability and validity levels to warrant and continued to be employed at the use and study in this area. Difficulties time of follow-up. of the IPI include its reliance on largely (2) ACCEPTABLE officers: face valid items, a reliance on self Those candidates who had been hired reported background information and by the agency, had one or more formal the fact that a number of scales are disciplinary infractions during their comprised of very few items. Few employment, yet had continued to be items per scale may lead to a large employed at the time of follow-up. variation of scores based on responses (3) DROPOUTS: Those to a few items, leading one to question candidates who had been hired and the validity. Finally, questionable intel' employed without formal disciplinary scale reliability issues have been raised infraction, yet had left the department and must be considered in evaluation voluntarily. These officers had left the .. efforts using the IFI. agency under positive conditions and Dep~ndent/Performance would be considered for re Criterion. As stated previously, a employment, should they desire. number of the officer candidates were (4) FAILURES: Those not offered positions, with a few offers candidates who had been hired by the being turned down by candidates for agency but were no longer employed various reasons. In these cases, no due to termination or forced follow-up performance measures were resignation. These officers would not 51 Correlates of Performance Griffith be considered for re-employment. hired by the referring agencies, to some (5) NOT HIRED: Those degree due to their assessment and / or candidates who either were not offered interview data. Obviously no follow-up positions or declined offers. performance criteria was available for Unlike previous studies in this this group, although MMPIIIPI scale area, the criterion grouping described correlations of this "not hired" group above allowed for comparisons of were included. Preliminary analyses of officers integrating both their a sample of this group (Griffith, performance and their employment unpublished thesis) confirmed that this status. While IPI results reported to "not hired" group does, in fact, date have focused on the prediction of significantly differ from other groups individual performance criterion (such on a number of scales. This "restriction as disciplinary infractions, sick days, of range" problem is an inherent termination/retention and others), limitation of the methodology of the typically as a dichotomy, the current current study and in similar research. study was designed to integrate the A second limitation of the study two major criterion measures to form lies in the fact that subjects experienced, four unique and independent different lengths of employment performance groups. (months) due to the fact that this was Statistical Analyses. Analyses an ongoing program. As a result, some were conducted using the Statistical subjects were employed up to 52 Package for the Social Sciences (SPSS, months with others as few as one Nie et al., 1970). Briefly, Pearson month. Such a design does not allow Bivariate correlations were conducted for the control of external variables as to understand MMPIIIPI scale inter an experimental design might. relationships. HoteIling's T results were Regarding the specific sample next obtained to determine any overall used, it is important to note that the multivariate relationships between the officer candidates were employed by inventories and job performance, with agencies in Southeast Florida; the two separate analyses for road patrol and counties are populated by correctional officers. Follow-up approximately 95,000 residents and the univariate ANOVA's were performed neighboring city has approximately on significant overall relationships. 49,000. Generalization of current Stepwise Discriminant Analyses (using results to officers and candidates in Rao's V criteria) were next conducted various sizes and locale should be done to determine the contributions of the with caution. significant univariate scales. Finally Finally, validity issues inherent classification accuracy of the derived in the instruments themselves and to functions were determined. their use in the area of pre Limitations of the Current employment screening are inherent Study:. A major limitation of the limitations to research of this type. As current study lies in the fact that a described previously, the IPI may be large number of subjects were not faulted due to the number of scales 52 Correlates of Performance Griffith comprised of only a few items, the Pearson Bivariate Correlation questionable inter-scale reliability and Results. Information on Pearson the reliance on self-report to largely Bivariate coefficients (r values) for all face valid items (particularly to MMPI and IPI scales used in the study biographical items). may be found in Appendix A. Critical values necessary for significance are RESULTS indicated. All candidates were included Following a preliminary in the bivariate correlations (n=590). examination of all variables to verify Multivariate Tests of assumptions of normality, linearity and Significance. Next, Hotelling's T scores homogeneity of variance, four major were calculated to determine overall questions were addressed using each of differences between the four groups of the officer groups (road patrol and subjects. Three analyses using the 15 corrections) with each of the MMPI variables, the 27 IPI variables inventories (MMPI and IPI). As a and the 42 combined MMPI and IPI result, four separate null hypotheses variables were analyzed with job were tested: performance, using separate samples of Hypothesis 1: There are no police and correctional officers. The six statistically significant relationships resulting omnibus measures are between MMPI scales and job Table 5 performance of road patrol officers. Hotelling's T Results Hypothesis 2: There are no statistically significant relationships Value Approx Hypoth Error Sign between IPI scales and job performance F d.f. dJ. of F of road patrol officers. Police Hypothesis 3: There are no MMPI .52 1.09 45 284 .33 statistically significant relationships IPI 1.10 1.12 81 248 .25 between MMPI scales and job performance of correctional officers. Combined 2.07 1.11 126 203 .25 Hypothesis 4: There are no Corrections statistically significant relationships MMPI .27 1.24 45 617 .14 between IPI scales and job performance IPI .71 1.69 81 of correctional officers. 581 <.001 Combined .98 1.40 126 540 <.001 A final question addressed in c.... the current study involved the utility in classification of the MMPI and IPI with presented in Table 5 and were used to the samples of road patrol and address the first four (4) hypotheses. correctional applicants, as follows: As may be seen in Table 5, there Hypothesis 5: There is no was no significant overall relationship statistically significant interaction observed between the 15 MMPI scales among significant variables with and job performance of the road patrol respect to job performance of a) road sample (.52, F=1.09, p=.33). Similarly, patrol and b) correctional officers. no significant overall differences were 53 Correlates of Performance Griffith job performance of the road patrol Table 6 officers (1.10, F=1.12. p=.25). Significant IPI Univariate Results: Correctional Officer~ Combined, the two inventories were not significant (2.07, F=1.11, p=.25). IPI Scale x Perfonnance Group mean Regarding the job performance (standard deviation) of the correctional sample, again no 1 2 3 4 overall group differences were Exemplary Acceptable Dropout Failure F p observed using the 15 MMPI variables (.27, F=1.24, p=.14). A significant Critical Items overall difference was observed 6.61 5.08 4.15 6.31 between the four groups of correctional (4.67) (5.86) (3.74) (4.80) 4.04 .00, officers using the 27 IPI scales, Guardedness (GD) however (.71, F=2.23, p<.OOl). The 42 7.68 8.64 8.84 10.45 combined MMPI and IPI variables (3.77) (3.94) (3.46) (4.08) 4.25 .00 ... were also significant (.98, F=1.40, Job Difficulty (JD) p<.OO1). Follow-up univariate F tests 2.78 3.11 2.12 3.07 (1.93 (2.05) (1.94) (1.69) 3.02 .031 were next conducted on the 27 IPI variables; significant «.05) differences Substance Abuse (SA) were observed on ten of the IPI scales 3.64 3.83 4.21 5.21 and are presented in Table 6. Follow- (2.18) (2.55) (2.59) (2.47) 3.31 .021 up univariate measures of the 15 Antisocial Attitudes (AS) MMPI variables (in the combined 42 4.97 5.81 5.57 7.28 variable model) were not significantly (2.63) (3.33) (3.04) (2.72) 4.78 .003 --- different between the groups of Hyperactivity (HP) correctional officers. Discriminant Analyses. Finally, 8.50 8.64 9.29 11.55 (4.10) (5.55) (4.90) (4.30) 3.38 .019 Hypothesis 5 was concerned with the interaction among the significant Rigid Typp. (RT) variables with respect to job 6.23 6.83 6.34 7.62 performance of a) police officers and b) (2.28) (2.42) (2.32) (2.26) 3.04 .030 correctional officers. As results on Unusual Experiences (UE) Hypotheses 1 and 2 were not 1.21 1.33 1.31 2.07 significant, the conclusion was that (1.25) (1.55) (1.32) (1.53) 2.74 .044 there was no significant variable Undue Suspiciousness (US) interaction between job performance of road patrol officers using either the 3.63 4.22 3.82 5.45 (2.50) MMPI or IPI scales. (2.80) (2.92) (3.15) 3.36 .020 Given that ten IPI scales were Family Conflicts (FC) found to be significantly different 1.83 2.89 2.65 2.62 between the groups of correctional (1.57) (2.62) (2.51) (1.95) 3.25 .023 officers (Hypothesis 4), these scales dJ. 3,221 were introduced into a reduced model 54 ------.------ Correlates of Performance Griffith discriminant analysis to determine the percent of the correctional cases, as contributions of the individual shown in Table 10. variables and their values, with results Table 8 presented in Table 7. Next, seven of Rao's V Canonical Discriminant Function these variables were introduced into a Coefficients stepwise discriminant analysis (using Rao's V criteria). These seven were IPI Variable Function 1 Function 2 chosen due to their contribution to Critical Items -.57843 .48925 optimal separation of the groups. IFI Guardedness (GD) .57801 .57475 scales RT (Rigid Type), UE (Unusual Job Difficulties (]D) -.18254 .69884 Experiences) and US (Undue Substance Abuse (SA) .50588 -.40378 1 Table 7 Reduced Model Discriminant Function Antisocial Attitudes (AS) .21236 .29807 Coefficients Hyperactivity (HP) .28576 .44570 IPI Variable Function 1 Function 2 Family Conflicts (FC) .24730 -.52766 Critical Items -59 -.48 Guardedness (GD) .58 -52 Job Difficulties (]D) -.19 -.65 DISCUSSION Substance Abuse (SA) .51 .38 As seen in the previous section, Antisocial Attitudes (AS) .13 -.14 significant between-group differences Hyperactivity (HP) .33 -.44 were observed only with the IPI and the correctional sample. The MMPI was Rigid Type (RT) -.06 .06 not found to significantly differentiate Unusual Experience (UE) .20 -.17 between performance groups of either Undue Suspiciousness (US) .05 -.13 correctional or road patrol officers. Family Conflicts (FC) .18 58 Likewise, significant findings with the Table 9 Rao's V Stepwise Discriminant Model Results Suspiciousness) were excluded from these analyses due to their lack of Function Eigen % Wilks dJ. p value Variance additive value. The reduced model is Lambda presented in Table 8. 1 .21 58.75 .72 21 <.001 Two significant reduced 2 .11 30.13 .87 12 .002 stepwise discriminant functions, 3 .21 11.12 .96 5 .13 presented in Table 9, indicate that the seven IFI variables contributed to two discriminant functions in their ability combined MMPI and IPI model were to classify the correctional officer found to be due to the contribution of sample. Finally, the derived the IFI variables. Significant differences discriminant functions were found to between the four correctional officer be successful in classifying 53.33 performance groups were found on IPI 55 Correlates' of Performance Griffith Difficulties OD), Substance Abuse (SA), of the variance observed. The functions Antisocial Attitudes (AS), shared information provided by the Hyperactivity (HP), Rigid Type (R1), Guardedness (GD) scale, which Unusual Experiences (UE), Undue weighed heavily in both functions and Suspiciousness (US), Family Conflicts is the major validity indicator. The (FC) and the critical items score. unique contributions are discussed Generally, the correctional officer below. Failure group (4) scored higher than Function 1 accounted for 58.75 the others on these scales. percent of the variance and was largely composed of the Substance Abuse (SA) Table 10 Discriminant Model Classification Results and critical items scores. Interpretation of this function is made difficult by the Actual Number Predicted Group n (%) presumed variety of the critical items, Group of Cases 1 2 3 4 and might best be described as a "Red Exemplary 92 62 5 13 12 Flag" model. Correctional (1) (67.4) (5.4) (14.1) (13.0) administrators would likely concur that Acceptable 36 8 12 7 9 officers should be substance free and (2) (22.2) (33.3) (19.4) (25.0) relatively free of behavior making up Dropout 68 26 4 26 12 critical items to red uce the likelihood (3) (38.2) (5.9) (38.2) (17.6) of manipulative or dangerous Failure 29 3 2 4 20 situations. (4) (10.3) (6.9) (13.8) (69.0) Function 2 contributed an Total n 225 additional 30.13 percent of the Percent of Grouped Cases Correctly Classified: 53.33% explained variance and was made up of IPI scales FC (Family Conflicts), HP (Hyperactivity), JD Gob Difficulties) It is important to note, however, and AS (Antisocial Attitudes). Together that differing responses on only one or these scales may represent "Conflict" two items may lead to significant and make compliance with the para between-group differences on scales military structure of the correctional FC, JD and UE, leaving validity and agency difficult. Hence these reliability questions open in their use correctional officers may have difficulty in employee selection. With this in complying with rules and regulations mind, current significant findings on or in their interpersonal relationships. IPI scales JD, AS, HP, RT and UE are Regarding the classification of consistent with significant findings the correctional officers using the reported in previous IPI validation discriminant functions derived from efforts. Of these, RT, UE and US were the seven IPI variables, it was found not found to contribute additively to that 53.33 percent of the officers were the stepwise discriminant functions. correctly placed in the four groups. The two derived functions were More importantly, however, the IPI found to be significant at the .05 level was found to be especially accurate in and were able to explain 88.88 percent classifying to groups 1 (Exemplary; 56 Correlates of Performance Griffith 67.4 percent) and 4 (Failure; 69 some previous studies suggest. percent). These are the problem free However, without any way of knowing officers of at least twelve months how the "Not Hired" group might have (Exemplary) and those who were performed, it is premature to discount terminated or forced to resign due to the inventories as useless. On the improper behavior or serious contrary, the instruments may disciplinary infractions. Identification successfully screen out a large of the Failure gruup is theoretically percentage of potential Failures. essential to minimizing the "high . Because the IFI was successful in visibility" negative events some officers predicting the correctional sample, it is may engage in. possible that the failure of the IFI with Given the theoretical make-up of the road patrol sample was due to a groups 2 (Acceptable) and 3 (Dropout), criterion deficiency, indicating the need it is not surprising that classification for unique criterion measures for the was less accurate (33.3 and 38.2 occupations. This would seem to percent, respectively). Recalling that indicate a validity problem in using the group 2 subjects had no serious MMPI with correctional officers, misconduct while employed, these although this question cannot be officers would likely pose little risk to sufficiently addressed unless a long the safety of the institution, staff or term study in which all officers are inmates. Likewise Dropouts (3) had hired, regardless of test results. resigned without disciplinary Given that the goal of most infractions and could be re-hired. Their criminal justice agencies is to typically reasons for leaving a seemingly "screen outll potential problem successful position are difficult to employees, the IPI in the current study predict, and would be interesting seems to be quite effective with this further study. correctional sample, although future Several issues are raised in the researchers may wish to consider current study. It is clear that the MMPI unique criterion measures for road was not successful in predicting job patrol officers. Further, the current performance of road patrol or stud y provides support for the need correctional officers. This could be due for long-term investigations (Le., over to an inherent validity problem one year) due to the homogeneity of (possibly due to a "curvilinear" officers employed less than one year. relativnship of both high and low scores being related to job performance) or a criterion problem. On the other hand, the IPI was successful in predicting performance of correctional, but not police, officers. Several explanations are again possible; perhaps the use of either inventory is invalid with road patrol officers, as 57 Correlates of Performance Griffith REFERENCES Burbeck, E., & Fumham, A. (1985). Police officer selection: A critical review of Adorno, T.W., Frankel-Brunswik, E., Levinson, the literature. TournaI of Police Science D.J., and Sanford, R.N. (1950). The and Administration, g 58-69. Authoritarian Personality. New York: Norton. Buras, O.K. (Ed.). (1982). The ninth mental measures yearbook. Highland Park, American Psychological Association (1980). N.J.: Gryphon Press. Principles for the validation and use of personnel selection procedures. Byrd, D.A. (1976). Impact of physical fitness on Washington, D.C.: Author. police performance. The Police Chief, .§ 30-32. Aylward, J. (1985). Psychological testing and police selection. TournaI of Police Campbell, D.T. & Stanley, J.e. (1963). Science and Administration, NJ3), Experimental and gunsi-experimeptal 201-210. designs for research. Chicago: Rand McNally. Azen, S.P., Snibbe, H.M., & Montgomery, H.R. (1973). A longitudinal predictive study Cattell, R.B., Eber, H.W., & Tatsouka, M.M. of success and performance of law (1949). The Sixteen Personality Factor '. enforcement officers. TournaI of Questionnaire. Illinois: Institute for Applied Psychology, 2.Z 190-192. Personality and Ability Testing, Inc. Balch, R.W. (1972). The police personality: Fact Cohen, B., & Chaiken, J. (1972). Police or Fiction? Journal of Criminal Law, background characteristics and Criminology, & Police Science, g performano:;:.Summary report, (No. 106-119. R-999-DOJ). New York: Rand Institute. Bartol, c.R. (1982). Psychological characteristics Commission on Accredition for Law of small town police officers. TournaI of Enforcement Agencies, Inc. (1985). Police Science and Administration, Standards for law enforcement 10(1),58-63. agencies. Fairfax, VA: Author. Behrens, G. (1985). Psychological screening Dawes, R.M., Faust, D., & Meehl, P.E. (1989). trends in the selection of police and Clinical versus actuarial judgment. corrections personnel. Criminal Justice Science, 243, 1668-1674. Digest, 1(7), 6-10. Dillman, E.G. (1963). Role-playing as a Blum, R.H. (1964). Police Selection. Springfield, technique in police selection. Public Ill.: Thomas. Personnel Review, M, 116-118-.-- Bonsignore v. City of New York. (1981). Edwards, A.L. (1953). The Edwards Personal 78-0240, New York Law TournaI, Sept. Preference Schedule. The PsycholOgical 14. Corporation. Bowen, Jr., D.N. (1984). Personality and Fabricatore, J.M. (1979). Pre-entry assessment demographic characteristics of road and training: PerfoJ.lnance evaluation deputies and correctional officers. of police officers. In e.D. Spielberger Unpublished doctoral dissertation, (Ed.), Police selection and evaluation. Florida Institute of Technology. New York: Hemisphere. 58 Correlates of Performance Griffith Froemel, E. (1979). Objective and subjective 250-258. measures of pollee officer perfonnance, pp. 87-111. In Charles Spielberger (Ed.), Hess, L. (1972). Police entry tests and their Police seiection and evaluation. predictability of score in police Washington, D.e.: Hemisphere Pub. academy and subsequent job perfonnance. Dissertation Abstracts Gettinger, S. (1981). Psychological testing of International, 5552, 33-B. recruits can screen out the real turkeys. Police Chief, March, 29-40. Hiatt, D., & Hargrave, G.E. (1988). Predicting job performance problems with Goldstein, L.S. (1974). Comment on "A psychological screening. TournaI of question of height" by R. 1. Hoobler & Police Science and Administration, J. A. McQueeny, in Police Chief, 41(11). ]6(2), 122-]25. Gottesman, J. (1969). Personality patterns of Hogan, R (1971). Personality characteristics of urban police applicants as measured by highly rated policemen. Personnel the MMPI. Hoboken, N.J.: Stevens Psychology, .M, 679-686. Institute of Technology. Hogan, R. & Kurtines, W. (1975). Gough, H.G. (1956). California Psychological Personological correlates of police Inventory. PaJo Alto, CA: Consulting effectiveness. Journal of Psychology, Psychologists, Inc. 21 289-29]. Greene, R.L. (1980). The MMPI: An interpretive Inwald, R (] 982). Psychological screening test manual. New York: Grune & Stratton. for police/ corrections. Criminal Iustice TournaI, l(]). Griffith, T.L. (1989). The use of the MMPI and IPI in police and correctional officer Inwald, R (] 984). Issues and guidelines for selection. Unpublished thesis, Florida mental health professionals conducting State University. pre-employment psychological screening programs in law enforcement Hagan, F.E. (1982). Research methods in agencies. Proceedings of the National criminal justice and criminology. New Symposium on Police Psychological York: Macmillan Pub. Services. Quantico, VA: FBI Academy. Hartman, B.J. (1987). Psychological screening of Inwald, R., Kratz, H., & Shusman, E. (1982). law enforcement candidates. American The Inwald Personality Inventory TournaI of Forensic PsycholoB)'t.1 5-10. Manual. New York: Hilson. Hathaway, S.R., & McKinley, J.e. (1940). A Inwald, R., & Sakales, S.R. (1982). A multiphasic personality schedule personality screening measure for law (Minnesota): 1. Construction of the enforcement officers. Paper presented schedule. TournaI of Psychology, ill at the American PsycholOgical 249-254. Association, Montreal, Canada. Hargrave, G. (1985). Using the MMPI and CPI Inwald, R, & Shusman, E. (1984a). The IPI and to screen law enforcement applicants: MMPI as predictors of academy A study of reliability and validity of perfonnance for police recruits. Journal clinicians' decisions. Tournai of Police of Police Science and Administration, Science and Administration, ]4(3), ]2(1). 59 Correlates of Performance Griffith Inwald, R., & Shusman, E. (1984b). Personality performance. Police, 58-64. and performance sex differences of law enforcement officer recruits. Journal of Melnicoe, W.J., & Mennig, J. (1978). Elements Police Science and Administration, of police supervisions. Encino, CA: 12(339-347). Glencoe Pub. Landy, F., Farr, J., Saal, F., & Freytag, W. Mills, R.B. (1976). Simulated stress in police (1976). Behaviorally anchored scales for recruit selection. Journal of Police rating the performance of police Science and Administration, 1" 179-186. officers. Journal of Applied Psychology, &1, 750-758. Mills, c., & Bohannon, W. (1980). Personality characteristics of effective state police Lefkowitz, J. (1975). Psychological attributes of officers. Journal of Applied policemen: A review of research and mchology, 22, 680-684. opinion. Journal of Social Issues, .2.1 3-26. Mills, M.C., & Stratton, J.G. (1982). MMPI and the prediction of job perfonnance. FBI Lester, D. (1979). Predictors of graduation from Law Enforcement Bulletin, 2110-15. a police training academy. Psychological Reports, 11, 362. Mitchell, J.V. (Ed.). (1985). The Ninth Mental Measurement Yearbook. University of Lester, D., Babcock, D., Cassisi, J., & Brunetta, Nebraska Press. M. (1980). Hiring despite the psychologist'S objections: An More, H.W" & Unsinger, P.c. (Eds.). (1987). evaluation of psychological evaluations Police managerial lise of psychology of police officers. Criminal Justice and and psycholOgists. Springield, Ill.: Behavior, Z 41-49. Thomas Pub. Levy, R J. (1967). Predicting police failures. National Advisory Commission on Criminal TournaI of Criminal Law, Criminology, Justice Standards and Goals. (1973). and Police Science,2§., 265-276. :rhe police. Washington, D.C.: U.S. Government Printing Office. Mandel, K. (1970). The predictive validity of on-the-job perfonnance from Nie, N.C., Hull, J., Jenkins, Steinbrenner, J., & recruitment selection information. Bent, D. (1975). Statistical Package for Unpublished doctoral dissertation, the Social Sciences. New York: University of Utah. McGraw HilI. Marsh, S.H. (1962). Validating the selection of Nowicki, S. (1966). A study of the personality deputy sheriffs. Public Personnel characteristics of successful policemen. Review, ~ 41-44. Police, ill 39-41. Matarazzo, J.D., Allen, B.V., Saslow, G., & O'Connor, G.W. (1962). Survey of selection Weins, A.N. (1973). Characteristics of methods. The Police Chief, 31(8). successful policemen and firemen applicants. TournaI of Applied Ostrov, E. (1986). Police/law enforcement and Psychology,.1§, 123-133. psychology. Behavioral Sciences and the Law, 1(4),353-370. McAllister, J.A. (1970). A study of the prediction and measurement of police Pendergrass, V.E. (1984). Psychological 60 Correlates of Performance Griffith assessment of police for entry-level Washington, D.C.: Hemisphere selection. Washington, D.e.: PubliShing. American Federation of Police. Spielberger, CD., Spaulding, H.e., & Vagg, Saxe, S. & Reiser, M. (1976). A comparison of P.R. (1981). Professional manual for the three police applicant groups using the Florida police standards psychological MMPI. Journal of Police Science and test battery. Tampa: Human Resources Administration, .1, 419-425. Institute. Schoenfeld, L.S., Kobos, J.L., & Phinney, L.R. Sweda, M.G. (1988). The Iowa Jaw enforcement (1980). Screening police applicants: A personnel study: Prediction of law study of reliability with the MMPI. enforcement job performance from Psychological Reports, 10 419-425. biographical and personality variables. Unpublished doctoral dissertation, Schultz, w.e. & Wood, M. (1957). Fundamental University of Iowa. Interpersonal Relations Orientation - Behavior. Palo Alto, CA: Consulting Terman, L.M. (1917). A trial of mental and Psychologists Press, Inc. pedagogical tests in a civil service examination for policemen and Shusman, E.}. (1987). A redundancy analysis firemen. TournaI of Applied for the Inwald Personality Inventory Psychology, .117-19. and the MMPI. Journal of Personality Assessment, 51(3),433-440. Thurstone, L.L. (1922). The intelligence of policemen. Journal of Personnel Shusman, E.}., Inwald, R.E., & Knatz, H. (1987). Research, .1 64-74. A cross validation study of police recruit performance Wechsler, D. (1981). Manual for the Wechsler predicted by the IPI and MMPI. Adult Intelligence Scale-Revised Journal of Police Science and Edition. San Antonio, TX: The Administration, 1.2, 162-169. Psychological Corporation, Harcomt Brace Jovanovich, Inc. Shusman, E.J., Inwald, R.E., & Landa, B. (1984). Correction officer job performance as Wright, B. (1988). Psychological evaluations as predicted by the IPI and MMPI. predictors of police recruit Criminal Justice and Behavior, 11 performance. Unpublished doctoral 309-329. dissertation, Florida State University. Speir, J.e. (1988). Police performance and Yuille, J.e. (Ed.). (1986). Police Selection and psychological testing: Application of Training: The Role of Psychology. the Megargee-Bohn MMPI-based .Boston: Martinus Nijhoff, Pub. classification system. Unpublished doctoral dissertation, Florida State University. Spielberger, e.D. (1968). State-Trait Anxiety Inventory. Palo Alto, CA: Consulting Psychologists Press, Inc. Spielberger, C.D. (Ed.). (1979). Police selection and evaluation: Issues and Techniques. 61 An Analysis of Differences in Role Perceptions Among Senior Patrol Officers and Campus Judicial Officers in Selected Public Higher Education Institutions in the United States Diane Marie Waryold, Ed.D. INTRODUCTION the nation, many would argue that assuming a posture of differential law Campus police/public safety enforcement on a campus places a officers, like officers of the law in any limit on the officers' authority and state, county, or municipal agency, creates role ambituity and a resultant exercise a certain amount of discretion feelin of frustration (Bordner & in deciding what course of action is to Petersen, 1983; Powell, 1981; Rousch, be followed in enforcing the law 1981). (Banton, 1964; Reiss 1971; Teasley, The purpose of this study was to 1973). The disposition of offenders determine whether differences exist in however, may differ in a campus the attitudes of senior patrol officers setting because of the unique nature of and campus judicial officers regarding the population served, and the interests the role of campus police/public and expectations of the institution safety. The role perceptions of senior (Meadows, 1982). According to patrol officers and campus judicial Bordner and Petersen (1983), officers at selected public higher "university administrators assume a education institutions of the United proprietary responsibility for State were assessed through the use of institutional members and, thus, a survey instrument containing encourage a discretionary, non-punitive questions and several case scenarios. approach to policing" (p.209). For This study was directed by the example, policies have been established following research questions: at most institutions for handHng 1. What are the perceptions of student conduct issues as an alternative senior patrol officers and campus to formal criminal charges. Although judicial officers regarding the role of this practice is very common in campus police / public safety relative to institutions of higher learning across the service versus law enforcement This paper is from Dr. Waryold's dissertation by the same title, Joseph c. Beckham, Ph.D, J.D., major professor. Dr. Waryold has served as Coordinator of Residential Judicial Programs/Security at the University of Florida since 1988. She received her Bachelor of Science degree in education from the Stat.! University College of New York at Cortland, her Master's of Education degree frOIl! the University oj Florida and her Doctorate in Educational Leadership from the Florida State University. Dr. Waryold has presented programs in Judicial Affairs, victims rights and campus police role perceptions throughout the countn). She is a charter member of the Association of Student Judicial Affairs and is the editor of the Southern Association of College Student Affairs newsletter. 62 Role Perceptions Waryold orientation? (p. 104). These criteria were 2. What are the demographic advantageous to this study because correlates of variations in perceptions? members have established their 3. Do significant differences commitment to upholding the mission exist in the perceptions of senior patrol of the institution and they constitute officers and campus judicial officers one identifiable group. The researcher relative to the service versus law secured the endorsement of IACLEA enforcement function? and the Association of Student Judicial Affairs (ASJ A) to encourage a high METHOD participation rate. Seventy-nine institutions were included under these Population guidelines. These institutions enroll Four-year public research and approximately 1,800,600 students in 39 doctoral-granting institutions of higher states. learning with an enrollment of 10,000 The sampling procedure or more students was the focus of the included the use of a survey study. Institutions of this classification questionnaire. A perceptual survey became the unit of study for this was administrated to three senior research because of their relatively patrol officers and the campus judicial homogenous nature as classified by the officer on each campus. r olice Carnegie Commission (1976). organizations in general are described Institutions of this size generally as bureaucratic with a formalized employ their own campus law hierarchy. The senior patrol officer enforcement department and also have position was chosen as the position to an established position responsible for study because this position assumes student conduct. This study was the responsibility of making the limited to those institutions that were decisions regarding the means of members of the International adjudication of incidents in the field. Association of Campus Law The captain, lieutenant, sergeant or Enforcement Administrators (IACLEA) corporal positions are the positions during the calendar year of 1990-199l. charged with this responsibility as they According to the bylaws of IACLEA, are generally assumed by more membership in this association seasoned officers. endorses the promotion of The researcher received 260 "professional ideals and standards so returned questionnaires from the total as to better serve the educational of 316 questionnaires mailed, yielding objectives of institutions of higher an overall rate of return of 82.3 education (p. 104)." The criteria for percent. Of this total, 184 of the membership also stipulates the returned questionnaires were from member institution must "provide its senior patrol officers, and 76 were from own functional, administrative campus judicial officel~s. Three of the organization to care for its needs returned questionnaires from senior involving law enforcement" on campus patrol officers were not included in the 63 ------ Role Perceptions Waryold data analysis because they were of questionnaire design. The first returned after the established cut-off group is a group of colleagues, the data. The usable respt:)TIses therefore second is a group of potential "users" consisted of a total of 257 of the data, and the third group questionnaires, of which 181 were from includes individuals drawn from the- senior patrol officers, and 76 were from population under study. campus judicial officers. Reliability was established using Dillman's Total Design Method the Cronbach Alpha formula was used as a guide for survey (MacMillan & Schumacher, 1984). construction and mail processing Questions were assigned a label of law (Dillman, 1978). The survey consisted enforcement or service for analysis of a modified version of a purposes. In applying the Cronbach questionnaire that was developed and Alpha test, internal consistency is validated by James Allen Telb (1980) measured by the use of a split-half and later used by Robert J. Meadows correlation in which the questionnaire (1982) and William David Nichols is divided into two subsets (Borg and (1985). The researcher requested the Gall, 1983). A logical division of the .. permission of the original author of the questionnaire for the purpose of this instrument to utilize the instrument study was a division between the role """lith modification to fit the i1eeds of criterion (questions 1-J 6) and the this particular study. Since the behavior criterion (questions 17-26) modification of the instrument was with their respective labels as noted drastic, efforts were made to validate above. The role items revealed a it. Content validity was established by: coefficient of internal consistency of .72 (1) asking the central staff of the and the behavior items a .79. Division of Housing at the University of Florida to critique it, (2) asking Analysis and Results members of the executive board of Research question one. In ASJA to critique it, and (3) pretesting examining the frequencies, means, the instrument with three senior patrol standard deviations, and t values for officers from the University of Florida all of the questions in the survey, the. Police Department. The purpose of data indicated that senior patrol these three steps was to review for officers and campus judicial officers correct spelling, grammar, readability appear to vary in their perception of and content validity. All feedback was the role of campus police/public safety carefull y considered and changes were relative to the law enforcement versus made accordingly. According to service orientation. When tetests were Babbie, "any members of the applied to the 26 questions, 21 of the population under study or any persons 26 questions yielded significant values similar to that population may serve as indicating possible differences between pretest subjects" (1970:207). Dillman the means of the two groups for (1978), notes that three types of people individual questions. should be utilized in the pre-test step ,Research question two. The 64 Role Perceptions Waryold demographic variables of size of items, indicating that the younger institution, rank, age, gender, length of professional favored the law time employed as a campus law enforcement orientation. A weak but enforcement officer, prior employment negative relationship existed bep.N'een experience, total number of years in time employed in the profession and law enforcement, previous experience role and behavior criterion items. As in military, training, and educational time employed in the profession level were analyzed for senior patrol increased in a negative direction, the officers. The same demographic law enforcement orientation decreased: variables were analyzed for campus A,trong positive relationship existed judicial officers with title substituted between education and role criterion for rank, length of time employed as a items, and a weak but negative judicial officer substituted for campus relationship existed between the law enforcement officer, total number behavior cd terion items. As the level of years in student affairs substituted of education increased, so did the for years in law enforcement, and the mean of the group towards the service training variable was excluded. orientation. Several analyses were undertaken to accommodate the differing types of Table 1 data. The variables of age, size of Regression Analysis for Demographic Variables institution, length of time employed as , a campus law enforcement Variable r r Signif. officer / campus judicial officer, total Age R .18761 .03520 .0030* years in law enforcement/student B .00611 .00004 .9237 affairs, and level of education were considered to be ordinal or interval Size of R .06933 .00489 .2736 Institution data. Therefore, the application of a B .03265 .00107 .6096 regression analysis was appropriate to Time R -.15871 .02519 .0J.32* determine if there was a relationship Employed between these variables and the B .18174 .03303 .0046* respondents preference towards the Total Years R .06972 .00486 .2780 law enforcement versus service B .07062 .00499 .2719 orientation. The results are indicated Education R .37493 in Table 1. Significance was based on .14057 .0000* the Critical Values for Pearson's B -.18681 .03490 .0033* Product-Moment Correlation (r) Table R = role criterion, B = behavior criterion, = significant (Best & Kahn, 1986, p.342) at an alpha at 0.05 level=.05 for n=200. Age, time employed as a campus law enforcement/ campus judicial officer, The variables of gender, and education showed a degree of previous experience in the interaction. Age showed a weak but military, prior employment experience, positive relationship to role criterion and attendance at the police academy 65 Role Perceptions Waryold constituted categorical or nominal data. the law enforcement orientation for Therefore, the application of the Chi behavior items. Square test was appropriate. The data is 'reported in Table 2. Gender was the only variable that showed a CONCLUSIONS AND relationship to the law enforcement IMPLICA TIONS FOR PRACTICE versus security orientation. Females tended to favor the service orientation The modern college and of the behavior criterion items. university community demands multiple serves from a professional Table 2 Chi-Square Values Analysis for Demographic campus police force. The individual Variables officer who perceives his/her role to be more law enforcement oriented than Variable d.f. n "/: service oriented, may have a difficult Gender R 1 249 .64670 time in meeting the demands of this B 1 249 5.95338* diversified community. The interpersonal relationships between Military Experience R 1 246 .45143 officers and the community may be B 1 246 3.70854 jeopardized when officers appear to Law Enforcement R 1 177 .02742 have an inconsistent approach to their Experience role, as perceived by other constituents B 1 175 .69760 of the community. The following SA Experience R 1 71 .0000 guidelines have emerged from this B 1 73 .12094 study: Police Academy R 1 176 2.73199 1. Beginning with the initial inquiry by potential applicants for a B 1 174 .00998 position with a department of campus R = role criterion, B = behavior criterion, • significant police / public safety, it may be valuable for p < 0.05 for the institution to provide accurate information which clearly defines Research questiofL three. T-tests departmental and university were computed to determine if philosophy and the uniqueness of the significant differences existed between population to be served. the two sample means of senior patrol 2. In the recruitment process, it officers and campus judicial officers as . may be valuable for departments to groups. The data indicated that a actively recruit a diverse staff of significant difference existed between campus police/public safety officers. the two sample means of senior patrol Diversity should be reflected in the officers and campus judicial officers for area of gender, and should accurately both role and behavior criterion items. represent the population served. Senior patrol officers supported the law 3. Departments may want to enforcement orientation for role items, actively recruit applicants with and campus judicial officers favored advanced degrees, or applicants who 66 Role Perceptions Waryold have the desire to further their REFERENCES education. They may also want to be supportive of the individual who Babbie, E.R. (1973). Survey research methods. actively furthers his/her education Belmont, CA: Wadsworth Publishing while employed. Company. 4. If the orientation of the Babbie, E.R. (1979). The practice of social institution is towards service, it may be research. Belmont, CA: Wadsworth valuable for younger recruits who are Publishing Company. fresh out of the police academy to receive additional specialized training Banton, M. (1964). The police in the in courses pertaining to human community. New York: Basic. development and student development Best, J.W., & Kahn, J.Y. (1986). Research in theory. education. Englewood Cliffs, NJ: 5. It may be of value for Pren tice-Hall. departments to conduct an annual review of their operational procedures Borg, W.R., & Gall, M.D. (1983). Educational including all policies, -programs, and research: An introduction. New York: Longman. practices to ensure that they are effectively serving the university Bordner, D.C., & Petersen, D.M. (1983). community. Campus policing: The nature of 6. Officers and administrators university police work. New York: are encouraged to jointly attend University Press of America. periodic in-service training programs to Carnegie Council on Policy Studies in Higher provide for opportunities for Education. (1976). A classification of continuous dialogue and involvement. institutions of higher education, 7. If departments of campus revised edition. Washington, D.C.: police / public safety tend to be image Carnegie Council on Policy Studies in conscious, efforts to recognize and Higher Education. reward campus police / public safety Dillman, D.A. (1978). Mail and telephone departments for the job they do may surveys: The total design method. be valuable. Raising morale may have New York: John Wiley and Sons. a positive effect on overall university relations, and the retention, and International Association of Campus Law recruitment of officers. Enforcement Administrators 8. It appears that basic training Membership Directory (1990-1991). Available from IACLEA, 638 Prospect in the police academy and other I!venue, Hartford, CT 06105. specialized programs may need to be reviewed. If training is misdirected, McMillan, J .H., & Schumacher, S. (1984). then it should be amended to fit the Research in education: A conceptual specialized needs of a university introduction. Boston: Little, Brown setting. and Co. Meadows, R.J. (1982). A study on the relationship between demographic 67 Role Perce'ptions Waryold characteristics of college and university safety officers and their orientation toward service or law enforcement. Unpublished doctoral dissertation, Pepperdine University. Nichols, W.o. (1985). Perceptions of the role of: campus safety departments by public safety directors and their immediate supervisors (Doctoral dissertation, University of Alabama, 1985). Dissertation Abstracts International, 1§., 1848A. Powell, J.W. (1981). Campus security and law enforcement. Boston: Butterworth. Reiss, A.J. (1971). The police and the public. New Haven: Yale Univer<;l.y Press. Rousch, R.E (1981). Philosophy of public safety within the college environment. Campus Law Enforcement TournaI, Marchi April, 30-31. Teasley, C.E. (1973). Police role perceptions. Unpublished doctoral dissertation, University of Georgia, ] 973. Telb, J.A. (1980). The relationships between personal characteristics of campus security officers and their role orientation (Doctoral dissertation, University of Toledo, 1980). Dissertation Abstracts International, 41, 1337A. 68 An Analysis of Factors Influencing Probation Outcome Kathryn D. Morgan, Ph.D THE PROBLEM' probationers under supervision is high. Tippman (1976), who studied 790 male For much of correctional history, felons, found the failure rate to be 40%. there has been reliance upon various Rogers (1981), with a sample of 1,014 programs designed to serve as probationers, concluded that 60% of alternatives to incarceration. Probation those placed under probation has been and continues to be used as supervision fail. Many studies have an alternative to corrections. Recent cited such factors as age, gender, statistics published by the Department employment, educational attainment of Justice (1990) indicate that in and marital stability as related to December, 1990, there were probation success or failure. Roundtree approximately three million offenders (1984) and Tippman (1976) found prior.. under federal and state probation criminal record to be the most supervision with 83% of the total under significant predictor of probation active supervision. Petersilia (1985) outcome. Other studies have identified estimates that between 60% and 80% of probation supervision variables all convicted criminals are sentenced to includ ing the length of sentence and probation. Despite the proliferation of types of conditions assessed as being intermediate punishments such as predictors of probation outcome. intensive probation, home confinement and electronic surveillance, probation The Present Study has survived as a sentencing The present study was alternative. prompted by two major concerns: the Although used extensively, continuing and widespread use of probation has been the center of much probation as a sentencing alternative, controversy and criticism for its failure and the paucity of research devoted to to rehabilitate and deter offenders the examination of probation placed under the supervision of effectiveness. probation departments. Many studies The focus of this study has been have shown that the failure rate for to identify factors related to the success This paper is from Dr. Morgan's dissertation by the same title, Dr. Gordon P. Waldo, Major Professor. Kathryn Diane Morgan iF a native of TeX11S. She attended TeX11S Woman's University where she received B.S. and M.A.. degree>; in SociologrJ. She was awarded a Ph.D in Criminology from the Florida State University in 1991. Dr. Morgan has been a Sociology Instructor at Lamar University in Beaumont, Texas; a Felony Adllit Probation Officer in Jefferson County, Texas; and Assistant Professor of Criminal Justice at the UniversitN of Tennessee at Chattanooga. She is presently Assistant Professor of Criminal Justice at the University of Alabama at Birmingham, where her research interests include probation and parole, and race and crime. She is married to Rev. Richard L. Morgan and has one daughter, Moya Elyse. 69 Probation Outcome Morgan or failure of probationers while under attachment to conventional society and probation supervision. Efforts were more sensitivity to the opinions of made to determine if significant othel's, females are not as likely as relationships exist between: (1) males to deviate from conventional offender characteristics and probation norms. Empirical evidence has also outcome; (2) prior criminal record and shown that females perform better probation outcome; or (3) probation under probation supervision and have sentence characteristics a.l1d probation more successful outcomes. Davis outcome. Offender characteristic (1964), Wisconsin Department of variables refer to those Corrections (1972), and Frease (1964) sociodemographic characteristics of all concluded that females are more probationers including age, gender, successful under probation supervision marital status, employment status, than males. Norland and Munn (1984) hourly wage, educational status, and found that female probationers have race. Prior criminal record variables lower violation and revocation rates include prior misdemeanor and felony with a majority of violations by female convictions, prior juvenile or adult offend ers being technical or the probation sentences, and previous violation of some administrative rule institutional commitments. Probation (failure to report, failure to pay sentence variables include conviction assessed fees). offense, Jength of probation term, There is much empirical support supervision level at the time of for the association of age with termination and behavioral and probation outcome. With age as a financial conditions of probation. predictor of probation outcome, most Therefore, this study attempted to studies have concluded that older identify variables most significantly offenders are often more successful related to the success or failure of those under probation supervision than under probation supervision. younger ones (England, 1955; Caldwell, 1951; Davis, 1964; Kusuda, 1976; Discussion of Variables Used in the McCarthy and Langworthy 1988). Study Petersilia (1985) found that Offender Characteristic probationers between the ages of 26 Variables: the sociodemographic and 30 are less likely to recidivate and variables gender, age, marital status, older offenders are less likely to be educational attainment and race. reconvicted for property offenses. Only With the variable, gender, one study (Menard and Covey, 1983) empirical evidence indicates that found that age was not significantly females have lower violation and related to probation outcome. revocation rates and are more Numerous studies have shown successful under probation supervision. the existence of a relationship between Hirschi (1969) contends that females marital status and probation outcome. are more attached and less likely to These studies have shown a stable deviate. Because of the greater marriage as being a significant factor 70 Probation Outcome Morgan associated with probation success concluded that being in a racial group (Landis, 1969; Frease, 1964; Kusuda, other than "white" significantly 1976; Missouri Department of increases the chances of probation and Corrections, 1976) Landis (1969) post-probation failure as measured by attributed the probation success of revocations and reconvictions. married offenders to the stability and There is much evidence to "settling down" effect resulting from support the importance of employment family life. as a predictor of probation outcome Educational attainment has also (Kusuda, 1976; Caldwell, 1951; Landis, been shown by research to be 1969; Wisconsin Department of significantly related to probation Corrections, 1972; and Petersilia, 1985). outcome. The empirical evidence Still other evidence (Caldwell; 1951; offered by .Frease, 1964, Landis, 1969 England, 1955) refers to low Petersilia, 1985, and the Missouri socioeconomic status as being related Department of Corrections suggests to probation failure. that the higher the educational Prior Criminal History: includes attainment, the lower the possibility of prior probation, prior incareration, andJ failure. Lower educational attainment the number of felony commitments. is likely to result in fewer "stakes in Many studies and prediction conformity". instruments have used prior criminal In traditional criminological history as a significant factor in theory, the variable race has always predicting probation outcome. In their been associated with serious persistent study of a group of Wisconsin adult criminal conduct. However, the data probationers, Gillin and Hill (1940) seem to be ambigious in explaining the found that prior arrests and juvenile relationship between race and criminal history were significantly performance under probation related to probation failure. Other supervision. Some studies conclude research has also concluded that there that there is no major difference exists a significant relationship between between the races regarding probation probation failure and prior criminal performance; while some indicate that history (Caldwell, 1955; England, 1955; there are significant racial differences Davis, 1955; Frease, 1964; Landis, 1969; especially between blacks and other Irish, 1972; Irish, 1977; Missouri groups. Other studies infer that being Division of Probation and Parole, 1976). non-black or non-minority improves Petersilia (1987) concluded that the the chances of probation success (Irish, greater the number of convictions, the 1977). Petersilia (1985) found that when higher the probability of failure. Scott predicting probation success or failure, and Carey (1983) suggested that prior the variable race contributed nothing to incarcerations were likely to result in the model. Other research has also probation failure; while prior found no significant racial differences convictions showed no significant between probation successes and relationship to probation outcome. failures (Landis, 1969). Davis (1964) Bartell (1977) found that prior arrests 71 Probation Outcome Morgan ======.. ======for burglary serve as one of the best failure. As probation sentence length predictors of probation failure. increased up to five years, probabilities Probation Sentence Variables: of success increased with each year. includes conviction offense and After five years, the success rate sentence length. Past research indicates steadily declined. (Frease, 1964; that they are significant in explaining Landis, 1969; Renner, 1978). probation outcome. Prior research has shown a relationship between property offending and recidivism; property Research Hypotheses offenders are more likely to recidivate. The present study has attempted Likewise, research examining probation to identify factors influencing outcome has indicated higher probation outcome. Relationships probabilities of failure for property between offender characteristics and offenders. In comparing property and probation outcome; prior criminal violent offenders who were on history and probation outcome; and probation, Holt and Beckett (1982) probation sentence characteristics and conduded that failure among probation outcome. nonviolent property offenders is more common. They suggested that whereas • Hypothesis 1: Females will a violent act represents a transitory experience more success on probation psychological state and is usually a one than males. time occurrence, persistent nonviolent • Hypothesis 2: As age increases, the property offending is an indicator of a likelihood of probation success propensity for criminality. Petersilia increases. (1987) concluded that property • Hypothesis 3: Married offenders will offenders are more likely to fail. have more success on probation than Property offenders tend to be arrested unmarried probationers. more quickly than violent offenders - • Hypothesis 4: The higher the usually within five months of being educational attainment, the greater the placed under probation supervision. likelihood of probation success. Cuniff (1986) reported that those • Hypothesis 5: Whites are more likely probationers convicted of burglary, a to be successful than blacks. property offense, had the highest • Hypothesis 6: Being employed rearrest rate with the new offense often increases the likelihood of probation being for another burglary. Bartell success. (1977) concluded that one of the best • Hypothesis 7: The higher the hourly predictors of probation failure is the wage, the greater the likelihood of number of prior arrests for burglary. probation success. The length of the probation • Hypothesis 8: The greater the sentence influences probation outcome. number of felony commitments, the Some studies have reported that the lower the probation success rate. longer the probation sentence, the • Hypothesis 9: Probationers with prior greater the likelihood of probation probation experience will be less likely 72 Probation Outcome Morgan to succeed on probation. those sentenced for new offenses are • Hypothesis 10: Probationers with counted as probation failures. prior institutional commitments are less A random sample of 350 cases likely to succeed on probation. was drawn from the population. • Hypothesis 11: Probationers Because of lost files; files whose last convicted of property offenses are names differed from those on the more likely to experience probation reference cards; cases closed because of failure than offenders convicted of death; and cases actually terminated other crimes. prior to 1980, 266 (N=266; 75% of the • Hypothesis 12: The longer the original sample) probationers make up probation sentence the greater the the final sample used for the study. A likelihood of probation failure. majority of the 25% of the cases not included in the final sample were lost or misplaced case files. RESEARCH METHODS AND PROCEDURES Data Collection The test instrument was divided Study Population and Sample into five sections: Personal Data; Prior The population for the study Criminal History; Probation Sentence; consisted of convicted adult felony Probation Adjustment; and Probation probationers in the Southeast Region of Outcome. For the 266 subjects selected Tennessee whose probation cases were for the study, closed case files were terminated in the period January 1, used to secure the information needed 1980 through December 31, 1989. This for the study. Information was Southeast Region encompasses six transferred from the case files to the counties: Hamilton, Bledsoe, data-gathering instrument. There was Sequatchie, Bradley, Grundy and no direct contact with the probationers. Marion. The study population consisted of 2508 adult felony probationers who FINDINGS had been given probation terms of two years or more and whose cases ha.d Bivariate Analysis been terminated in the specified time On the bivariate level, Chi period by completion of the probation square was used to determine if term, revocation or administrative relationships between the dependent closure. Administrative closure occurs variable, probation outcome, which when the supervising officer terminates was measured by success or failure, supervision of the case because the and selected independent variables probationer has died, absconded, or were statistically significant at the .05 has been sentenced for another offense probability level (see Table 1). within the jurisdiction or in another Offender Characteristics. With jurisdiction. In the present study, death regards to offender characteristics, data cases are excluded; but absconders and indicated that significant relationships 73 Probation Outcome Morgan indicate that those who were Table 1 Chi-Square Values for Independent Variables unemployed (49%), failed to complete and the Dependent Variable high school (35%), and had a lower PROBATION OUTCOME income (26%) had higher failure rates. These findings support the fact that Independent Chi-Square Statistical, Variables Values Significance (p<.05,\ when individuals have more to lose and a greater stake in conformity, the Gender 5.37 d.f.=l Sighificilnt likelihood of deviation diminishes. Age 6.19 d.f.=3 Not Significant Prior Criminal Record. The data Marital Status 13.83 d.f.=l Significant showed that prior criminal history significantly affects probation outcome. Education 6.45 d.f.=3 Significant The data showed that probationers Race 2.85 d.f.=l Not SigrJficant with three or more prior felony Employment 43.46 d.f.=l Significant commitments were more likely to fail than those with none or 1-2 prior Hourly Wage 47.51 d.f.=2 Significant felonies (58% versus 20% and 37% Prior Felonies 24.70 d.f.=2 Significant respectively: Chi-Square = 24.7, df=2, Prior 15.70 d.f.=l Significant p=.05». Each felony commitment Probation increases the likelihood of fail ure Prior 17.05 d.f.=l Significant (None - 20%; 1-2 - 37%). Those who Institutional had prior juvenile or adult probation Conviction 5.47 d.f.=2 Not Significant sentences (50%, Chi-Square = 15.70, Offense df=1, p=.05) and those with previous Length of 3.59 d.f.=l Significant incarceration terms (47%, Chi-Square = Sentence 17.05; df=l, p=.05) are more likely to experience probation failure. existed between the variables marital Probation Sentence. Regarding status, gender, employment status, variables related to the probation hourly wage, educational attainment sentence -- conviction offense and and the dependent variable, probation sentence length -- the data supported outcome. Not surprisingly, males were onI y the hypothesis which predicted a more likely to experience probation relationship between sentence length failure than females (35% versus 19% and probation outcome. Probationers respectively: Chi-Square = 5.37, df=1, receiving probation terms of more than p=.05). Those probationers who were five years (> 5) had slightly higher not married which included the single, failure rates than those given sentences never married; and the divorced had of five years or less (39% versus 27%: higher failure rates than those who Chi-Square = 3.59, df=l, p=.05). were married (37% versus 15%: Chi Although the relationship between Square = 13.83, df=1, p=.05) indicating conviction offense and probation that stability and commitment to outcome was not statistically family are possible related to significant, the data did go in the conformity. Likewise, the data also predicted direction. Property offenders 74 Probation Outcome Morgan had higher failure rates (35%) than probabilities. In order to determine the violent offenders (28%) or public order probability that probation outcome will criminal offenders (6%). be failure, i.e., Pr(Y = 1), another equation will be used: Multivariate Analysis: The Logit Model Pr(Y = 1) = e Because the dependent variable 1 + eX probation outcome is dichotomous and consists of only two categories - where e is the base of the natural success and failure, the use of Ordinary logarithm with a value of 2.718. It is Least Squares regression is raised to the power of the odds inappropriate because it would yield coefficient as represented by x. The misleading estimates. Therefore, the result of the computation represents logit model is used in the present the probability of failure given a study to determine the relationship particular independent variable. As a between probation outcome and a result, the interpretation of the group of independent variables. coefficients for logistic regression is TI1e fulliogit model contained similar to that for linear regression. all of the variables used in the bivariate Although less straightforward, the analysis. With the exception of age, interpretation states that "with a unit's race and conviction offense, all of the increase or decrease in the value of x, variables in the full model were the probability of failure increases or statistically significant in the bivariate decreases by the computed value of analysis. All of the relationships Pr(Y = 1)". In interpretation, the including those which were not computed value bears the same statistically significant went in the positive or negative sign as the logit predicted direction. coefficient (x) (Aldrich and Nelson, T-ratios were used to determine 1984:32-35). the statistical significance of the logit To determine the fit of the coefficients. The computed T -ratio was model, logit and probit do not have the compared with the critical value of 1.96 R-Square as OLS regression. Howeverp (.05 probability level) to determine the there are measures "in the spirit of the significance of the coefficient. The logit R-Square" which can be considered. A coefficients represe,nt the odds of an measure proposed by Aldrich and individual failing under probation Nelson is a Psuedo R-Square which -is supervision given a particular computed by taking the Chi square independent variable. This is statistic that measures the overall fit of represented by the equation: the model and dividing it by the Chi Square statistic added to the sample odds(failure) = P size. 1 - P These values are indirect 75 ------ Probation Outcome Morgan Pseudo R-Square = C/ (C + N), status, and hourly wage were still the only variables in the model whose where coefficients were statistically C is the Chi-square statistic significant. Marital status remains N is the sample size slightly below the significance level of 1.96. Correlation coefficients showed The value ranges from 0 to 1 that employment stahlS and hourly with zero showing evidence of poor fit wage (.91) and prior felonies and prior and one indicating good fit. Aldrich probation (.54) were highly correlated. and Nelson (1984:57) caution that care Results also showed that the fit of the must be exercised when using this model has not significantly improved statistic because it is not universally (p-suedo R2 = .48). used or accepted. The advantages of The third model is presented in this statistic are that it is easily Table 4. It was estimated with six computed from data given in the variables: sex, employment status, computer output and it ranges behveen education, marital status, prior felonies zero and one "approaching zero as the and conviction offense. Hourly wage . quality of the fit diminishes and 1 as it and prior probation were deleted improves" (Aldrich and because of being highly correlated with Nelson:1984:57). other predictors in the model. Hourly Table 2 presents the results of wage was the less significant of the the first lugit model containing twelve two correlated variables (employment variables. Coefficients for the variables status and hourly wage). In the sex, hourly wage and employment bivariate analysis, prior probation had status are significant leading to the a weaker association with the conclusion that those who are fema12, dependent variable; therefore, it was full-time employed and making dropped from the model. Prior minimum wage or above have a better institutional commitrnent was deleted chance of succeeding on probation. because of its consistently low level of These variables are the only ones tha.c significance in all of the models. are significant at the .05 level (1.96). Results in Model 3 showed that sex, TIle coefficients for variables marital marital status, employment status, and status and probation sentence length prior felonies were statistically are just below this significance level. significanti conviction offense was The Pseudo R2 of .48 indicates that the slightly below the significance level. fit of the model to the data is not very The fit of the model has improved as good. the Pseudo R-Square approaches a The second mod el was value of 1 (.5311). estimated with ten variables, excluding In the final model, the most age and race whose coefficients statistically insignificant variables were continued to be statistically deleted from the model, education insignificant. Table 3 gives the results (1.70311) Ch'ld probation sentence length which indic(lte that sex, employment (1.7403]). The model was reestimated 76 ~ Table 2 Table 3 g.>-t Legit Estimates for Felony Probationers Terminated from Logit Estimates for Felony Probationers Terminated from ~ OJ Probation Between 1980 and 1989 Probation Between 1980 and 1989 ::t 0 FULL MOl)EL MODEL 2 ::I 0 Variable Coefficient Standard T-Ratio P(Y=1) Variable Coefficient Stnndard T-Ratio P(Y=l) C (P/1 - P) Error (P/1 - P) Error rl-n 0 Sex -.53871 .21615 -2.49235 .36&5 Sex -.53525 .21444 S -2.49603 .3693 ro Race .01603 .19383 .08269 .5040 Employment Status 1.98791 .49804 3.99146 .8795 Age -.10734 .10477 -1.02456 .4732 Hourly Wage -.70470 .31654 -2.22627 .3308 Employment Status 2.0455:1 .50281 4.06819 .8854 Education -.24815 .14570 -1.70311 .4383 Hourly Wage -.74161 .31892 2.32539 .3226 Marital Status .40713 .21452 1.89786 .6003 Education -.24931 .14833 -1.68078 .4380 Prior Probation .25687 .24291 1.05749 .5633 Marital Status .40003 .21751 1.83912 .5986 Prior Prior ~ II Institutional .09019 .22699 .39732 .5224 Probation .24588 .24432 1.00640 .5611 Prior Felonies .17241 .17172 1.00404 .5430 Prior I nsti tu tional .09701 .22910 .42342 .5242 Conviction Offense -.2065i .13930 -1.48249 .4485 Prior Felonies .14242 .17531 .81238 .5355 Probation Conviction Sentence .29102 .18326 1.58796 .5722 Offense -.16001 .14428 -1.10901 .4600 Constant 2.56371 .66606 Probation Sentence .35132 .18925 1.85618 .5869 PEARSON GOODNESS-OF-FIT 252.945 Length PSEUDO R-Squared .4874 Constant 2.71782 .69289 PEARSON GOODNESS-OF-FIT 250.805 PSEUDO R-Squared .4852 ~ 0 >-t OQ OJ ::I Probation Outcome Morgan Table 4 Logit Estimates for Felony Probationers Terminated from Probation Between 1980 and 1989 MODEL 3 Variable Coefficient Standard T-Ratio P(Y:l) (P/1 - P) Error Sex -.46556 .20911 2.22639 .3856 Education -.24862 .14207 -1.74997 .4381 Marital Status .46363 .20988 2.20907 .6138 Employment Status .97057 .18615 5.21383 .7252 Prior Felonies .34402 .10944 3.14355 .5851 Conviction Offense -.24535 .13326 1.84118 .4389 Constant 2.89769 .55723 PEARSON GOODNESS-OF-FIT 284.407 PSEUDO R-Sguared .5311 Table 5 Logit Estimates for Felony Probationers Tenninated from Probation Between 1980 and 1989 FINAL MODEL Coefficient Standard T-Ratio P(Y=I) Variable (P /1 - P) Error Sex -.50575 .20748 -2.43764 .3762 Marital Status .44278 .20756 2.13324 .6089 Employment Status 1.00143 .18450 5.42771 .7313 Prior Felonies .35595 .10773 3.30398 .5880 Conviction Offense -.28476 .12996 -2.19120 .4292 Constant 2.6]343 .53237 PEARSON GOODNESS-OF-FIT 294.194 PSEUDO R~Sgllan:d .5396 78 Probation Outcome Morgan with the variables sex, marital status, increases the probability of probation employment status, prior felonies and failure by .5880 when the effects of conviction offense. Results in Table 5 other variables are held ,:;:onstant in the show that all of the variables are model. The final variable in the model statistically significant at the .05 is the conviction offense which resulted significance level. According to the in probation. When the effects of other final model, the best predictors of variables are held constant in the probation success are sex, marital model, a unit change in the conviction status, employment status, prior felony offense decreases the probability of commitments, and the offense for probation failure by 4292. A which the individual was placed on probationer placed on probation for a probation. Goodness-of-fit statistics violent offense or a public order crime indicate that the fit of the model has has a better chance of sUlceeding on improved. probation than the property offender. Holding other variables constant According to the final model, in the model, a unit change in the probationers who are married, variable sex decreases the probability employed, with no prior felonies, and ' of probation failure, {P(Y=l)}, by .3762. are on probation for offenses other Therefore, the probability of probation than property offenses have a g,Teater success is greater for females than probability of succeeding on probation. males. This finding is consistent with previous research. A unit change in the Conclusions variable marital status increases the The sample under study probability of probation failure by consisted primarily of males under the .6089 when other things are held age of thirty who were either single or constant in the model. Increased married and lived with family probabilities of probation failure occur members. The largest percentage of the when the marital status is single, probationers were unemployed; and divorced, or separated. This finding. over half did not have a high school has also been consistent with previous education. Regarding prior criminal research as well as the findings of the history, over half of the probationers present study. The variable had at least one misdemeanor employment status has consistently commitment; while less than fifty been correlated to probation outcome percent had previous felony both in this ~tudy and in previou~, commitments. Small percentages of studies. Holding the effects vf other probatic:mers in the sample had variables constant, the probability of previous probation supervision and probr>tion failure increases by .7313 institutional commitments. with a unit change in work status. If Over seventy percent of the the employment status is less than full probationers had been placed on time employment, tht probability of probation for property offenses; and a failure increases. A unit change in the majority were given sentences of five variable prior frlony commitments years or less. Very few were given 79 Probation Outcome Morgan behavioral conditions; and when successful; and the higher the behavioral conditions were assessed, educational attainment the greater the probationers were usually required to likelihood of probation success. The attend AA meetings, receive data supported these hypotheses and mandatory drug treatment or do showed them to be significantly related community service. to probation outcome. Hypotheses Regarding probation adjustment, . which predicted significant many of the probationers in the sample relationships between race and age and had no misdemeanor or felony arrests probation outcome failed to be or court referrals while under supported even though the probation supervision, When relationships were in the predicted probationers were arrested, it was direction. Consist~nt with previous usually for a felony offense and court research, the data were supportive of referrals were most often the result of a hypotheses which predicted that full new arrest violation. time employment and adequate wages A majority of the probationers in would result in probation success. the sample completed probation and Probationers who were employed ., were considered probation successes; a either full-time or part-time were more smaller percentage were considered likely to be successful on probation failures due to revocation, absconding than those who were unemployed. or being sentenced for another offense. Regarding prior criminal history, Felony rearrest was the primary reason hypotheses which predicted significant for revocation with the majority of relationships between prior crimbal those revoked being sent to prison. The history variables and probation data also showed that when outcome were supported as statistically probationers did commit a new significant by the data. The data not offense.. :t was usually a property only supported the hypotheses but also offense. demonstrated that prior felonies, prior probation and prior institutional Confirmation of Hypotheses commitments were significantly related The research hypotheses to probation failure for they dimini$h predicted relationships between the likelihood of probation success. probation outcome and selected The hypothesis which predicted independent variables. For offender that property offenders wou!d be Jess characteristics, the data supported likely to experience probation success predicted significant relationships failed to find support in the data. Even between probation outcome and the though the data showed that a larger variables gender, marital status, and percentage of property offenders fail educational attainment. The data when compared to other offense supported hypotheses which predicted categories, the relationship was not that females would have greater statistically significant. The data did success on probation than males; support the prediction of a significant married probationers would be more relationship between probation 80 I Probation Outcome Morgan sentence length and probation which should be addressed by outcome. probation research in the future are: In all of the hypotheses the 1. To what extent does relationships went in the predicted probation IIwiden the netll of sodal direction even when they were not control? Blomberg (1980:582), who sees statistically significant. net-widening as increasing the Selection of IIBestll Predictors of numbers of people subject to some Probation Outcome. On the kind of social control makes this multivariate level, there was an observation: attempt to determine the IIbestll set of predictors of the dependent variable. "The net-widening trend can be associated with Using logit regression, several models a number of tum-of-the-century correctional were estimated. The full model refonns, including parole, probation and the juvenile court. Since the early nineteenth contained the twelve variables used in century, there have been a series of refonns the bivariate analysis. In each considered as alternatives to imprisonment. succeeding model, variables were Thus, it is not surprising that the history of deleted for two reasons: coefficients fell u.s. correctional reform has consistently below the significance level (1.96); and resulted in net-widening whereby an correlation coefficients indicated that increasing proportion of the population has become subject to some form of institutional or they were highly correlated with other community-based control." variables in the model. The final model contained five variables which ware all 2. What can probation do or not statistically significant and cited as the do, and for whom does probation work IIbestll predictors of probation outcome ll ll best? Is there a IItype of offender who in the present study. The "best is likely to benefit most from probation predictors were sex, marital status, supervision? Although this present work status, prior felonies, and study identifies some characteristics conviction offense. Goodness-of-fit associated with success or failure, more statistics used in the present study research needs to be done in this area. showed that this final model had 3. \Vhat are unanticipated explanatory and predictive ability. outcomes of probation supervision? Discussion. The purpose of this There are other research study was to determine factors that questions that can best be answered influence probation outcome. A major through evaluation research. Babbie underlying theme of this research and (1964) defines evaluation research as past res~arch has been whether applied research when the plimary probati6'n as a nonincarcerative purpose is to evaluate the impact of alternative is fulfilling its objectives. some social intervention. Lawrence Recognizing the limitations of past Bennett (1978) argues that evaluations probation research, there is a of correctional programs are very poor. tremendous need for more empirical Most program evaluations do not go investigation of probation as a beyond statistical headcounting, correctional alternative. Questions 81 Probation Outcome Morgan describing activities or offering job referral services represent some subjective opinions of the value of and alternatives to addressing this major contributions made by the program problem. usually without supporting evidence. Evaluative research can provide understanding of program outcomes, unanticipated consequences, programs and services and client types. REFERENCES Ultimately, evaluation research can help to "tell the story" of the program. Aldrich, John and Forrest Nelson. Linear Practice. In the present study, it Probability, Logit and Prob'i'tModels. seems that probationers who have little Beverly Hills: Sage Publications, 1984. commitment to conformity are those Babbie, Earl. The Practice of Social Research. who are unemployed. Perceptions of New York: Wadsworth Publishing "low stakes in conformity" lead Company, 1986. individuals to conclude that there is little to lose. Without employment, Bartell, Ted and Winfree Thomas. "Recidivist conditional freedom is often Impacts of Differential Sentencing Practices for Felony Offenders". insufficient to ensure conformity. With Criminology. 15:387-396, 1977. the perception that there is little to lose, many offenders are free to deviate Bennett, Lawrence. "Evaluation of Probation from society's norms and u.ltimately . Parole and Community Corrections" from the rules and regulations Probation Parole and Community governing their conditional freedom. Criminology. Carter, Glaser and Wilkins (eds.). New York: John Wiley A major finding of the study and Sons, 1978. shows that inadequate employment and unemployment to be major Blomberg, Thomas. "Widening the Net: An impediments to achieving successful Anamoly in the Evaluation of probation adjustment and ultimately Diversion Programs". Handbook of successful probation outcome. For Crimina11ustice Evaluation. Malcolm Klein and Katherine Teilman (eds.). many of the probationers who were Beverly Hills: Sage Publications, 1980. revoked and even a substantial number of those completing probation Caldwell, Morris "A Review of a Different successfully, employment was a major Type of Probation Tennination Data problem. Although unemployment is a Made in Alabama". Federal Probation. major issue, inadequate employment 15 aune) 1951. also has implications for probation California Bureau of Statistics. "Superior Court supervision. The study has shown that Probation and/or Jail Sample". 1977. a majority of those under supervision have been convicted of property Comptroller General of the United States. State offenses. The data have further shown and Countv Probation: Systems in- that most of the new arrests were for Crisis. Washington, DC: US property offenses. Job counseling and Government Printing Office, 1976. 82 Probation Outcome Morgan Cuniff, Mark. A Sentencing Postscript: Felony Landis, Judson, James Mercer and Carole Probationers Under Supervision in the Wolff. "Success and Failure of Adult Community. 1986. Probationers in California". TournaI of Research in Crime and Delinguency. 6 Davis, George. "A Study of Adult Probation January 1969. Violation Rates by Means of the Cohort Approach". Journal of Criminal Law, McCarthy, Belinda and Robert Langworthy. Criminology and Police Science. 55 Older Offenders: Perspectives in March, 1964. Criminology an~ Criminal Justice. New York: Praeger, Inc. 1988. England, Ralph. "A Study of PostProbation Recidivism Among Five Hundred McGinnis, Robert and Kenneth Klocksiem. Federal Offenders". Federal Probation. Probation and Employment: A Report 19,1955. to the Bergen County New Jersey Probation Department. 1976. Frease, Dean. "Factors Related to Probation Ou tcome". Board of Prison Tenns and New Jersey Administrative Office of the Parole. State of Washington. April, Courts. Adult Probation in New Iersey: 1964. A Studv of Recidivism and a Determination of the Predictive Gillin, J.L. and R. L. Hill. "Success and Failure Utilities of a Risk Assessment Model. of Adult Probationers in Wisconsin". Trenton, 1980. Journal of Criminal Law and Criminology. 30:807-829, 1940. Norland, Stephen. and Priscilla Munn. "Being Troublesome: Women on Probation". Hirschi, Travis. Causes of Delinquency. Criminal Justice and Behavior. 11 Berkeley, CA: University of California (2):115-135, 1984. Press, 1969. Petersilia, Joan. GrBnting Felons Probation: Hirschi, Travis and Michael Gottfredson. "Age Public Risks and Alternatives. and the Explanation of Crime". California: The Rand Corporation, American Tournai of Sociologv. 89 1985. (3):552-584, 1983. ____. "Probation and Felony Offenders". Holt, Norman and Gerald Beckett. "Prediction Federal Probation. 51 (2), 1987. of Violent Versus Nonviolent Recidivism from Prior Violent and "Probation in Missouri July I, 1968 to June 30, Nonviolent Criminality". TournaI of 1970: Characteristics, Performance and Abnormal Psychology. 91 (3):178-182, Criminal Reinvolvement". Jefferson 1982. County, MO: Division of Probation and Parole. Irish, James. "Probation and Its Effect on Recidivism: An Evaluative Study of Renner, John. "The Adult Probationer in Probation in Nassau County New Ontario". Ministry of Correctional York". Mineola NY: Nassau County Services, 1978. Probation Department, 1977. Rogers, Sally. Factors Related to Recidivism Kusuda, Paul. "1974 Probation and Parole Among Adult Probationers in Ontario. Terminations". Madison Wisconsin: Toronto, Ontario, 1981. Department of Corrections. 1974. 83 Probation Outcome Morgan Roundtree, George, Dan Edwards and Jack Parker. "A Study of Personal Characteristics Related to Recidivism." Journal of Offender Counseling Services and Rehabilitr.ltion. 8 (3) Spring, 1984. Scott, Menard and Herbert Carey. "Community Alternatives in Colorado". Criminal Justice and Behavior. 10(1):93-108, 1983. Tippman, Donald. "Probation as a Treatment Alternative for Criminal Offender". Unpublished Manuscript, 1976. 84 Did I Cause This and Can I Change It: The Role of Perceived Responsibility in Treatment of Adolescent Offenders Cynthia Cupit Swenson, Ph.D. INTRODUCTION treatment of antisocial adolescents in Juvenile crime represents particular are generally pessimistic astronomical costs to society regarding success rates. To encourage considering damage to property and more productive adult lives for persons, as well as treatment of adolescent offenders and to decrease delinquents and their victims. In the emotional and physical costs to addition to criminal behavior, society, the search for successful adolescent offenders tend to exhibit a intervention strategies must continue. higher prevalence of learning Among the many treatment programs disabilities, hyperactivity, attention utilized with juvenile offenders, deficit disorders, mental retardation, problem areas such as anger control, and substance abuse problems; and sexual identity problems, substance they are less skilled in problem-solving abuse, and institutional adjustment than the general adolescent population have been addressed (Davis & (Farrington, 1983; Kazdin, 1985; Larson, Leitenberg, 1987; Feindler, Ecton, 1988; Quay & Werry, 1986; West, 1982). Kingsley, & Dubey, 1986; Friedman, Even though billions of dollars Glickman, & Morrissey, 1986; Phillips, are spent annually in the United States Phillips, Fixsen, & Wolf, 1971; to operate the juvenile justice system, Swenson, Butler, Kennedy & Baum, costs to society continue when 1989). However, alterable factors that delinquents reach adulthood due to the predict which treatment will be propensity of this population for successful with which individuals unemployment, accidents, divorce, and remain unknown. Intellectual levet welfare services (Caspi, Elder, & Bern, chronicity, and severity of the problem 1987). have all been related to treatment Presently, published studies on outcome (Quay & Werry, 1986); This paper is from 01'. Swenson's dissertation, "Perceived Control as a Predictor Of Treatment Outcome with Chronic Adolescent Offenders." Wallace Kennedy, Ph.D., her major professor, is co-author of the paper. Cynthia Cupit Swenson completed her Masters degree in Clinical Psychology at Northeast Louisiana University in 1981. She practiced psychology in the school system in Louisiana for five years. In 1991, she earned her Ph.D. in Clinical PsychologrJ at the Florida State University, where she worked with severe juvenile offenders for three years. Dr. Swenson completed an internship in child clinical/pediatric psychologrJ at the Medical University of South Carolina, where she cllrrently holds the position of post-doctoral fellow in clinical psychology. The authors gratefully acknowledge the administrative staff of the Dozier School, Roy McKay, Danny Pate and Billy Baxter, as well as colleagues who provided support, Congressman Pete Peterson, Joseph G. Ball1ll and Marshall E. Swenson. They note their appreciation for Pat Kennedy's editorial assistance and Ron Polland's assistance in statistical analysis, and thank the FSU lIndergraduates who assisted in data collection and entry. 85 --- ~~------ Did I Cause This Swenson however, these factors are not alterable. oriented toward building or One alterable factor which strengthening self-concept. With appears promising in predicting aggressive children, on the other hand, individuals who will potentially benefit strategies emphasizing personal from treatment is perceived control. responsibility for behavior might be Weisz (1983, 1986) defined perceived more appropriate. control as a joint function of ou tcome Given the empirical data contingency (perceived contingency) indicating the importance of perceived and personal competence (perceived control beliefs, that this construct has competence). Outcome contingency not been studied as a possible invoives the degree to, which outcomes predictor of treatment outcome with depend on the behavior of the chronic adolescent offenders is individual (Le., does what I do make a surprising. These individuals may hold difference in what happens to me). maladaptive control beliefs Personal competence refers to the contributing to and intensifying their individual's capacity to produce the problems, as well as negatively behavior upon which the outcome is influencing their progress in solving contingent (Le., can I do wl1at it takes these problems. If perceived control, an to be successful on this task). These alterable factor, relates to two factors are reported to be psychotherapy success, this variable independent, with a nonsignificant may indicate those chronic adolescent correlation of .18. Research indicates offend ers who would be amenable to that children as young as seven years treatment (Kazdin, ] 987). Those of age can make the distinction adolescent offenders who appear between these concepts (Weisz, 1986). unamenable to treatment when initial The distinction between assessments are conducted, may be perceived contingency and perceived prepared to benefit from specific skills competence is important because the treatment through perceived control two beliefs have different implications training. for behavior and treatment. For The present stud y concerns the example, perceived contingency has relationship between perceived control been shown to predict progress in and treatment outcome with chronic treatment among primarily adolescent offenders who exhibit two externalizing-problem children (Weisz, broad types of behavior difficulties, 1986). In contrast, studies addressing internalizing and externalizing internalizing-behavior problems show a problems. Perceived control was positive correlation between depression addressed from the recent! y and low perceived competence but not reconceptualized two-dimensional between depression and perceived model (Weisz, 1983)986; Weisz & contingency (Weisz, et al., 1989; Weisz, Cameron,1985;Weisz & Stipek,1982). Weiss, Wasserman, & Rintoul, 1987). Therefore, when treating depressed children, the therapy of choice may be 86 Did I Cause This Swenson METHOD operates a behavior-management system focusing on positive Population reinforcement. Students can earn daily Three hundred seven (307) points for performing certain tasks or adolescents committed to a state for exhibiting specific positive training school in the southeast behaviors. Points are traded for valued participated in this study. Criteria for items on a weekly basis at the campus commitment to this facility consists of point store. Positive write-ups are commission of a felony plus a previous earned for appropriate behavior; criminal history, and includes an negative write-ups, for inappropriate expected stay ranging from six months behavior (see inset). Release from the to several years, depending on the facility is contingent upon meeting severity of the offense (Dozier certain criteria, including number of Classification, 1988). All students were negative write-ups. males between the ages of 14 to 18 (M 15.5); 97 were White, 201 were Black, Actions Warranting Major Write-Ups and 9 were Hispanic. 1: Student inflicts physical harm to self fir others or causes permanent damage to property. 2: Student steals or has in possession Therapeutic Milieu something that he does not have permission of Upon commitment to the state the owner to have. student undergoes 3: Student engages in sexual misbehavior, such training school, each as indecent exposure, public masturbation, or a psychological evaluation to aid in coercing or bartering for sexual acts. development of an individual 4: Student attempts to escape. treatment plan. Following S: Student pops electric sockets. 6. Student engages in behavior that threatens psychological assessment, each student the security of self or others. is assigned to participate in weekly 7. Student has 4 minor write-ups in 1 hour. individual and/or group therapy, 8. Student possesses harmful or controlled based on his/her presenting problem substances or objects (such as smoking material or sharp objects). at the time of commitment. Therapists 9. Student threatens the health or safety of are clinical psychology graduate others (such as excessive horseplay, threatening students and/ or social workers under a fight with closed fist or weapon). 10. Student refuses to attend crew, school, the supervision of a licensed group, or individual therapy. psychologist. Group therapy focuses on 11. Student is off-bounds without permission. individual problems, such as substance 12. Student fails to complete assigned detail. abuse, anger control, or commission of 13. Stadent directs profanity toward staff or peers. a sexual offense. Students also attend 14. Student is truant, i.e., departing cottage, an on-campus high school featuring school, crew, group, or therapy without cooperative or team learning and permission or without intending to escape. computer assisted instruction, with a . Double Majors Reality-Therapy-based discipline program. All students reside within In addition to psychotherapy cottages and are supervised by and academic intervention, the campus houseparents. Students who indicate 87 Did I Cause This Swenson they may be a danger to themselves or 80, yes or no questions, and yields a others are placed in a time-out cottage, global measure of self-concept, along the "Intensive Supervision Program," with six individual domains: 1) until the problem is resolved. Behavior -- acknowledgement or denial of behavior problems; 2) Intelligence - l\tfeasures self assessment of intellectual or Child Behavior Checklist. The academic abilities; 3) Physical-self teacher report form of the Child assessment of physical competence; 4) Behavior Checklist (Achenbach & Anxiety-assessment of self as anxious, Edelbrock, 1983) lists 118 child worried, or sad; 5) Popularity - behavior problems and provides assessment of popularity with peers; summary scores for internalizing and and 6) Happiness -- reflection of externalizing problem areas. This scale general satisfaction with life. has been shown to classify effecti vel y Treatment outcome. Treatment the different behavioral patterns of outcome was measured by decrease in juvenile offenders (Brannon & negative write-ups. Because positive Williams, 1986). write-ups typically are earned for Multidimensional Measure of doing extra work and do not appear to Children's Perceptions of Control represent significant behavior change, (MMCPC). This scale measures they were not considered a treatment perceived contingency assessed outcome measure. Data from negative through 48 items, yielding a global write-ups during the tvvo week period perceived contingency score and prior to initiation of treatment and data subscale scores across three domains: during the two week period following social/ cognitive, and physical (Connell, treatment were used to gauge 1985). In addition, the scale addresses treatment outcome:. three possible sources of control for outcomes, including internal, powerful Procedure others, and unknown. This study includes data Perceived Contingency gathered over a one-year period. Behavioral Domain Scale. To measure Following an adjustment period of at perceived contingency concerning least three weeks, teachers completed behavior, a 12-itern scale was used the Child Behavior Checklist on all (Hartsfield, Licht, Swenson & Thiele, students assigned to their classroom. 1989). This scale had not been used Perceived-control measures were previously with juvenile delinquents; administered to small groups of 4-6 therefore, internal consistency measures students. Questionnaires were read were computed on the present data, aloud to all students by the first author yielding coefficient alphas ranging with assistance from upper-level from .43 to .71. undergraduates majoring in The Piers-Harris Self-Concept psychology at a nearby university. Scale. The Piers-Harris (Piers, 1984), Points were provided to students for assesses perceived competence through completion of questionnaires and these 88 Did I Cause This Swenson points could be spent at the campus canteen for snacks or other items. Table 1 Perceived Control Belief Measures as Correlates with Child Behavior Checklist Scores RESULTS Global Measures: To determine the relationship DOMAIN CBCLE CBCLI between internalizing and externalizing Overall self-concept -.14' -.21" scores on the Child Behavior Checklist and perceived control measures, zero Specific Domain Measures on the Piers-Harris order correlations were calculated DOMAIN CBCLE CBCLI (Table 1). Six perceived-competence Denial of behavior problems -.20" -.18' measures were significantly related to student behavior problems as rated by Self-assessment of intelligence -.12 -.15' teachers. Scores on the Chlld Behavior Self-assessment of physical Checklist Internalizing Scale (CBCLI) competence -.07 -.17' negatively correlated with total self Assessment of self as concept, denial of behavior problems, nonanxious -.06 -.14' self-assessment of intelligence or Assessment of self as generally academic skill, self-assessment of happy -.13 -.21" physical competence, selfassessment as Specific Domain Measures on MMCPC Success Scales: worry free or nonanxious, and self perception as generally happy. No DOMAIN CBCLE CBCLI CBCLI scores related significantly to Attributing 5uccessful perceived-contingency scores on the outcomes of behavior to self -.15' -.06 Multidimensional Measure of Specific Domain Measures on MMCPC Failure Scales: Children's Perception of Control. DOMAIN CBCLE CBCLI The Chlld Behavior Checklist Externalizing Scale (CBCLE) related Attributing cognitive failures to self -.19" -.08 significantly to subscales of both the perceived-competence and perceived . Significant at .01 level, "Significant at .001 level contingency measures. Specifically, the CBCLE correlated negatively with analysis revealed a strong, positive overall self-concept and denial of correlation (.63) between the CBCLI behavior p.:oblems. 0{1. perceived and the CBCLE, indicating overlap contingency domains the CBCLE among these students in the two correlated negatively with attributing problems areas. successful outcomes of behavior to self To assess the relationship of and attributing cognitive failures to perceived control to treatment self. A significant positive correlation outcome, a stepwise multiple was found between CBCLE and regression model was used. The attributing successful outcomes of dependent variable (treatment behavior to unknown causes. Further outcome) was the decrease in absolute 89 Did I Cause This Swenson number of behavioral write-ups before those subjects with Child Behavior and after treatment. Perceived Checklist Externalizing Scale scores competence and perceived-contingency higher than their Child Behavior varIables, age of the student, number Checklist Internalizing Scale scores, of commitments to previous programs, indicating their primary behavior Full Scale Wechsler IQ, and Reading problem to be external in nature. Two Achievement were assessed also as variables were significant predictors of predictors of treatment outcome. treatment outcome. The first was Predictors of Ail Students perception of self as generally happy, Treatment Outcome.. First, a forward accounting for 7 percent of the variance stepwise regression was done using all (adjusted R2 = .06658). The second was students because the overlap on the attributing overall successful outcomes Child Behavior Checklist Internalizing to self, contributing 5 percent of the Scale and Child Behavior Checklist variance (adjusted R2 = .11689). Externalizing Scale indicated high Predictors of Treatment comorbidity. In this procedure each Outcome for Internalizers. A third variable was entered one at a time on regression was conducted on students the basis of its F to enter (Le., the whose CBCLI scores were higher than proportion of variance explained by their CBCLE scores, indicating their each independent variable). An primary behavior problem to be analysis of the resid uals indicated all internal in nature. Two variables were assumptions were met except a normal significant predictors of treatment distribution. The normal probability outcome for these students. The plot was s-shaped. Therefore, to variable accounting for most of the normalize the distribution, natural logs variance (9 percent) was self were computed for each score. assessment of physical competence Three variables were significant, (adjusted R2 = .09319). Attributing albeit weak, predictors of treatment overall failure outcomes to self outcome for all students. Self contributed 6 percent of the variance in assessment of Physical Competence treatment outcome (adjusted R2 = accounted for 5 percent of the variance .15960). in treatment outcome (adjusted R2 = Predictors of Treatment .04954). Age added significantly to the Outcome for Severe Externalizers. A explanatory power of the model fourth regression was conducted on (adjusted R2 = .08113), contributing 3 students rated by teachers as exhibiting percent of the variance in treatment severe forms of externalizing behavior outcome. The third significant predictor problems. The criteria for admission of treatment outcome, attributing social into this group induded CBCL failures to unknown factors, externalizing T scores of at least 70 and contributed 1 percent of the variance CBCL internalizing T scores no more (adjusted R2 = .09496). than 60. One significant predictor WaS Predictors of Externalizers. The revealed in this analysis. Assessment of second regression was conducted on self as anxious or worried contributed 90 Did I Cause This Swenson 63 percent of the variance in treatment externalizing behavior problems tended outcome (adjusted R2 = .62642). A to not attribute successful outcomes of regression was not conducted on behavior to their own actions and they severe internalizers due to the small tended to not attribute responsibility sample size. for failure outcomes on cognitive tasks to themselves. They did tend to attribute successful outcomes of DISCUSSION behavior to unknown factors. Perceived Control and This study assessed the Internalizing Problems. Perceived relationship between perceived control competence was significantly related to and type of behavioral difficulty and internalizing behavior problems in addressed whether perceived these offenders. Specifically, students contingency and perceived competency with internalizing problems tended to were differential predictors of have poor overall self-concepts and to treatment outcome with adolescent assess their ability or academic skill, as offenders whose behavior problems well as physical competence, as low. were either primarily internalizing or They also tended to view themselves as externalizing. having a lot of worries and fears, being Perceived Control and unhappy and they acknowledged Externalizing Problems. Both perceived behavioral problems. Perceived competence and perceived contingency contingency was not significantly were significantly related to related to internalizing behavior externalizing, maladaptive behavior in problems in these adolescent offenders. these adolescent offenders. In the area Predictors of Treatment of perceived competence, students with Outcome for Externalizers. Offenders externalizing problems tended to have with primarily externalizing problems a low overall self-concept. On the tended to have a better treatment Piers-Harris, overall self-concept outcome when they attributed overall encompasses all domains, including the successes to their own behavior. On the behavior problem domain. These other hand, they tended to do worse in particular adolescents tended to treatment when they attributed overall acknowledge their behavior problems. negative consequences to their own Therefore, the overall self-concept score behavior. These results suggest that may have been lowered by items claiming responsibility for one's endorsed on the behavior problems failures may be maladaptive, while domain. Rather than a reflection of low perceiving one's self as responsible for self-concept, the low total score may one's successes may be adaptive. The have been a reflection of offenders in this study have likely met acknowledgement of behavior with much failure in their history in problems. schPol, home, and relationships. To In the area of perceived focus on their responsibility for failures contingency, students with may lead to decreased hope for the 91 Did I Cause This Swenson future and they may give up in their improved. treatment programs. In contrast, Internalizers who While focusing on one's own attributed social failures to internal responsibility for failure was a negative factors tended not to do well in predictor of treatment progress, the . treatment. Internal factors include the more Externalizers viewed themselves attribution of effort. Physical failures, as worried or anxious, the more such as missing a thrown ball, may successful was their treatment. In fact~ have less impact if explained as caused this variable was the strongest by insufficient effort. However, if a predictor of treatment outcome for student attributes social failures, such Externalizers, highlig~ting the as poor peer relations, to lack of effort importance of acknowledging one's (Le., self) and conveys that attitude to problems. authority figures, then those authority Further, the more Externalizers figures may be less likely to offer viewed themselves as generally happy assistance. In the present study, the and physically competent, the less measure of treatment outcome was progress they made in treatment. number of negative write-ups, which Externalizers who viewed themselves were administered by houseparents as happy and perhaps without and teachers, as well as by therapists. problems may be entrenched in denial, It is possible that if students viewed believing they do not need assistance themselves as not trying to make in changing behavior problems that in friends or to get along socially, and their view do not exist. The latter conveyed this attitude to staff, then finding suggests that aggressive staff would be less likely to be lenient students may see themselves as in write-ups. Several lines of research invincible and may associate strength have shown that those who are and physical competence with being perceived as acting deliberately are less untouchable by consequences. It is also likely to receive help (Piliavin, Rodin, plausible that with their focus on how & Piliavin, 1969) and more likely to physically competent they are, they receive criticism and punishment may tend to use physical means (Brophy & Rohrkemper, 1981). (aggression) for solving problems, reinforcing a "macho image." Predictors of Treatment SUMMARY Outcome for Internalizers. Consistent \o\ith Extemalizers, Internalizers who Overall, global measures of tended to attribute overall failures to perceived competence did not predict themselves and those who viewed treatment outcome. However, global themselves as high in physical perceived contingency measures were competence did not do well in significant predictors of treatment treatment. However, if the Internalizers outcome for both Internalizers and attributed their physical failures to Externalizers. That is, success in themselves, the treatment outcome was treatment for both Internalizers cLl1d 92 Did I Cause This Swenson Externalizers was related to viewing then perhaps the group in the Brannon overall outcomes of their behavior as and Williams study did not exhibit a contingent on what they do. In general pathology and were therefore, addition, several domains of perceived more likely to show primarily contingency and perceived competence internalizing or externalizing problems did predict treatment outcome. than were students in the present The results of this study support study, who with their perhaps greater the importance of considering domains pathology, were not so clearly defined of perceived control. The efficacy of by this rating scale. using two measures of perceived Consistent with previous control for predicting treatment studies, Edelbrock and Costello (1988) outcome appears warranted. However, also found that the Internalizing and using the two measures to predict Externalizing Scales related to different treatment for different types of diagnoses with psychiatric in-patients. disorders is not as strongly supported, However, depression correlated at least with chronic adolescent significantly with both Internalizing offenders. and Externalizing Scales. As in the One influence on the results of current investigation, studies with this study appears to be the children who were psychiatric in relationship between the Child patients indicated a high level of Behavior Checklist Externalizing Scale comorbidity on the Child Behavior (CBCLE) and the Child Behavior Checklist Gones, Latkowski, Kircher, & Checklist Internalizing Scale (CBCLl). McMahon, 1988; Woolston et a1., 1989). These results were contrary to previous A possible explanation for the studies using the Child Behavior high correlation between the CBCLI Checklist with juvenile delinquents. In and the CBCLE in this study is that the a study assessing juvenile offenders teachers who completed the ratings do committed to a state juvenile services not teach nondelinquent students and agency, Brannon and Williams (1986) therefore have little basis for a day to found .that the Child Behavior day comparison of various degrees of Checklist effectively substantiated behavior problems. Therefore, they differential behavioral patterns: 61 may view their students as exhibiting a percent of their subjects were classified general pathology because they all are as Externalizers; 11 percent as involved with delinquency. A second Internalizers. Students in this study explanation may be that many of these were from three levels of the agency, chronic delinquents do exhibit general including training schools, non-secure pathology, displaying worried, anxious park camps, and community group behavior as well as aggression. homes. They may have differed from The present study is consistent students in the present study in wi th theories supporting the severity of behavior problems and importance to treatment of accepting length of criminal history. If their responsibility, at least for one's behavior problems were less severe; successes. If Internalizers and 93 Did I Cause This Swenson Externalizers attribute their failures to cause of the problem, the student self, but do not pErceive successes as might concentrate attention on negative contingent on their behavior, they may self-blame. This focus on negative minimize their role in improvement. behaviors may elicit feelings of This attribution style has been shown hopelessness in solving the problem. to be common among depressed For some students, targeting blame for children, who are more likely to show the problem may lead to denial, thus self-blame for negative events than avoiding solving the problem. On the nondepressed children (Seligman et aI, other hand, if the focus of therapy 1984). Similarly, those offenders who were on teaching students that they are blame themselves for negative events responsible for solving their problems may feel helpless to solve their and successes are contingent on their problems, which may in turn further behavior, they may feel more hope for exacerbate their difficulties. improvement. The present results suggest that If students believe that solving treatment for juvenile delinquents their problems is contingent on their should focus on changing perceived behavior, but they lack the skills to contingency beliefs to teach adaptive solve the problem, they probably will belief styles. Brickman et a1. (1982) incur failure. A combination of training make a distinction between control in skill defici ts pI us training to beliefs for the source of the problem attribute responsibility for problem and for the solution of the problem. solving to one's self may be the most They assert that an individual is not valuable treatment for adolescent responsible for the source of the offenders. To work first on changing problem, but is held responsible for the faulty beliefs and then provide skill solution. When this model is applied, training may be the most beneficial Brickman et al. argue that individuals strategy. If adolescent offenders view view themselves as having to themselves as responsible for solving compensate, through effort, ingenuity, their problems, they may be more and collaboration, for the handicap receptive to skills training. imposed on them. This model allows In summary, the present study the individual to focus on solving the has implications for treatment and problem without dwelling on guilt future research. For treatment, two feelings for creating the problem. important points emerge. First, guiding A predictor of treatment offenders to view themselves as having outcome for Internalizers and a problem is essential to improvement Externalizers was attributing successes in treatment, especially for aggressive to self. Given that attributing successes students. Second, focusing on their to self is important to treatment, successes and the view that they are perhaps treating delinquents based on responsible for their successes and the compensatory model would best resolving their problems appears emphasize this belief. If therapy important for actual treatment gains. focuses on the student's being the In an institutional setting, where 94 Did I Cause This Swenson numerous students have treatment that control beliefs predict treatment needs and placement is time-limited, outcome. Past research has shown that triage methods may be most effective. various therapies change control beliefs That is, students who score low on the (Long & Sherer, 1984; Thatcher, 1983; Piers-Harris Anxiety Scale, indicating Waksman, 1984; Yarish, 1986). acknowledgement of anxiety and Additional work is needed to worries, may be immediately amenable determine which therapies are most to a combination of attribution and effective in changing perceived control skills training. Other students, who beliefs with chronic adolescent score high on the Piers-Harris Anxiety offenders. Development of a perceived Scale, will be less amenable to control attribution training package for treatment. They will require work on behavioral areas (as opposed to the denial of their problems before academic attribution training) may be attribution and skills training can helpful in guiding these offenders occur. Given the limited resources toward significant behavior change. often found in institutions, students in Consistent with previous the former group should be first research, the value of teaching self priority for treatment programs. responsibility for behavior through The second important point therapy is highlighted. Intensive from the present study relates to training of perceived control beliefs directions for future research. may bring about an increase in positive Considering the overlap between the behaviors and perhaps elicit adaptive internalizing and externalizing coping and long-term maintenance of behavior problem areas shown in this behavior change. study, future research is needed to further clarify the efficacy of making the distinction between perceived contingency and perceived competence REFERENCES beliefs for different behavior disorders. Measures that more clearly Achenbach, T. M., & Edelbrock, C. (1983) differentiate internalizing and Manual for the child behavior checklist externalizing problem behaviors should and revised child behavior profile, Burlington, Vermont: University be used. For example, self-report Associates in Psychiatry. measures of depression and aggression in lieu of, or in addition to, teacher Brannon, J. M., & Williams, D. (1986). The reports may better distinguish the effectiveness of the Child Behavior groups. With the severe juvenile Checklist in identifying the behavioral offenders in this study, the Child patterns and security requirements of juvenile otfenders. International TournaI Behavior Checklist did not distinctly of Offender Therapy and Comparative diagnose internalizing and Criminologv".2Q, 195-201. externalizing behavior problems. Research in the treatment area is Brickman, P., Rabinowitz, V. c., Karuza, J., Jr., also needed. The present study verifies Coates, D., Cohn, E., & Kidder. L. 95 Did I Cause This Swenson (1982). Models of helping and coping. characteristics and retention in American Psychologist,E, 368-384. treatment in adolescent drug treatment programs: A large-scale cross Brophy,]. E., & Rohrkemper, M. M. (1981). The validation.lournal of Drug Education, influence of problem ownership on lEt 149-165. teachers' perceptions of and strategies for coping with problem students. Hartsfield, F., Licht, B. G., Swenson, C. c., & Journal of Educational Psychology, n, Thiele, C. (1989). Control beliefs and 295-311. behavior problems of elementary school children. Paper presented at Caspi, A., Elder, G. H., & Bem, D.]. (1987). annual meeting of the American Moving against the world: life-course Psychological Association, New patterns of explosive children. Orleans, Louisiana. Developmental Psychology,~ 308-313. Jones, R. N., Latkowski, M. E., Kircher, ]. c., & Connell,]. P. (1985). A new multidimensional McMahon, W. M. (1988). The Child measure of children's perception of Behavior Checklist: Nonnative ,:ontroJ. Child Development,Q.§, 1018- infonnation for impati.ents. TournaI of 1041. the American Academy of Child and Adolescen.t Psychiatry, 'JZ 632-635. Davis, G. E., & Leitenberg, H. (1987). Adolescent sex offenders. Psychological Kazdin, A. E. (1985). Treatment of antisocial Bulletin, 101. 417-427. behavior in children and adolescents. Homewood, JIJinois: Dorsey Press. Dozier classification of length of stav. (1988). A. G. Dozier School for Boys. Kazdin, A. E. (1987). Treatment of antisocial Marianna, Florida. behavior in children: current status and future directions. Psychological Edelbrock, c., & Costello, A. ]. (1988). Bulletin, 102, 187-203. Convergence between statistically derived behavior problem syndromes Larson, K. A. (1988). A research review and and child psychiatric diagnoses. alternative Journal of Abnonnal Child Psychology, hypothesis explaining the link beh"leen 1.§, 219-231. learning disability and delinquency. Journal of Learning Disabilities, ~ Farrington, D. P. (1983). Offending from 10 to 357-369. 25 years of age. In K. T. Van Dusen & S. A. Mednick (Eds.) Prospective Long, S.J., & Sherer, M. (1984). Social skills studies of crime and delinquency, training with juvenile offenders. Child Boston: KIuwer-Nijhoff. and Family Behavior Therapy, .2t 1-11. Feindler, E. L., Ecton, R. B., Kingsley, D., & Phillips, E. L., Phillips, E. A., Fixsen, D. L., & Dubey, D. R. (1986). Group anger Wolf, M. M. (1971). Achievement Place: control training for institu tionalized Modification of the behaviors of psychiatriC male adolescents. Behavior predelinquent boys within a token Thera,py, lZ, 109-123. economy. TournaI of Applied Behavior AnalysiS,l, 45-59. Friedman, A. 5., Glickman, N. W., & Morrissey, M. R. (1986). Prediction of Piers, E. V. (1984). Revised manual for the successful treatment outcome by client Piers-Harris Self-Concept Scale. Los 96 Did I Cause This Swenson Angeles: Western Psychological Weisz, J. R., & Cameron, A. M. (1985). Services. Individual differences in the student's sense of control. In C. Ames & R. Piliavin, I. M., Rodin, J., & Piliavin, J. A. (1969). Ames (Eds.), Rese~ll'Ch on motivation in Good samaritanism: An underground education: Vol. 2. The classroom milieu phenomenon? Journal of Personality (pp. 93-140). New York: Academic and Social Psycholo,gy, g 289-299. Press. Quay, H. C. & Werry, J. S. (1986). Conduct Weisz, J. R., Stevens, J. S., Curry, J. F., Cohen, disorders. In H.C. Quay & J. S. Werry R., Craighead, E., Burlingame, W. V., (Eds.) Psychopathological disorders Smith, A., Weiss, B., & Parmelee, D. X. ofchild.b,ood, New York: John Wiley & (1989). Control-related cognitions and Sons. depression among impatient children and adolescents. TournaI of the Seligman, M.E.P., Peterson, c., Kaslow, N.]., American Academy of Child and Tanenbaum, R.L., Alloy, L.B., & Adolescent Psychiatry, ~ 358-363. Abramson, L.Y. (1984). Attributional style and depressive symptoms among Weisz, J. R., & Stipek, D. J. (1982). Competence, children. TournaI of Abnonnal contingency and the development of Psychology, 22, 235-238. perceived control. Human Development, ~ 250-281. Swenson, C. c., Butler, A. c., Kennedy, W. A., & Baum, J. (1989, May). Group Weisz, J. R., Weiss, B., Wasserman, A. A., & treatment for anger control with Rintoul, B. (1987). Control related institutionalized adolescent offenders. beliefs and depression among clinic Poster presented at the meeting of the referred children and adolescents. Association for Behavior Analysis, Abnormal Psychology, .2§., 58-63. Milwaukee, Wisconsin. West, D. J, (1982). Delinquency: Its roots, Thatcher, J. A. (1983). Effects of reality therapy careers and prospects. Cambridge, MA: upon self-concept and locus of control Harvard University Press. for juvenile delinquents. TournaI of Reality Therapy, 2, 31. Woolston, J. L., Rosenthal, S. L., Riddle, M. A., Sparrow, S. S., Cicchetti, D., & Waksman, S.A. (1984). Assertion training with Zimmerman, L. D. (1989). Childhood adolescents. Adolescence, n, 123-130. comorbidity of anxiety / affective disorders and behavior disorders. Weisz, J. R. (1983). Can I control it? The Journal of the American Academy of pursuit of veridical answers across the Child and Adolescent Psychiatry, ~ life-span. In P. B. Baltes & O. G. Brim 707-7]3. (Eds.), Life-span development and behavior (Vol. 5, pp. 233-300). New Yarish, P. (1986). Reality therapy and the locus York: Academic Press. of control ofjuvenile offenders. lournal of Reality Therapy, .Y1 3-10. Weisz, J. R. (1986). Contingency and control beliefs as predictors of psychotherapy outcDmes among children and adolescents. TournaI of Consulting and Clinical Psychology, M, 789-795. 97 DUI Offenders' Pt:fceptions of Recovering and Nonalcoholic Counselors Shirley A. Spooner, Ph.D. INTRODUCTION Haskins, 1985). Politicians, employee Driving a motor vehicle is assistance programs, other private and associated with leisure, pleasure, public health care industries, and the necessity, aggravation, danger, and media (Bradstock, Marks, Forman, sometimes death. Elements of danger Gentry, Hogelin, Binklin, & Towbridge, and death are compounded when 1987) have also joined forces in efforts psychoactive substances impair driving to prevent the maiming and / or death ability (Donelson, 1988). of many innocent victims of drunken Of all the drugs which affect drivers (Roizen, 1982). mood, reasoning, and control, alcohol Approximately every twenty is the most socially acceptable (Ray, minutes someone dies in an alcohol - 1983). It is readily available for legal related traffic accident and another purchase to those of age and also 650,000 are injured yearly (Podolsky, accessible directly or indirectly to many 1985). Although alcohol is implicated minors. When drugs are implicated in in over 50 percent of all traffic automobile accidents, alcohol is most fatalities, it is unknown to what extent often the culprit (Podolsky, 1985). other drugs, like marijuana and Although previously tolerated in cocaine, may be involved. Routine the United States, drunken driving and testing for other drugs is currently its consequences have become a ineffective and inexpensive. Unlike national problem (Borkenstein, 1985). alcohol, for which the legal limit is The Surgeon General's Report of 1979 .10% in Florida and in most other helped bring this issue to the national states, dangerous levels of other drugs forefront. Organizations such as have not been established (Fitzgerald & Mothers Against Drunk Drivers Hume, 1987). (MADD), Remove Intoxicated Drivers Driving Under the Influence (RID), and Students Against Driving Countermeasures. Attempts at Drunk (SADD) have been largely controlling the substance abuse driving responsible for increased public problem have included such awareness of the problem (Haskins & suggestions as imposing insurance sur- This paper is from Dr. Spooner's dissertation, "DW Offenders' Perceptions of Counselor Empathy and Helpfulness," R. Myrick, Chairperson. The paper is co-authored by Harry L. Spooner, DW Coordinator, Vol usia-Flagler Safety Council, and Helen B. Justice, Executive Director, Tampa-Hillsborough DW COllnterllttack. Shirley Weedman Spooner, lMHC, has worked in Florida's DW system in various professional capacities jor 11 years. She is cllrrently in full time private practice in Daytona Beach, Florida. 98 DUI Offenders Spooner charges and denying insurance benefits Additionally, as arrest rates fluctuate to guilty drivers Gacobs, 1988). within a precinct, the percent of Bartenders and owners have been offenders with a drinking problem charged with the liabili ty of serving diagnosis also varies (B. Gross, "one for the road" in some instances. personal communication, August 10, Besides taverns and bars, private hosts 1989). For example, when officers have also been held responsible when arrest relatively few drinking drivers in injury occurred after a.n impaired a given month, the chronic alcoholic driver left their parties (Bonnie, 1985). with a high breathalizer reading is Random breath testing to deter more likely to receive police attention. the potential drinking driver has been Florida's Rehabilitation Program. used in some locations (Hamel, 1988). In Florida's rehabilitation system, first In Massachusetts automatic inpatient offenders are required to attend a treatment has been court-ordererl for primarily didactic secondary multiple offenders as an alternative to prevention course for a minimum of 12 jail (McCarty & Argeriou, 1988). But hours. Multiple offenders, or the single greatest countermeasure, individuals who have previous DUI aside from the actual charge of driving convictions, must successfully complete under the influence (DUI) (e.g., Florh:la at least 21 hours of group-process Statute 316.193) and subsequent license oriented education. revocation, probation, fines, possible Both courses incl ude an jail, and community service, is evaluation of the offender to determine mandatory education and rehabilitation if more extended rehabilitation services of the offender (Wilde, 1985). are needed. The ultimate question Malfetti and Yaksich (1984) each counselor must as is, "Will reported the existence of at least 1,500 education alone be enough to prevent rehabilitation programs in the United another DDI arrest?" Throughout the States and Canada. While these procedures of arrest and rehabilitation, programs are geared toward imparting the offender may experience reactions of general information, many include of "resentment, indignation, shame, referral for more intense treatment frustration, and fear" (Malfetti & when an alcohol/ drug problem exists. Winter, 1980, p. 5). How he or she According to Pursch (1981), perceives the counselor may exacerbate approximately 50 percent of first or diminish these negative reactions. offenders display alcoholic behavior Statement of the Problem. which would require rehabilitation Offenders are mandated to complete a beyond education. However, arrest substance abuse course and an rates of drinking drivers vary from alcohol/ drug screening through the precinct to precinct, depending on DUI program by the judiciary. This factors such as politics, cummunity and action often makes them resistant, national interest, funding, and police uncooperative, and even hostile. sophistication (D. Manning, personal Course registrars, instructors, and communication, July 29, 1989). particularly the counselors are the main 99 OUIOffenders Spooner recipients of mild to aggressive verbal bond in counseling relationships assaults when they do occur. Even (Rogers, 1975). Moreover, Kirk, Best, when no word are spoken, the "I'm and Irwin (1986) stated, "It has been angry!" message is often conveyed assumed that some similarity in back through door slamming and tire ground helps to enhance empathy~' squealing once the evaluative interview (p. 82). Alcoholism may provide the is concluded. common bond. Hostility toward personnel may Patterson (1988) further believed be viewed as a defense (Malfetti and that empathy is a method of delaying Winter, 1980). Typically, the more judgment. Since OUI clients are prone defensive or 'hostile an offender, the to be hostile, especially because they more likely he or she is to be referred feel judged, would DUI counselors by the counselor for treatment beyond presented as "recovering" be received the twelve or twenty-one hours more readily and regarded as more required for OUI course completion. empathic than "nonalcoholic" More expenses arise from the cost of counselors? Ooes referral or treatment and the probability of nonreferral by a recovering or extended probation, contributing to nonalcoholic counselor playa role in offenders' feelings of frustration, being perceived empathy? out of control, and misunderstood. Purpose of the Study. The The counselor, attempting to be a purpose of this study was to examine helper, is frequently viewed by the the effects of two counselor offender as being in an adversarial role presentations (types) as perceived by similar to the legal system. How might our offenders. More specifically, this this problem be addressed? Perhaps study focused on how male OUI empathy plays a vital role. offenders i:1 a Florida evaluation center A prevalent assumption in the perceived counselor empathy and alcohol! drug counseling field is that, helpfulness based on how the "It ta.1(es one to know one." In other counselors were introduced to them, as words, alcoholics are best suited to either recovering or nonalcoholic diagnose and rehabilitate other themselves. Evaluation outcome, or alcoholics, by modeling recovery, referral status to extended treatment, recognizing and confronting defenses was also examined. they once used themselves, and Need for the Study. The specific offering hope (Kirk, Best, & Irwin, assumption that" ... the recovered 1986). Are these factors likely to alcoholic's drinking history is thought enhance OUI offenders' perceptions of to give him or her something in empathy in counselors in the our common with his subjects ... has setting? remained untested" (Kirk, Best, & Of all the facilitative dimensions Irwin, 1986). Kirk et a1. (1986) in therapy, empathy has been conducted the only study found in the described as the prime ingredient for literature regarding client perception of establishing rapport and building a empathy and helpfulness in 100 ------ DUI Offenders Spooner "recovering" counselors. Examining process the counselor becomes more DUI offenders' perceptions of empathy and more attractive and a power base in counselors presented as recovering is established, much like the and nonalcoholic will not only help fill parent-child relationship (Strupp, a gap in the literature, but may be 1973). directly useful for treatment in this How the therapist exerts an effect setting. upon the client is a social influence The results of the present phenomenon (Corrigan, 1980; Heppner investigation have contributed to the & Dixon, 1981; Heppner & Heesacker, knowledge regarding offenders' 1982). Empathy and helping are the perceptions of empathy in counselors therapeutic elements to be directly presented as recovering and examined in this investigation from the nonalcoholic. It has helped answer the social influence model of intervention question, "ls it a positive or a negative (Barak, Watkin, & Dell, 1982; Heppner attribute to be recovering in assessing & Heesacker, 1982; Vargas & substance abusers in the DUI Borkowski, 1982). population'?" Can we manipulate The Offenders. Offenders may perceptions of the offender to view fall anywhere on the drink-user counselors as more empathic and continuum from social to alcoholic helpful so they become more (Hoffmann, Ninonuevo, Mozey, & cooperative? Is there a way to create Luxenberg, 1987). But the typical DUI bonding with the helper, promoting offender is not the friendly neighbor compliance and reducing hostility and and outstanding citizen next door, one resistance? who just happened to have one sip too Theoretical Framework. much. Rather, this person is more Existential, Gestalt, and client-centered likely to be a man with a rather shady approaches to psychotherapy rely history and symptoms of problem heavily on the philosophical premise of drinking (McCord, 1984). phenomenology (Waldron-Skinner, Men are the more represented 1986). Inherent in this phenomenologi gender in the adjudicated or the cal psychology is the attempt to adjudication-pending drinking-driver evaluate perceived internal and population (Clay & Swenson, 1978; external events. Client-centered Hoffmann, Ninoneuvo, Mozey, & therapy, on which this study is based, Luxenberg, 1987). Several reasons may performs evaluative tasks by analyzing account for the difference of 9-20 more the helper's power of social influence men than women being arrested for through a number of facilitative driving under the influence of alcohol. dimensions. When there is a couple in the car, Social influence, or change, then, the male is more likely to be driving. occurs when the therapist competently In addition, there are more licensed conveys certain attributes, such as drivers who are male and, as in the respect, genuineness, and empathy to past, more men than women tend to the client (Egan, 1975). Through this drive after drinking (Cosper & 101 DUI Offenders Spooner Mozersky, 1968; Hoffmann, other, particularly through group Ninoneuvo, Mozey, & Luxenberg, fellowship (Alibrandi, 1982). Those 1987). with established sobriety have also . As a consequence, men have been long been contributors on treatment more intensively studied as the teams, regardless of profession or drinking-driver offender (Shore, credentials (Royce, 1981). Yet, the McCoy, Toonen, & Kuntz, 1988). Some premise on which they operate, the research on drinking drivers also intangible bond of alcoholism, has focuses on youth, but less often, mainly been assumed to be effective women. during intervention. As Zir1berg Assessing the Offender. The (1982) stated, "There is no scientific, professional counselor must stand medical, or psychological theory that ready to use every skill acquired in serves as the basis for the folk psycho training during the clinical interview therapy ... " (p. 1002). The model for (S. Reed, personal communication, July this folk psychotherapy is Alcoholics 1, 1990). Offenders expect to be Anonymous. Of the leaderless peer condemned for their behavior and as a group, Zimberg (1982) further added, result most often present a defensive, "Alcoholics Anonymous, or "AA," has defiant, and obstinant posture, been one of the most successful especially during feedback. treatment approaches for alcoholism. Additionally, other psychopathology .. because it (AA) implicitly frequently accompanies substance understands the psychological conflicts abuse (Meyer, 1986; Mirin, Weiss, and psychological needs of the Mich.:~el, & Griffin, 1988), requiring a alcoholic" (p. 1001-1002). "dual diagnosis," a swift accurate Though the self-help group is assessment by the counselor, and widely lauded (e.g., McInerney, 1973; proper routing for assistance. Nace, 1987; Staub & Kent, 1973), their Robert W. Davis (1985) offered a anonymity makes it difficult to more correctional assessment and treatment objectively define the term "success," stage model which fits nicely with the and little hard data contribute to the perceptions of counselors employed in validity of such statements. Mental the DUI field (1. Armstrong, personal health professionals and physicians communication, March, 1991). Similar have historically not been effective in to Elizabeth Kubler-Ross' (1969) model, treating alcoholics. As a result, a rift the "Stages of Incarceration" include (1) between recovering and nonalcoholic denial and disbelief, (b) bargaining, (c) and professional and paraprofessional anger, (d) depression, and (e) helpers has ensued (Royce, 1987). Each acceptance. The majority of clients extolls the merits of his approach, and enrolling in the DUI program appear to scarce research exists to warrant be in an angered state (Malfetti & endorsing one philosophy and Winter, 1980). background over another. Recovering Helpers. For many years counselors, with or without credentials, alcoholics have been helping each are widely employed for generally 102 DUI Offenders Spooner accepted rationale. devised for this study, rated a .82 :Many recovering individuals, coefficient at a test-retest interval of regardless of academic background, two weeks. It, too, employs have attempted to abolish the Likert-type responses. para-professional image and create a Proced ures. The four groups new profession (e.g., Certified compared were: (a) recovering Addictions Professional) through counselor / clients referred, (b) credentialing in various states (Staub & recovering counselor / clients not Kent 1973). Some are even considering referred, (c) nonalcoholic lobbying for insurance reimbursement counselor / clients referred, and (d) for their counseling services (B. Gross, nonalcoholic counselor / ( . :ents not personal communication, August, referred. Subjects vieweu a 15 minute 1990). segment of a simulated screening interview of counselor "A" or "B" METHOD interacting with a male posing as a Participants. Eighty first-offender client. Both middle age, nonalcoholic, male volunteers were solicited from the femal e counselors possessed at least a Tampa-Hillborough DUI Counterattack bachelors degree, were certified as program to participate in the study. evaluators by Florida's DUI rehabilita Their age ranges were 18 to 72. All tion system, and held more than 4 were Caucasian. years experience in the DUl system. Instruments. Offenders' The confederate offender was actually perceptions of counselor empathy were a certified DUl instructor. measured b)T the Barrett-Lennard At random, 20 volunteers were Relationship Inventory (Barrett-Len told counselor "Atl was recovering, nard, 1962). Counselor helpfulness while another 20 were told counselor was assessed by the Client Opinion "A" was nonalcoholic. Likewise, 20 Questionnaire. The 64-item volunteers were told counselor "B" was Relationship Inventory scales measure recovering, while another 20 were told empathy, regard, unconditionality, and counselor "B" was nonalcoholic. Also congruence in a Likert-type format. randomly, half the subjects from each The total score reflects the subject's condition were told the client was overall perception of the relationship. referred to extended rehabilitation by This study employed a 28-item version the counselor. The other half were told of the inventory previously used by the client did not need further Kirk et al. (1986) in their study of intervention for a drinking or a drug perceptions of empathy in therapists problem, according to the counselor. presented as recovering and Upon viewing the tape, program nonalcoholic. The modified form registrars administered the two instru yielded a .78 coefficient at an interval ments. of two weeks in a test-retest study by this investigator. The helpfulness scale, specifically 103 -----~------ DUI Offenders Spooner RESULTS Both univariate and multivariate A between-subjects factorial analyses on the dependent variables MANOVA was performed to evaluate were employed. The results of the two the null hypotheses. Because of the separate univariate analyses are given lack of hypothesized intercorrelation in Table 1. between the two dependent variables, No effects are statistically counselor helpfulness and counselor significant for counselor empathy as empathy Cr = .05, 12. < .66), the measured by the Barrett-Lennard univariate results from the MANOV A Relationship Inventory, or for proced ure were also examined for each counselor helpfulness, as measured by individual dependent variable. the Client Opinion Questionnaire. Four parallel tests of significance MANOV A test criteria and exact were evaluated for each of the three F statistics for the hypothesis of no effects: client disposition (referred or overall effect are given in Table 2 for not referred), counselor status the three effects utilizing Type III sums (recovering or nonalcoholic), and the of squares and cross-prod ucts matrices. interaction between counselor status No statistically significant results and client disposition. are present for any of the effects. Because the Client Opinion Questionnaire originated with this Table 1 Between-Subjects Factorial Multivariate study, however, it was analyzed on the Analysis: Univariate Result same dimensions. Of the 10 ite~.;, the following two were found to be Dependent Variable: Counselor Helpfulness significant on univariate ANOV AS. Source d.f. E-Value E-Statistic Model 3 0.86 0.4672 Item Two: Fro111 the way the counselor assisted the client, it seemed like she Error 76 also had a Dill at one time. Counselor Status 1 0.47 0.4940 Client Disposition 1 1.42 0.2373 A significant main effect was found for counselor status E (1,76) = Status x 1 0.68 0.4121 Disposition 17.94,12.< .001 (Table 3). A comparison of means indicates that recovering Dependent Variable: Counselor Empathy counselors were perceived as having Source d.f. E-Value E-Statistic had a DUI at one time more than Model 3 0.71 0.5481 nonalcoholic counselors (2.92 and 4.32, respsectively). (See Table 4.) Error 76 Counselor Status 1 0.83 0.3659 Item Six: The counselor was Client Disposition 1 1.20 0.2760 judgmental. Status x 1 0.10 0.7494 Disposition A significant main effect was found for counselor status, f, (1, 76) = 104 DUI Offenders Spooner Table 2 Results of Between Subjects Factorial Multivariate Analysis: Manova Statement Hypothesis of No Overall Counselor Status Effect Statistic F-Value Num d.f. Den d.f. E-Statistic Wilks'Lambda 0.60 2 75 0.5492 Pill ai's Trace 0.60 2 75 0.5492 Hotelling-Lawley Trace 0.60 2 75 0.5492 Roy's Greatest Root 0.60 2 75 0.5492 Hypothesis of No Overall Client Disposition Effect Statistic F-Value NUl;: dJ. Den dJ. E-Statistic Wilks' Lambda 1.38 2 75 0.2572 Pillai's Trace 1.38 2 75 0.2572 Hotelling-Lawley Trace 1.38 2 75 0.2572 Roy's Greatest Root 1.38 2 75 0.2572 Hypothesis of No Overall Status x Disposition Effect Statistic F-Value Num dJ. Den d.f. 12-Statistic Wilks' Lambda 0.40 2 75 0.6686 Pillai's Trace 0.40 2 75 0.6686 Hotelling-Lawley Trace 0.40 2 75 0.6686 Roy's Greatest Root 0.40 2 75 0.6686 Tahle 3 Table 4 Results of Univariate Analysis Results of Univariate Analysis Dependent Variable: Item Two Dependent Variable: Item Six (The counselor was (From the way the counselor assisted the client, it judgemental). seemed like she also had a DUI at one time). Source d.f. F-Value E-Statistic Source dJ. F-Value E-Statistic Model 3 3.71 0.0151 Model 3 6.20 0.0008 Error 76 Error 76 Counselor Status 1 6.03 0.0164 Counselor Status 1 17.94 0.0001 Client Disposition 1 4.19 0.0442 Client Disposition 1 0.57 0.4518 Status x Disposition 1 0.19 0.3428 Status x Disposition 1 0.09 0.7631 105 our Offenders Spooner 6.03, l2. < .01 and a main effect was Questionnaire may have failed to found for client disposition f (I, 76) = detect any significant differences 4.1912. < .04. (See table 4.) A compari because first offender DUI students son of means indicated that recovering who really have a drinking problem counselors were perceived as (an estimated 50 percent), versus those judgmental more than nonalcoholic who do not fit criteria for further counselors (3.50 and 4.30, respectively. referral, may have answered questions For client disposition, a comparison of at opposite extremes. This means revealed that those who v\Tere phenomenon would serve to cancel out told the client was referred perceived effects of the combined group. the counselor as more judgmental than Analyses of the items on the those who were told the client was not Client Opinion Questionnaire, referred (3.48 and 4.22, respectively). however, offer clues as to how OUI clients perceive counselor status and DISCUSSION AND IMPLICATIONS the effect of client disposition. Subjects Improved techniques and perceived recovering counselors as methods of influencing our offenders more likely to have had a our them to address their psychoactive substance selves, a shared commonality. abuse are sorely needed in the our Theoretically, then, the counselor population. The premise of this study, should have more credibility and the powerful social influence theory influence with DUI students because of inherent in Client Centered Therapy a membership-group similarity effect (Egan, 1975), suggests that our clients (Atkinson, Ponce, & Martinez, 1984; might be persuaded to cooperate, Atkinison & Schein, 1986; Porche & change, or otherwise red lice their Banikiotes, 1982). resistant stance if the therapist On the other hand, responses on competently conveys empathy and Item 6 suggest that counselors posing helpfulness. The counselor would as recovering were perceived as more become more attractive, thus judgmental than their nonalcoholic establishing a power base (Strupp, colleagues. A possibility exists that 1975). these students viewed the recovering This study, which has examined counselor as lithe pot calling the kettle actual DUI students' responses to black." In other words, how dare perceived counselor status as well as recovering counselors attempt to their reactions to client disposition, evaluate others since they have made provided information where little has the same mistake. existed. However, the efficacy of using Results on Item 6 also reflect that one type of counselor over the other both types of counselors who referred for the DUI population was not clients for further treatment were seen supported by either dependent variable as more judgmental than those who on the measures used. failed to refer. This finding is entirely The Barrett-Lennard Relationship consistent with the clinical experience Inventory and the Client Opinion conveyed -by seasoned recovering and 106 DUI Offenders Spooner nonalcoholic counselors in the DUI guilt, remorse, embarrassment, field. To this investigator's knowledge, humiliation, and other feelings which the present study is the first experi are painful. The DUI offender would mental attempt to objectively evaluate have ample reason to calIon his DUI students' perceptions of the defenses to deal with the consequences evaluative and referral process. of sustaining a DUI. More research is Whether a counselor is recovering needed to identify and thus modify or nonalcoholic represents a type of factors in the evaluative process which therapist self-disclosure. Choosing to seem to dredge up these behaviors. make this information known to the Performing an analysis of client, even upon request, is apt to respondents in the four experimental make a difference in offenders' groups who were ultimately referred to perceptions of the counselor on the treatment and those who were not dimension of being judgmental (and referred to treatment could provide possibly other, as yet untested, insight into any bias their own dimensions). Moreover, true psychoactive substance use may have differences may exist between contributed. The present study solicited recovering and nonalcoholic only Caucasian volunteers age 18 and counselors. Self-disclosure is a much over. Although these are by far the studied source of research interest, one most common DUI offenders, future which has yielded mixed results studies should also include women, (peca-Baker & Friedlander, 1989; minorities, and youthful offenders. Watkins & Schneider, 1989). There is also a possibility that responses may differ in those who SUMMARY AND SUGGESTIONS have accumulated more than one Neither the Barrett-Lennard offense. Examining the variable of not Relationship Inventory or the Client only the first offender, but those with Opinion Questionnaire offered multiple arrests would be wise. Using significant data on the four male counselors may also make a experimental groups. Two items, difference. however, on the Client Opinion Matching problem drinkers with Questionnaire tentatively suggest that recovering counselors and pairing offender resistance and defense those without drinking problems (those mechanisms (as reflected most fre not in need of additional intervention) quently by anger) may be triggered by with non-recovering counselors may be perceiving the counseior as recovering informative. Finally, the social and judgmental. All referred clients influence premise is credible from seemed to also be viewing the counse logical and research perspectives. The lor as more judgmental, regardless of ways in which the pertinent variables counselor status. of empathy and helpfulness are Dorpat (1987) believed that measured need to be refined. defenses are largely unconscious and automatic, serving to mitigate anxiety, 107 DUIOffenders Spooner REFERENCES Clay, T. R, & Swenson, P. R (1978). Selective enforcement of drunken driving in Alibrandi, L. A. (1982). The fellowship of Phoenix, Arizona. Journal of Studies on Alcoholics Anonymous. In E. Pattison & Alcohol, 12., 516-521. E. Kaufman (Eds.), Encyclopedic handbook of alcoholism (pp. 979-986). Corrigan, J. D. (1980). Counseling as a social New York: Gardner Press. influence process: A review. Journal of Counseling Psycholog)'t E 395-441. Atkinson, D. R., Ponce, F. Q., & Martinez, F. M. (1984). Effects of ethnic, sex, and Cosper, R, & Mozersky, K. (1968). Social attitude similarity on counselor con-elates of drinking and driving. credibility. Journal of Counseling Quarterly TournaI of Studies on Alcohol, Psychology, ~ 588-590. .i. 58-117. Atkinson, D. R., & Schein, S. (1986). Similarity Davis, R. W. (1985). Assessing offenders: in counseling. Counseling Psychologist, dying for reprieve. Tournai of 1.1, 319-354. Personality Assessment, Q. 605-612. Barak, A., Watkin, J., & Dell, D. M. (1982). Donelson, A. C. (1988). The alcohol-crash Effects of certain counselor behaviors on problem. In M. Laurence, ]. Snortum, & perceived expertness and attractiveness. F. Zimring (Eds.), Social control of the Tournai of Counseling Psychology, ~ drinking driver (pp. 3-42). Chicago: 261-267. The University of Chicago Press. Ban-ett-Lennard, G. T. (1962). Dimensions of Dorpat, T. L. (1987). A new look at denial therapist response as causal factors in and defense. Annual of Psychoanalysis, therapeutic change. Psychological 1.2, 23-57. Monolrraphs, Z§, 1-36. Egan, G. (1975). The skilled helper: A model Bonnie, R. J. (1985). Regulating conditions of for svstematic helping an interpersonal alcohol availability: Possible effects on relating. Monterey: Brooks/Cole. highwa y safety. TournaI of Studies on Alcohol, Supplement, 1.Q. 3-12. Fitzgerald, E. F., & Hume, D. N. (1987). Intoxication test evidence: Criminal and Borkenstein, R. F. (1985). Historical civil. Rochester, NY: The Lawyers perspective: North American traditional Cooperative Publishing Company. and experimental response. Iournalof Studies on AlcohoL Supplement, ill Haskins, J. B., & Haskins, T. S. (1985). Major 3-12. social action groups against drunken drivers. Tournai of Studies on Alcohol, Bradstock, M., Marks, J., Fonnan, M., Gentry, ill 192-195. ;., Hogelin, M., Binklin, N., & Trowbridge, F. (1987). Drinking, Heppner, P.O., & Dixon, D. N. (1981). A driving and health lifestyle in the United review of the interpersonal process in States: Behavioral risk factors surveys. psychotherapy. Personal Guidance TournaI of Studies on Alcohol, b 147-152. Journal, .22, 542-550. 108 DUI Offenders Spooner Heppner, P.O., & Heesacker, M. (1982). The McInerney, J. (1973). Alcoholics anonymous interpersonal influence process in members as alcoholism counselors. In real-life counseling. Journal of G. Staub & L. Kent (Eds.), The Counseling Psychology, ~ 215-223. para-professional in the treatment of alcoholism (A new profession) Hoffmann, N. G., Ninonuevo, F., Mozey, J., & (pp. 91-105). Springfield: Charles C. Luxenberg, M. G. (1987). Factors Thomas. associated with motor vehicle accidents among male alcoholics. Journal of Malfetti, J. L., & Winter, D. J. (1980). Studies on Alcohol, ~ 586-590. Counseling manual for educational and rehabilitative programs for persons Holland, A. L., Atkinson, D. R., & Johnson, convicted of driving while intoxicated M. E. (1987). Effects of sexual attitude .cmY!l. Falls Church, VA: AAA and sex similarity on perceptions of the Foundation for Traffic Safety. counselor. Journal of Counseling Psychology, M. 322-325. Malfetti, ]. L., & Yaksich, S. (1984). DWI reeducation and rehabilitation Homel, R. (1988). Policing and punishing the programs--successful results and the drinking driver. New York: future. Journal of Traffic Safety Springer-Verlag. Education,21 7. Jacobs, J. B. (1988). The impact of insurance Meyer, R. E. (Ed.). (1986). Psychopathology and civil law sanctions on drunk and addictive disorders. New York: driving. In M. D. Laurence, J. Snortum, The Guilford Press. and F. Zimring (Eds.), Social control of the drinking driver, (pp. 227-247). Mirin, S. M., Weiss, R. D., Michael, ]., & Chicago: The University of Chicago Griffin, M. L. (1988, December). Press. Psychopathology in substance abusers and their families. Resident and Staff Kirk, J., Best, J., & Irwin, P. (1986). The Physician, pp. 61-65. perception of empathy in alcoholism counselor" TournaI of Studies on Nace, E. P. (1987). The treatment of Alcohol, £., 82-84. alcoholism. New York: Brunner/Maze!. Kubler-Ross, E. (1969). On death and dving. Patterson, L. W. (1988). The function of New York: MacMillan. automaticity in counselor information processing. Counselor Education and McCarty, D., & Argeriou, M. (1988). Rearrest Supervision, 'J2 305-314. following residential treatment for repeat offender drunken drivers. TournaI of Peca-Baker, T. M., & Friedlander, M. L. Studies on Alcohol, i2, 1-6. (1989). Why are self-disclosing counselors attractive? TournaI of McCord, J. (1984). Drunken drivers in Counseling and Development, g longitudinal perspective. TournaI of 279-282. Studies on Alcohol, .1.2, 316-320. Poldolsky, D. M. (1985). Alcohol, other drugs, and traffic safety. Alcohol, Health and Research World,.2" 16-23. 109 Dur Offenders Spooner Porche, L. M., & Banikiotes, P. G. (1982). Waldron-Skinner, ]. A. (1986). A dictionary of Racial and attitudinal factors affecting psychology. New York: Routledge & the perceptions of counselors by black Kegan Paul. adolescents. Journal of Counseling Psychology, ~ 305-314. Watkins, C. D., Jr., & Schneider, L.]. (1989). Self-involving versus self-disclosing Pursch, J. (1981, January). An interview. counselor statements during an initial Alcoholism: The National Magazine, interview. TournaI of Counseling and p.12. Development, §L 345-349. Ray, O. (1983). Drugs, society, and human Wilde, G.]. (1985). The use of incentives for behavior. st. Louis: The C. V. Mosby the promotion of accident-free driving. Company. Journal of Studies on Alcohol, .lQ, 161-167. Rogers, C. (1975). Empathic: An unappreciated way of being. The Zimberg, S. (1982). Psychotherapy in the Counseling PsycholoS>ist, 2t 2-10. treatment of alcoholism. In E. Pattison & E. Kaufman (Eds.)., Encyclopedic Roizen, J: (1982). Estimating alcohol handbook of alcoholism (pp. 999-1010). involvement in serious events. In New York: Gardner Press. National Institute on Alcohol Abuse and Alcoholism, Alcohol Consumption and Related Problems, Alcohol and Health Monograph No.1, DHHS Publication No. (ADM) 82-1190. Washington: Government Printing Office, 179-219. Royce, J. E. (1981). Alcohol problems and alcoholism. A comprehensive survev. New York: The Free Press. Shore, E. R., McCoy, M. L., Toonen, L. A., & Kuntz, E. J. (1988). Arrests of women for driving under the influence. TournaI of Studies on Alcohol, .12, 7-10. Staub, G. E., & Kent, L. M. (Eds.). (1973). The para-professional in the treatment of alcoholism (A new profession). Springfield, IL: Charles C. Thomas. Strupp, H. H. (1973). On the basic ingredients of psychotherapy. TournaI of Consulting and Clinical Psychology, 1113-15. Vargas A. M., & Borkowski, J. G. (1982). Physical attractiveness and counseling skills. TournaI of Counseling Psycholofm .£?, 246-255. 110 The Fight for Victim's Rights in Florida: The Realities of Social Reform Susan Louise Sayles, PhD. Events, ideologies and social changing social trends in America by reform movements do not occur in an examining the space provided to them historical vacuum. External forces and in local newspapers. Because the events pave the way for synergistic amount of space provided for news ideas to take root. Sometimes, all the items remains relatively constant, factors which appear to be necessary popular causes, i.e., reform movements, for a thought to become nurtured and generally are allotted only so much take fruition are present, and yet the space. Newspapers are continually fragile idea is unable to germinate. adding and subtracting popular causes. Many forces present during the This establishes a forced choice, and late 1960's and earl .-v 1970's nurtured one becomes able to measure the the formation of the victim's rights "share of the market" commanded by movement. The setting was ripe for competing societal concerns. Societies, change. This dissertation explores why, like human beings, can only stay after centuries of neglect, the victims of concerned about a relatively short list crime finally became a popular cause, of issues at a time. The most striking and how a coalition group, the Florida problems rise to the forefront, while Network for Victim Witness Services, other issues spiral downward into facilitated this development in Florida. ultimate oblivion. Naisbitt found that The victim's movement in most social invention in the United Florida was spearheaded by factors States takes place in just five which were very similar to those "bellwether" states. California is the .key responsible for sparking its birth indicator state, Florida a close second, nation-wide. In the 1982 book, followed by Washington, Colorado and Megatrends, John Naisbitt explained Connecticut. All of these states have this phenomenon. Using content interesting mixtures of ethnic and analysis, Naisbitt (1982) monitored demographic populations which tend This paper is from Dr. Sayles' dissertation by the same title, Dr. Leroy C. Gould, major professor. Dr. Sayles received both her undergraduate (1973) and doctoral degree (1991) in criminology from the Florida State University. She also holds a Masters in Business Administration from Nova University (1985). Dr. Sayles was hired in 1973 as one of the first women investigators for the State Attorney's Office in Florida, and assisted in establishing several comprehensive victim-centered programs in Palm Beach County and throughout the State. In 1990 she received the Florida Network of Victim Witness Services, Inc., James Fogarty Award in recognition of dedicated and continuous leadership and outstanding achievement in the pursuit of victim rights. The Palm Beach Victim Rights Coalition presented her with the 1991 Mary S. Greene Award for "oulstanding and dedicated service to the victims of crime" during her keynote address at their annual victim rights luncheon. Dr. Sayles teaches in the Departme11t of SocioloSlJ and AnthropoloSlJ at the University of North Carolina at Wilmington. 111 Victim's Rights Sayles to be more willing to listen to differing of violent crime who faced "serious ideas, creative and innovative financial hardship." The major experimentation, and change. problem with crimes compensation was The victim's rights movement that it had too few funds and, what confinns Naisbitt's predictions: other funds were available were limited to states model their legislation and specific, extreme circumstances. No one organization upon Florida's example. could receive money for property Indeed, Florida recently led the nation damage or monetary losses. Only when it passed a constitutional medical injuries were reimbursed, and amendment recognizing the rights of then only at a fraction of the amount victims to be heard at all crucial needed. When a victim was related to hearings in the processing of criminal the defendant, either as a spouse or cases. child, reimbursements were, by law, The history of the Florida systematically denied. Doerner (1978) movement for victim's rights began found that victims reimbursed under with factors and forces congruent with these conditions, receiving small those that took place nationally. In payments and having to deal with a their infancy, programs for victims very slow, bureaucratic process, were distributed in various agencies, actually resented the meager sum more with differing concerns, all across the than if there had been no promises of state of Florida -- from law funds at all. enforcement and probation The collapse of the Law departments to county comprehensive Enforcement Assistance Administration facilities, mental health agencies and (LEAA) and related funding around prosecutorial offices. This brought 1982-83 initiated even fiercer considerable confusion and infighting competition among governmental and between agencies. In the beginning the grass-roots agencies. As a result of the battles centered around limited funding loss of funds, many programs perished. and were fought between the Some programs, having established institutionalized, Florida Prosecuting themselves as integral parts of the Attorneys Association (PP AA) and the criminal justice system, were able to less well organized, grass roots, cut back or to find more permanent service-oriented programs. funding within local communities. The creation of the Florida Agency heads found that the time Crimes Compensation Commission in involved in the crucial efforts of fund the Executive Office of the Governor in raising, left them little time and energy the mid-seventies was the Legislature's for administration and program confirmation that victims were a valid development. A more serious side political issue. The Commission was effect of chasing too few resources was established to raise funds via fines the over-use or staff and volunteers. imposed upon convicted criminals; the This caused a great many talented and proceeds of such revenues were to be dedicated agents of change to fall by placed in a fund to reimburse victims the way-side, the victims of acute 112 Victim's Rights Sayles burnout. haphazardly began to call upon one Prosecutors who, when funding another to discuss the inherent sources were available, encouraged and dilemmas caused by a severe lack of promoted witness management any previous examples (Gusfield, agendas, were now left to their own 1989). The Florida Network for Victim budgets to provide monetary support. Witness Services (FNVWS) evolved in Most State Attorney's offices abolished 1979 as a result of this need. The whatever previously-established FNVWS is an umbrella network programs existed in favor of hiring consisting of service providers, self additional support staff or prosecutors. help groups and individual survivors Any State Attorney, such as David from across the state. It is also Bludworth of the 15th Judicial Circuit, comprised of other "networks" and who left his or her program intact, was "coalitionsl! from across the state, which severely criticized for misappropriation include the Refuge Information of funds and lack of efficiency Network (RJN) and a loosely-based (Bludworth, 1990.; Braddy, 1991). Those sexual assault "network." The prosecutorial-based programs that expressed purpose of the organization remained concentrated on indicted is to provide training and technical defendants and witness facilitation for assistance, and to act as a unified mass persons who had filed cases. The to ed ucate the Legislature for the purpose of these programs was to keep funding, creation, and continuation of in better touch with witnesses, reduce the already-existing programs waiting time for canceled or delayed (McCormack, 1987). The latent, trials, and reintrod uce a more unExpressed motive is what Blumer outwardly appearing concern for the (1969) refers to as "peer group victim as a witness. The reasons for its counseling," and "esprit de corps." approval among some prosecutors was Later, the more "grass roots" victim the perception of an increased organizations, such as Mothers Against conviction rate, red uced victim or Drunk Driving (~vfADD), Justice for witness attrition, loss of contact, or Surviving Victims aSV), and Parents of hardships encountered during the trial Ivlurdered Children (POMC), joined the processes. Prosecutors, as elected Network to give the more officials, were not unaware that the professionally oriented programs, i.e., platform of victim concerns could be programs with paid staff, an emotional seen as a viable and successful and physical boost in terms of power argument towards reelection. and statewide recognition Because of the sporadic, 100sely (McCormack, 1987). formed foundation caused by a This dissertation examined the conglomeration of grass-root and leaders of the Florida Network and governmental agencies, a need found, as Blumer (1969) did, that there developed for some type of cohesive appeared to be a need and function for structure. Representatives of whatever typical personality types which agencies existed in the state included the charismatic, statesman, 113 Victim's Rights Sayles agitator and bureaucrat. Examples of (Gusfield, 1989). Some of the individual the agitators came with the loud, vocal leaders began to view themselves as "survivor" or feminist who became the representing the best interests of the squeaky wheel that needed to be oiled. movement, acting beyond the accepted They were instrumental in creating the scope of others. Resulting conflicts needed attention to the cause, but with illustrate Michael's theory of the "iron a few exceptions, did not stay within law of oligarchy" by the ruling the mainstream for long. Agitators, members of a movement (Michels, while very dedicated to the cause, did 1949; Selznick, 1966). not compromise, made enemies and The Florida coalition presents a burned bridges in order to get their pattern which is consistent with needs met. These tactics worked in the previous literature concerning short run but advocates found they coll£ctive behavior and social had to distance themselves from these movements (Turner and Killian, 1957; personalities in order to navigate the Blumer, 1969; Selznick, 1966; 1948; political waters. 1943; Messinger, 1955; Till y, 1978; Zald Of particular interest are the and Ash, 1969; Weber, 1968; 1947; individuals whose identities became so Perrow, 1972; Michels, 1949). A typical very ingrained within the cause that "life cycle" or "growth pattern" can be they were unable to let go even after seen in the formation, organization they had either lost their jobs or were process and subsequent development removed otherwise from the core of the of the FNVWS. This dissertation network. This is an indication of the examined the coalition work of the degree of "esprit de corps" which the FNVWS, documenting its gain in movement established. Although momentum, forced transitions, women still dominate both locally and bureaucratization, institutionalization, nationally, the feminists, who were advances, setbacks, and flirtation with largely responsible for the initiation of the dangers of cooptation crisis response centers, are largely (Staggenborg, 1986; Perrow, 1972; absent from present leadership. They Selznick, 1966). were the initial catalyst which gained Elias (1986; 1983) has noted that media and legislative attention. the progress which has been made by Because they were unable to victim advocates is largely symbolic in compromise, they were unable to work that new laws are on the books, and the system to their advantage, and advocate positions are available in thus, lost out to the more sophisticated most large metropolitan areas, but that statesmen and bureaucrats (Waters, the focus on victims is largely 1989; Dressel, 1987; Weber, 1968). rhetorical. Funding is not being Initially the FNVWS attempted afforded in large enough amounts to to establish a niche within the really accomplish any more than band governmental bureaucracy in order to aid treatment. Proof that the victim create job s(:curity for a newly movement has been truly ensconced emerging social welfare problem within the system will be evident when 114 Victim's Rights Sayles the money is actually provided or to become controlled through insuring adequate protection under the abstinence. law. Resources are extremely limited. The poor in this country have Witness an excerpt from the St. been used to an abusive criminal Petersburg times in May of 1991 justice system, and saw it as being the disclosing a rape victim's displeasure norm. Persons form the middle class over her remuneration from crimes were brought up to believe in the compensation: police and the courts as the upholders of their values. When unexpectedly "Anna Resnick, 75, is not ashamed to confronted with an unsympathetic, reveal that she was raped five years ago. What bureaucratic, often hostile court system, upsets her is the $28.81 check the state victim's many middle class participants felt assistance office just sent her as compensation. 'This is what they're offering me after betrayed and thus began to mobilize. five years,' Resnick said. 'I don't know what Taking this argument further, one can to do with it.' make comparisons between the The awards are based on an analysis of women's movement claim of rape as a the victim's medical bills, income, insurance, symbol of the oppression and and ability to pay. The money comes from domination of men, and other reform convicted criminals who are ordered to pay restitution to a trust fund for the victims." movements in this countris history. It was women who spearheaded the child Many victims would be satisfied saver movement at the turn of the knowing that the state provided no century (Platt, 1969), the prohibition compensation rather than have an movement (Gusfjeld, 1963) and the incorrect anticipation of adequate asylums for the poor and mentally ill reimbursement. In reality, it cost the (Rothman! 1971). state more than the amount reimbursed Perrow (1986) noted that, like to cut the check and process the claim. people! organizations experience II drift. II Many similarities exist between Unforeseen consequences of what the victims' movement and other past seemeJ to be harmless concessions, de~ reforms. Marwell et al. (1988) stated rail the original goals of an agency. The that movement mobilization tends to danger of this type of "evolution" is be more successful when participants that it eventually erodes the autonomy were related to each other through of the institution. Those organizations social circles and friendships. who fail to make concessions do not Essentially, major figures behind the survive. Thus! a ICatch~22" situation victim's movement came from occurs cClusing idealistic goals to be members of the middle class. This doomed to eventual cooptation. In mirrors G~sfield's (1963) analysis of Florida, the Attorney General's Office the Prohibition reform movement in has taken over almost all of the the 19th and 20th centuries. Prohibition Statewide victim issues including became the symbolic reform of this crimes compensation. The Governors's dominant population, with the intent of Office retains one advocate that was urging the lower classes to assimilate, not statutorily moved at the time of the 115 Victim's Rights Sayles consolidation in 1991. The FNVWS has Gusfield, Joseph R. circumvented some of the problem by 1963 Symbolic Crusade: Status Politics and removing the "heavy hitters" from the the American Temperance Movement. Urbana: University of Illinoj~ Press. board and having them placed as Killian, Lewis M. "advisory" members. The reality of the 1964 Social Movemellts. In R.E. Faris (ed.). victim's movement as social reform, Handbook of Modern Sociology, unfortunately, is that it is destined to Chicago: Rand McNally. become one more institutionalized, symbolic, and rhetorical arm of McCormack, Robert 1987 Marlene A. Young, executive director government. of the National Organization for Victim Assistance (NOV A). Law Enforcement News, A Publication of John Jay College of Criminal Justice/CUNY REFERENCES 13(259) November 24. Marwell, Gerald, Pamela E. Oliver and Ralph Bludworth, David Phral 1990 Telephone Interview, July 5. 1988 Social networks and collective action: 1991 Personal Interview, April 25. A theory of the critical mass. III. American Journal of Sociology Blumer, Herbert 94(3):502-34. 1969 Social movements. In Studies in Social Movements, Barry McLaughlin (ed.). MeSSinger, Sheldon L. New York: The Free Press. 1955 Organizational transformation: A case 1971 Social problems as collective behavior. Social Problems. study of a declining social movement. American Sociological Review 20:3-9. Braddy, Edward Lamar Michels, Roberto 1991 Personal Interview, July 1. 1949 Political Parties. Illinois: The Free Press. Doerner, William 1978 An examination of the alleged latent Naisbitt, John effects of victim compensation 1982 Megatrends. New York: Warner Books. programs upon crime reporting. LAE Journal 41:71-80. Perrow, Charles 1972 Complex Organizations: A Critical Dressel, Paula Essay. Glenview: Scott, Foresman and 1987 Patriarchy and social welfare work. Company. Social Problems 34:294-309. Platt, Anthony M. Elias, Robert 1969 The Child Savers: The Invention of 1983 The symbolic politics of victim Delinquency. Chicago: The University compensation. Victimology: An of Chicago Press. International Journal 8:213-24. 1986 The Politics of Victimization: Victims, Rothman, David J. Victimology and Human Rights. New York: Oxford University Press. 1971 The Discovery of the Asylum: Social Order and Disorder in the New Republic. Boston: Little, Brown and 116 Victim's Rights Sayles Company. Selznick, Phillip 1943 An approach to a theory of bureaucracy. American Sociological Review 13:25-35. 1948 Foundations of a theory of organizations. American Sociological Review 8:47-54. 1966 TV A and the Grass Roots: A Study in the Sociology of Formal Organiza tion. Berkeley: University of California Press. Staggenborg, Suzanne 1986 Coalition Work in the pro-choice movement: Organizational opportunities and obstacles. Social Problems 33:374-90. Tilly, Charles 1978 From Mobilization to Revolution. Reading: Addison-Wesley. Turner, Ralph H., and Lewis Killian 1957 Collective Behavior. Englewood Cliffs: Pren tice-Hall. Waters, Malcolm 1989 Collegiality, bureaucratization, and professionalization: A Weberian analysis. American Journal of Sociology, 94:945-972. Weber, Max 1968 Economy and Society. Guenther Roth and Claus Wittich (eds.), Vols. 1 and 3. New York: Bedminster Press. ZaId, Mayer N. and Roberta Ash 1969 Social movement organizations: Growth, decay and change. In Robert Mclaughlin (ed.), Studies in Social Movements. New York: The Free Press. 117 l!:::=11==APP=ENDIX=A ==~ Inwald Personality Inventory (IPI) Scales IPI SCALES ELEV ATED SCORES LOW SCORES Purpose may indicate: may indicate: 1. VALIDITY SCALE 1. GD (Guardedness). To identify Guardedness, denial, desire to Candor, self-awareness, ability to persons who have minimized appear socially "correct." Possibly admit to minor shortcomings, short-comings, denied faults, and invalid profile due to maturity regarding self- answered items in a "socially defensiveness of responses. Lack assessment. desirable" manner. of insight. II. "ACTING OUT" BEHA VIOR MEASURES A. SPECIFIC "EXTERNAL" BEHAVIORS 2. AL (Alcohol). To identify Alcohol abuse, admission of Denial of habitual use of alcohol. individuals who may use alcohol habitual and/ or excessive use of Infrequent use of alcohol. excessively. alcohol. 3. DG (Drugs). To identify Drug abuse, admission of habitual Denial of habitual use of drugs. individuals who may use drugs and/ or excessive use of drugs. Denial of experience with drugs. excessively. 4. DV (Driving Violation). To A pattern of moving violations, No evidence of driving violations, document problems in adhering io fines, accidents, and/or possible or difficulty following rules the law with regard to the disregard for the law regarding related to motor vehicles. operation of a motor vehicle. motor vehicle regulations. 5. JD (Job Difficulties). To indicate Past difficulties holding jobs, Little apparent difficulty obtaining those persons who have had a history of being terminated from or keeping employment. No history of difficulty in the area of job, problems with coworkers or evidence of job adj'.Istment job performance. bosses, and/or spotty employment difficul ties. record. 6. TL (Trouble with the Law and Antisocial behaviors and incidents Denial of a pattern of anti-social Society). To id.;!ntify any history of including arrests, convictions, behaviors in conflict with societal brushes with the law and with school suspensions, fist fighting rules and norms. sccial norms. and other brushes with authority. 7. AA (Absence Abuse). To Pattern of difficulty meeting job Little evidence of absenteeism, document tendencies towards responsibilities regarding frequent minor illness, or lateness excessive absence and/or lateness attendance and / or punctuali ty. on past jobs. on the job. History of abusing sick leave privileges may be evident. B. ATTl TUDES AND TEMPERAMENT 8. SA (Substance Abuse). To Impulsivity, thrill-seeking, No apparent tendency towards identify individuals with willingness and tendency to take risk taking or gambling. May be a substance abuse and risk-taking risks, pattern of a gambler who mature individual who is unlikely tendencies. may be a substance abuser as well.. to abuse substances. A - 1 Appendix A Inwald Personality Inventory (IPI) Scales IPI SCALES ELEV ATED SCORES LOW SCORES Purpose may indicate: may indicate: 9. AS (Antisocial Attitudes). To Antisocial attitudes, skepticism, Denial of antisocial attitudes in measure expressed antisocial feeling that those who break the conflict with the norms of society. attitudes which may affect rules in order to "beat the system· Positive and mature outlook behavior and compliance with are justified. regarding general need for social mles. authority and regulations. 10. HP (Hyperactivity). To identify Restless, overactive, impulsive Apparently average or below persons with impulsive, patterns of behavior. highly average activity level. Little hyperactive behavior patterns. energetic individual who may be evidence of tendency to overreact outspoken and/or impatient with or become unusually restless others. under stress. 11. RT (Rigid Type). To indicate a Impatient attitude towards others Flexible, nonrestrictedi patience quality of rigiditYi strict adherence for minor discretions, easily with minor shortcomings in self to rules, and inflexibility in irritated person who has tendency fu,d others. May show tolerance of interpersonal relations. to keep emotions hidden. others. 12. TA (Type "A"). To identify Driven, restless, competitive Non-competitive, relaxed attitudes characteristics similar to "Type A" personality with high standards towards work and recreation. personalities. and "workaholic" tendencies. May Little evidence of a driven, be prone to heart ailments or other aggressive personality. physical stress reactions. II. INTERNALIZED CONFLICT MEASURES 13. IC (Illness Concerns). To Frequent health concerns and/or No apparent tendency to develop document tendencies towards chronic physical symptoms or physical symptoms or illnesses physical ailments or unusual illness. Unusual or unwarranted under stress. Few complaints concerns about illness or bodily preoccupation with health issues. regarding current physical harm. Minor ailments may tend to condition. develop under stress. 14. TP (Treatment Programs). To Endorsement of at least one of No history of past therapy or identify persons who have had three items indicating past counseling, drug/alcohol past psychological counseling, psychotherapy, drug! alcohol treatment program participation substance abuse treatment, and/or treatment and/or history of taJrJng and/or taking of medications for mediation for nervous conditions. tranquilizers for a problem. an emotional problem. 15. AN (Anxiety). To indicate Admission of frequent anxiety and Denial of anxiety or symptoms individuals suffering from a high worry over problems. Poor coping related to chronic worry. May level of anxiety. skills in stressful situations may be I have sufficient stress coping skills. evident. 16. PH (Phobic Personality). To Symptoms characteristic of phobic Little evidence of phobic reactions identify a tendency towards patients in outpatient clinics. of avoidance behavior as coping fearful avoidance of situations not Strong anxiety reactions when mechanism. Denial of symptoms classified as dangerous or fearful facing situations such as crowds, common to phobic individuals. by most people. closed spaces, and / or travel. A-2 Appendix A Inwald Personality Inventory (IPI) Scales IP} SCALES ELEV ATED SCORES LOW SCORES Purpose may indicate: may indicate: 17. OB (Obsessive Personality). To Desire for a highly structured Denial or strong concern with identify individuals who are environment in order to function. details and order. May be able to particularly concerned with details Anxiety over details may be handle disorganized situations and order, who may react to evident, as well as anxiety when without excessive anxiety. unstructured situations with there is confusion or anxiety. disorganization. 18. DE (Depression). To indicak i Expressed discouragement and Little expression of dissatisfaction those persons who appear depression. perceived difficulty with life, or discouragement depressed and who exhibit a coping with daily stresses, and regarding the attainment of goals. generalized dissatisfaction with difficulty achieving personal goals. life. General dissatisfaction with progress in life. 19. LO (Loner). To identify Tendency to spend time alone, Apparently social and outgoing, individuals who tend to keep to have few close friends, and avoid may have a network of friends themselves, who may be "loners" such socializing with others. In and social contacts. with limited social contact with times of stress, human support others. network may be guite limited. 20. UE (Unusual Experiences/ Positive endorsement of items Denial of unusual or bizarre Thoughts). To document unusual rarely answered by non-clinical thoughts and experiences. Does experiences and though ts patients. May have unusual or not admit to symptoms commonly associated with bizarre thoughts and experiences. characteristic of thought disorders. emotional disturbance or thought High scores may be evidence of disorders. emotional disturbance. IV. INTERPERSONAL CONFLICT MEASURES 21. LA (Lack of Assertiveness). To Tendency to avoid confrontations Little apparent difficulty being measure level of assertiveness in with others. May be easily assertive in social interactions. social interactions. intimidated and lack assertiveness May be able to react to others in a in social interactions. straightforward manner. 22. ID (Interpersonal Difficulties). Indication of problems getting Denial of difficulty getting along To identify individuals who have along with others. May be moody, with and relating to others. May demonstrated a pattern of impatient, and demanding, with be able to work easily in a group interpersonal conflicts and strong expectations as to how setting. troubled relationships with others. others should behave. 23. US (Undue Suspiciousness). To Pattern of reacting with undue Little apparent suspicion of others identify a tendency to be suspiciousness regarding the or evidence of paranoid thoughts. unusually suspicious and wary of behavior of others. Tendency to May be open and easygoing. the motives and actions of others. feel few can be trusted. Paranoid ideas may be present. 24. FC (Family Conflicts). To Positive endorsement of items Few serious family conflicts or document a history of family indicating conflicts with family resentments evident. Apparently difficulties and personal conflict. members, resentment towards one stable family backgw.md. or both parents and/or relatives. A - 3 Appendix A Inwald Personality Inventory (IPI) Scales IPI SCALES ELEV ATED SCORES LOW SCORES Purpose may indicate: may indicate: 25. SC (Sexual Concerns). To Expressed hostility and/or Few endorsements of items identify individuals having confusion with regard to the indicating difficulty relating to or difficulty relating to members of opposite sex. May have some working with members of the the opposite sex. difficulty working with members opposite sex. of the opposite sex. 26. SP (Spouse/Mate Conflicts). To Admission of history of problems Denial of difficulties getting along document a history of troubled in relationship with spouse or with spouse or mate. relationships with spouse or mate. mate. Minnesota Multiphasic Personality Inventory (MMPI) Scales L = Lie F = Infrequency K Hs = Hypochondriasis D = Depression Hy = Hysteria Pd = Psychopathic Deviate Mf = Masculinity-Femininity Pa = Paranoia Pt = Psychasthenia Sc = Schizophrenia Ma = Hypomania Si = Social Introversion MacAndrews Alcoholism Scale Critical Items A-4 MMPI x IPI Pearson Bivariate Correlation Matrix IPI Scale MMPI Scale Critical Items L F K 1(Hs) 2(De) 3(Hy) 4(Pd) 5(Mf) 6(Pa) 7(Pt) 8(Sc) 9(Ma) O(Si) Mac Critical Items .73 -.32 .33 -,45 .07 .17 -.08 .05 .17 .24 .25 . .21 .26 .22 .28 I CD -.08 .60 .04 .21 .13 .07 -.03 -.08 -.16 -.14 -.04 .08 -.02 -.01 -.09 AL .20 -.22 .19 -.24 -.04 .01 -.12 -.05 -.05 .06 .06 .06 .22 -.03 .26 DC .14 -.20 .11 -.12 .u5 -.02 .02 .11 .06 .12 .07 .11 .20 -.07 21 DV .02 -.15 -.11 -.02 -.04 -.02 .05 .04 -.08 .12 -.07 -.08 -.07 -.07 .10 JD .27 -.24 .22 -.30 .01 .05 -.07 .03 .15 .11 .09 .07 .12 .14 .14 TL .24 -.25 .15 -.23 .00 .01 -.05 .15 -.05 .15 .06 .11 .20 .05 .29 AA .31 -.35 .24 -.36 .04 .07 -.08 .OD .13 .12 .11 .07 .19 .15 .21 SA .35 -.30 .23 -,41 -.02 .01 -.12 .01 .15 .10 .04 .08 .39 .02 .24 ::> ,42 '"d AS -.26 .35 -.64 -.12 .07 -.40 -.15 .12 -.01 -.01 .01 .33 ... ..35 .31 '"d ::> HP ,45 -.38 .29 -.63 -.12 .07 -.23 -.10 .16 .11 .01 -.03 .34 .19 .28 ro ::l RT .30 -.13 .26 .-48 -.19 .01 -,42 -.19 .01 -.03 -.05 -.07 .23 .29 .25 0. CJl TA ,42 -.30 .29 -.60 -.13 .12 -.26 -.13 .17 .15 .02 -.08 .21 .32 .23 S<" IC .46 -.17 .34 -.41 .19 .23 -.09 .01 .26 .18 .21 .16 .18 .28 .18 ::> TP .17 -.18 .13 -.11 .12 .09 .10 .14 .20 .15 .10 .07 .03 .07 .10 AN .38 -.25 .30 -.47 .00 .26 -.12 -.07 .16 .20 .17 .00 .03 .36 .04 PH .52 -.26 .36 -.56 -.02 .26 -.26 -.10 .21 .15 .18 .08 .02 .56 .13 OB ,42 -.26 .33 -.60 -.11 .17 -.30 -.19 .13 .08 .03 -.08 .16 .39 .19 DE .52 -.26 .42 -.51 .07 .39 -.11 .06 .28 .26 .30 .18 .13 ,47 .11 LO .23 -.07 .27 -.31 -.03 .26 -.25 -.04 .09 .06 .11 .04 -.05 .54 .05 UE .51 -.10 .44 -,45 .10 .22 -.23 -.04 .09 .11 .24 .26 .26 ,43 .22 LA .12 -.10 .19 -.08 .11 .21 .02 .06 .11 .11 .18 .11 -.28 .37 -.13 ID ,43 -.28 .35 -.59 -.08 .20 -.28 -.09 .12 .12 .04 -.02 .14 ,43 .20 US .44 -.22 .33 -.63 -.16 .13 -,45 -.18 .16 .00 .06 .01 .29 .41 .28 FC .32 -.14 .34 -.34 .05 .17 -.11 .14 .30 .17 .14 .17 .13 .22 .15 SC .29 -.03 .26 -.34 .01 .15 -.23 .00 .14 .06 .10 .13 .16 .30 .05 SP .27 -.15 .21 -.32 -.01 .15 -.12 .05 .2] .13 .11 .07 .14 .21 .12 n = 590 Pearson (r) values of: 0.07 significant at .05 level. 0.09 significant at .01 level. 0.11 significant at .00] level APPENDIX B Criminal Justice-Related Doctoral Research in Florida, 1991 Barlow, David. (1991). The Political Economy of Criminal Justice in the United States: 1948-1987. (Criminology, Florida State University) Barlow, Melissa. (1991). Popular Ideologies of Crime in the United States: 1947-1987. (Criminology, Florida State University) • Bullock, Carol. (1991). The Effect of Florida/s Post Furman Executions on Homicide Rates. (Criminology, Florida State University) Cable, Eric. (1991). Reconceptualizing Government Growth: An Examination of Three Alternative Models of Criminal Justice Expenditures in the American States for the 1948-1985 Period. (Public Administration, Florida State University) • Cintron, Myrna. (1991). An Assessment of an Evolving Control Strategy: Cocaine, the United States and Latin America. (Criminology, Florida State University) Daly, Charles. (1991). Coordination of Child Abuse Services: Beyond the Structural Determinants. (Social Work, Florida State University) Deitchman, Mary Ann. (1991). Factors Affecting Competency-for-Execution Decision Making in Florida Forensic Examiners. (psychology, Florida State University) • Gnffith, Tracy. (1991). Correlates of Police and Correctional Officer Performance. (Counseling Psychology and Human Systems, Florida State University) • Morgan, Kathryn. (1991). An Analysis of Factors Influencing Probation Outcome. (Criminology, Florida State University) Park, Won-Kyu. (1991). Patterns of Criminality in Postwar Japan: A Structural Perspective. (Criminology, Florida State University) Partyka, David. (1991). Analysis of the Link Between Delinquent Behavior and Reading Developmental Disorders in a Sample of Chronic Male Juvenile Offenders. (Psychology, Florida State University) Ruckman, Jr., Peter. (1991). The Confirmation of Supreme Court Justices: Constructing and Testing a Theoretical Model of a Dynamic Process. (Political Science, Florida State University) B-1 Appendix B Savestanan, Naras. (1991). Thai Police Cadets' Attitudes Toward Police Corruption in Thailand. (Criminology, Florida State University) 'Sayles, Susan. (1991). The Fight for Victim's Rights in Florida: The Realities of Social Reform. (Criminology, Florida State University) Shaw, Theodore. (1991). Offender Variables and Treatment Outcomes of Participants in a Residential Sex Offender Program. (Counseling Psychology, University of Florida) • Sollars, David. (1991). The Allocation of Law Enforcement Resources, Spillovers, and the Illicit Drug Market. (Economics, Florida State University) • Spooner, Shirley. (1991). DUI Offenders' Perceptions of Counselor Empathy and Helpfulness. (Agency, Correctional and Developmental Counseling, University of Florida) • Swenson, Cynthia. (1991). Perceived Control as a Predictor of Treatment Outcome with Chronic Adolescent Offenders. (Psychology, Florida State University) • Waryold, Diane. (1991). An Analysis of Differences in Role Perceptions Among Senior Patrol Officers and University Judicial Officers in Selected Public Higher Education Institutions in the United States. (Higher Education, Florida State University) NOTE: University Microfilms International annually lists all titles in Dissertation Abstracts. Copies of complete dissertations are available through University Microfilms, c/o The Bell and Howell Company, 300 North Zeeb Road, Ann Arbor, Michigan 48106-1346. • Asterisks indicate a paper based on the dissertation is included in this volume. B-2 '. ~, ' ;:'~+, .~. : ...: .,. ' ,,' .. \ .. ~ , "r \' ., , " .~. " , "''. " .f. ~ . . ~,.'''' , ~~,: .. , ,~ , " ',H ',' ,-' Academically Speaking.