FACTORS AFFECTING THE SEVERITY OF THE ‘ SENTENCES eF FEMALE OFFENDERS , ' Thesis for the Degree of‘ M. A. MICHIGAN STATE UNIVERSITY LINDA BETH GORNITSKY ' 1977 INF—FM FEW? FALI983 G}0L/C"// ABSTRACT FACTORS AFFECTING THE SEVERITY OF THE SENTENCES 0F FEMALE OFFENDERS By Linda Beth Gornitsky This study examined the effect of certain factors, extracted from presentence reports, on the severity of the sentences received by female offenders. It also investigated the effect of the recommenda- tion of the probation officer on the judicial outcome. The research was carried out in a county probation office in a midwestern state. Seven students coded the infbrmation contained in the 376 presentence reports, which dated from 1969 to l976, according to an established rating schedule. Two different types of designs were used: the first one was a description of the decision-making process and the variables included, and the second was a multivariate predictive one. The two multiple predictive techniques employed were multiple regression and discriminant function. In conjunction with these, two data reduction procedures, representing a rational and an empirical approach, were used to insure that the variables were orthogonal. The major finding was that the principal determinant of the sentence was the recommendation of the probation officer contained in the presentence report that he/she had prepared. Other variables which had a significant (p < .05) impact on the judicial outcome were Linda Beth Gornitsky the severity of the charge, the previous criminal history, the disposi- tion mode, and the defendant's living situation. From a methodological perspective, it appeared that the rational approach was more predictive than the empirical one and that the multiple regression and discrimin- ant function equations were identical in both content and predictive abilities. FACTORS AFFECTING THE SEVERITY OF THE SENTENCES OF FEMALE OFFENDERS By Linda Beth Gornitsky A THESIS . Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTERS OF ARTS Department of Psychology l977 Copyright LINDA BETH GORNITSKY 1977 To Harvey, who provided me with an infinite amount of love, support, and advice, and To my parents, who were always there when I needed them. ii TABLE OF CONTENTS INTRODUCTION ........................ Disparity ....................... Predispositions of Judges ............... Judicial Characteristics ............... Judicial Prejudices .................. Other Precipitants of Disparity ............ The Presentence Report ................ Rationale for the Present Research .......... METHOD ........................... Setting ........................ Subjects ....................... Design ........................ Procedure ....................... Instrument ...................... Data Collection Procedures .............. RESULTS .......................... Part I--Presentation of Data ............. Part II--Data Reduction and Decision Prediction . . . . DISCUSSION ......................... Part I--Summary of the Descriptive Results ...... Part II--The Decision Model and Its Implications . . . Part III--Nhere Do We Go From Here? .......... APPENDICES ......................... REFERENCES ......................... iii 124 Table acumen-boom TO IT 12 13 14 15 16 LIST OF TABLES Studies Relating the Race of the Offender to Judicial Sentencing ...................... Predictors of the Decision Criteria ......... Original Variables Contained in the Instrument . . . . Criminal Offenses and Statuatory Sentence Lengths Time Chart ...................... Coding Changes .................... Final Set of Variables Used in the Analyses ..... Absolute and Relative Frequences of Variables in the Circumstances of the Present Offense and Previous Criminal History Categories ............. Absolute and Relative Frequencies of Variables in the Defendant's Characteristics Category ......... The Absolute and Relative Frequencies of Variables in the Presentence Report and Court Proceedings Categories ...................... Contingency Table of Sentence by Recommendation of the Presentence Report ................ Contingency Table of Maximum Sentence by Recommenda- tion of the Presentence Report ............ Contingency Table of Sentence by Charge ....... Eigenvalues and Percent of Variance Accounted for by the T7 Empirical Factors ............... Factor Loadings of Variables when Severity of Sentence was the Dependent Variable .............. Variables Contained in Each of the l7 Factors and Their Respective Factor Loadings when Severity of Sentence was the Dependent Variable ......... iv Page 15 32 35 38 52 55 58 60 62 63 65 67 69 72 73 Table acumen-hook) 10 11 12 13 14 15 16 LIST OF TABLES Studies Relating the Race of the Offender to Judicial Sentencing ...................... Predictors of the Decision Criteria ......... Original Variables Contained in the Instrument . . . . Criminal Offenses and Statuatory Sentence Lengths Time Chart ...................... Coding Changes .................... Final Set of Variables Used in the Analyses ..... Absolute and Relative Frequences of Variables in the Circumstances of the Present Offense and Previous Criminal History Categories ............. Absolute and Relative Frequencies of Variables in the Defendant's Characteristics Category ......... The Absolute and Relative Frequencies of Variables in the Presentence Report and Court Proceedings Categories ...................... Contingency Table of Sentence by Recommendation of the Presentence Report ................ Contingency Table of Maximum Sentence by Recommenda- tion of the Presentence Report ............ Contingency Table of Sentence by Charge ....... Eigenvalues and Percent of Variance Accounted for by the l7 Empirical Factors .......... . . . . . Factor Loadings of Variables when Severity of Sentence was the Dependent Variable .............. Variables Contained in Each of the 17 Factors and Their Respective Factor Loadings when Severity of Sentence was the Dependent Variable ......... iv Page 15 32 35 38 52 55 58 6O 62 63 65 67 69 72 73 75 Table 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 The Rational Factors ................. Eigenvalues and Percent of Variance Accounted for by the 16 Empirical Factors ............... Factor Loadings of Variables when Recommendation of Presentence Report was the Dependent Variable Variables Contained in Each of the 16 Factors and Their Respective Factor Loadings when Recommendation of the Presentence Report was the Dependent Variable . A Comparison of the R;Square Values for the Complete and Reduced Multiple Regression Equations when Severity of Sentence was the Dependent Variable The Empirical Multiple Regression Solution when Severity of Sentence was the Dependent Variable The Rational Multiple Regression Solution when Severity of Sentence was the Dependent Variable A Comparison of the Variables Forming the Two Regression Equations when Severity of Sentence was the Dependent Variable ................ A Comparison of the Adjusted 375quare Values for the Two Multiple Regression Equations when Severity of Sentence was the Dependent Variable ......... Empirical Multiple Regression Solution when Recom- mendation of the Presentence Report was the Dependent Variable ....................... Rationa1 Multiple Regression Solution When Recommen- dation of the Presentence Report was the Dependent Variable ....................... A Comparison of the Variables Forming the Two Regres- sion Equations when Recommendation of the Presentence Report was the Dependent Variable .......... A Comparison of the Adjusted §7Square Values for the Two Multiple Regression Equations when Recommendation of the Presentence Report was the Dependent Variable . Nilks Lambda and Raos V Scores for the Variables in the Empirical Discriminant Function ......... The Discriminating Capabilities of the Empirical Discriminant Function ................ Page 76 8O 81 83 86 87 89 9O 92 94 96 97 Table 32 33 34 35 Page Nilks Lambda and Raos V Scores for the Variables in the Rational Discriminant Function ......... 104 The Predictive Ability of the Rational Discriminant Function ...................... l06 A Comparison of the Variables in the TWo Discriminant Functions ..................... l06 A Comparison of the Squared Canonical Correlations of the Two Discriminant Functions ......... l07 vi INTRODUCTION The crime problem in recent years has become increasingly salient and increasingly difficult to solve. From 1960 to 1972, the number of reported offenses and crime rates have continued to climb. During this 12 year period, the rate for murder and non-negligent manslaughter increased 70%, whereas the rates for forcible rape, robbery, aggravated assault, and burglary more than doubled (A National Strategy to Reduce Crime, 1973, p. 12). This augmentation in illegal activity is appar- ently felt in many walks of life. People of all ages and of both sexes are scared to walk alone at night (Hindelang, 1975) and citizen groups have begun to lobby for better street lighting and other com- munity improvements, and have organized neighborhood security programs. In a 1970 survey, 70% of the white respondents and 53% of the black ones felt that the system of law enforcement did not discourage people from committing crime. Thus a circular situation seems to have emerged--on account of the continually increasing crime rate and ineffectual efforts to halt it, citizens have become disenchanted with the correctional system and have begun to take precautionary measures on their own. However, this lack of confidence in political institu- tions (Election Time Series Analysis of Attitudes of Trust in Govern- ment, 1971) may have contributed to law-breaking because citizens do not recognize the legitimacy of the country's political institutions. Perhaps what all this is implying is that the criminal justice system is pleagued with problems and as a result, is unable to maintain order. This review is concerned with two areas of the justice system which have attracted considerable attention. These are female offenders and the sentencing process. The first topic is of major concern because the arrest rate is rising dramatically for women. In the same 12 year period considered earlier, the arrest rate among women rose nearly three times faster than it did for men (Adler, 1976). The sentencing process has also surfaced as an area of interest because as the back- bone of the correctional process, it should be operating swiftly and fairly, yet, in fact, it "has broken down under the burden of increased business" (A National Strategy to Reduce Crime, 1973, p. 93). Each of these topics will be discussed in greater detail. As we stated previously, there has been a dramatic increase in the number and types of offenses committed by women. According to the Uniform Crime Reports, in 1953, of the women arrested, 1 out of 12.8 were arrested for a serious crime. By 1973, the ratio had changed to 1 out of 4 women (Simon, 1975, p. 38). An interesting factor is that this increase in serious crimes was due almost wholly to a greater participation in property crimes such as larceny. In 1953, about 1 in every 20 arrests for women was for larceny, whereas by 1972, the ratio had shifted to l in every 5 (Simon, 1975, p. 41). The greatest in- creases in Type II offenses were for embezzlement and fraud and for fergery and counterfeiting. The offenses with the most dramatic differences between 1960 and 1972, were embezzlement (up 280% for women and 50% for men), larceny (up 303% fbr women and 82% for men) and burglary (up 168% fer women, and 63% for men) (Adler, 1976, p. 16). This same pattern was reflected in the 1974 Uniform Crime Report for the State of Michigan. 0f the women arrested in that year, for index crimes, 16.4% had committed murders, 6.1% robberies, and 58% larcenies. In Ingham county alone, in 1969, only 10 women comitted felonies whereas 46 women were arrested for felonies in the first 6 months of 1976 (Statistics from Ingham County Probation Office). Thus, the most obvious conclusion which can be deduced from these statistics is that r“: women are becoming active in crime, in general, and in white collar ones in particular. The second topic of concern is the criminal court system of the I United States. This has become the focus of much criticism recently I— because of its inability to function adequately. Specific faults Whi ch have been delineated by the National Advisory Committee on Criminal Justice Standards and Goals are: l) inconsistency in the processing of criminal defendants, 2) uncertainty as to the results attained, 3) unacceptable delays, and 4) alienation of the community (A National Strategy to Reduce Crime, 1973, p. 93). The first weakness is partially due to plea bargaining, a non-trial procedure which has been used in- consistently and has helped perpetuate sentencing disparities. Alie- I"ation of the public has occurred largely because the processes followed by court officials are not visible to the public. There is a definite 1 ack of conmunication between the outside and the court which has led to a cynical attitude on the part of citizens. According to a Gallop Doll conducted for Newsweek magazine (March, 1971), many American do hot have much faith in their courts. Citizen groups actively criticize the justice system for it's failure to be 'just' in it's treatment of criminals. This has resulted in a two-sided argument. Some groups claim that judges are too lenient and that this allows dangerous individuals to roam the streets and commit additional offenses. In contrast, more liberal thinkers argue that incarceration is not They "the" answer and that it, in fact, perpetuates criminality. insist that locking people up with no provisions for occupational and educational improvement is a temporary and "harmful" solution because once convicts are freed they are certainly no better off than before and are forced to return to their anti -social supportive In 1975, in Washington D.C., two out of three persons patterns . Six out of ten per- arrested for serious crimes were not convicted. sons who were arrested for felonies had prior criminal records. In Detroit, in 1975, of the total number of persons arrested on felony charges, 58% were convicted and of these only 20% were eventually Therefore more than half were returned Sent to jail or to prison. to the conmunity after being fined or placed on probationary status (U-S. News and World Report, May, 1976). One of the reasons why the courts seem to be operating in this state of confusion or mismanagement is that many different individuals act as sources of input thereby influencing the final decision reached. Thus the offender is channelled through a filtering system of sorts and at any branch in this network, a decision could be reached which Would dismiss the person or affect her eventual charge. For example, the suspect's initial encounter with the criminal justice process probably occurs when she is arrested by a police officer who is Suspicious of her behavior or possesses evidence suggesting her .involvement in a particular crime. The police have the choice of Thus Either invoking criminal proceedings or of dismissing the case. a wrong decision may result in liberating a guilty individual or detaining an innocent one. Another decision-affector is the prosecutor who, as society's representative in the court, is primarily responsible for determining the public's interests in each case and for charging the offender accordingly (Knudten, 1970). Thus he may decide to prosecute or not to prosecute depending on the circumstances of the case and depend- ing on his workload. According to a recent publication of U.S. News and World Mort (May, 1976), on the average, at least 24% of the pro- secutor's cases are dismissed because witnesses cannot be found and/ (Jr' [arosecutors do not feel they have ample evidence to prove that the defendant is legally and factually guilty. It is only at this point that the accused, if they have not t>¢a£ari previously dismissed, is confronted by the judge who has the l-I‘It'imate decision as to the length of the sentence. However this power is reduced considerably because of the former actions under- 't3Ei|‘Ierall sentencing structure by: l) prescribing the maximum penalties for criminal offenses; 2) establishing a partially indeterminate sentencing struc- ‘ture granting judges certain flexibility such as the power to suspend sentences; 3) establishing probation and parole services and giving the parole board the ability to determine the actual length of incar- ceration after the minimum has been set by the judge; and 4) ruling that all minimum sentences be within two-thirds of the legislatively set maximum (Palmer and Zalman, 1975). The second "modifier" is the probation officer who conducts the presentence investigation, a procedure required by a Michigan statute in all felony cases. , The information collected is then used to prepare the presentence report which is later reviewed by the judge and serves as a guide for determining the disposition of cases. Thus it would seem that the timeless proverb "Too many cooks Spoi l the broth" is particularly applicable to the judicial process Where the variety of officials, each with their own responsibilities a"(I decision-making capabilities, has resulted in a fairly incom- DY‘Ehensible and unorthodox fashion of sentencing. In summary, two major areas of the criminal justice system have been briefly examined. The first which concerned women offenders, revealed a definite and continuing trend in the increase in female pa r‘1:icipation in serious and white collar crimes. The second, which concerned the judicial system, exposed the courts as inefficient and plagued with uncertainty, inconsistency, and delay. These two Seemingly unrelated issues are connected by a crucial stepping stone Wtlich is that the sentencing practices of the courts determine the t)utcome of an offender, and the effectiveness of corrections. Thus, this study proposes to bridge these two areas by studying the factors which lead to severity of sentence in order to differentiate the effect of certain court practices, such as plea bargaining, on the defendants. Although this will not address the inmediate problem of why the crime rate is escalating for females; it will prOvide some much needed demographic i nformation' on female offenders and will re- veal sentencing practices as they relate to women. The review will now look at three of the major issues involved in sentencing. The first one, alluded to earlier, is disparity or inconsistency in sentencing practices within and between judges. The I: second is plea bargaining. aprocedure whereby the defendant and pro- _. secution negotiate the charge and sentence in a process of mutual advantage-taking. Finally, the third topic is the presentence report, a document prepared by the probation officer and given to the judge Dr‘i or to the trial, or hearing. It must be stressed that since there are no' studies specifically concerned with sentencing and it's rela- tion to women defendants, and in order to understand the global opera- ti on of the criminal justice system, the relevant issues will be ac[dressed by looking at their interaction with male offenders. \D 1. S Eari ty The disparity issue focuses on the existence of, and explana- t‘i ons for, inconsistencies in sentencing. Unfortunately the exact Qauses for this situation are in dispute, and much controversy has t1eveloped over this issue (A National Strategy to Reduce Crime, 1973). StJule of the factors responsible for disparity might be the previous experiences and social values of the judge, such as his prejudices, a11d the roles of the other court administrators. Each of these alter- natives will be examined separately. These will be followed by an examination of previous disparity-oriented studies which attempted r..- _Ie, to assess the prevalence of inconsistent sentencing in U.S. courts. Predispositions of Judges. It has been proposed that the decision-making process is affected by the attitudes an individual has previously acquired (Hogarth, 1971). If this is so, then it would seem likely that a person who is appointed to the position of judge has already formulated certain opinions and beliefs concerning . A ‘the world and his relation to it. In effect, he has developed broad ' ciispositions that serve as potentials for specific judicial attitudes that he will form in the future. The implication is that a judge, with mi ddle-class conservative values, would react negatively to female [IRKQ - offenders, in general, because they have violated his assumptions i31>c>ut the role and social behaviors of women, and especially to ones Who had other undesirable traits, such as numerous divorces or ille- Qi timate children. In other words, what were considered desirable attributes by magistrates was governed by their upbringing and the Viilues they had had instilled in them. Following this reasoning, it ‘“'<>uld seem useful to examine the backgrounds of judges to see what 1‘?3rpe of attitudes they might have had and, as an extension of this, 11|1e biases that might have influenced their decision-making processes. Judicial Characteristics. There were two contrasting theories Elrticulated by Haines (1922), which represented the two dominant t>eliefs about the place and function of the judge. The first, known iis the mechanical theory, proposed that the judge studied the facts (of each case and formulated his opinion solely on the basis of the information he had. Thus the judgment was not subject to any biases or idiosyncracies. The second, or theory of free legal decision, proposed that decision-making was a normative and subjective process. The judge's conclusion was expected to be influenced by the data at hand as well as by his previous experiences and believes, and by extran- eous conditions. These two theories, although coined some 50 years ago, still exist today (see for example, Karos and Mendelsohn, 1967) which may act as an indication of the pervasiveness of this state of controversy and of sentencing inconsistency. Proponents of the latter Itypothesis of free legal decision have studied the backgrounds of judges in order to demonstrate that variables such as age, ethnic and r121 igious affiliations, parental occupation, party membership, and education affected the decisions made by the subjects. One of the e2£1r~1 iest investigations into the experiences of magistrates was J. Schmidhauser's (1959) Collective Portrait of the Justices. He divided the years between 1789 and 1957 into six periods by following the general designation accorded each era by historians and then looked at the biographical data of the 91 judges who served on the Supreme CO urt during these time periods. In an attempt to determine the social status of these judges, he recorded the following variables-- Dianternal occupation, patterns of occupational heredity, ethnic origin, bejigious affiliation, educational background, and political party. ‘T‘}\£3 portrait he formed after collecting the data was that the typical SS‘JIDreme court justice had been white Protestants from socially pres- ti geful and politically active families whose ethnic stock originated ‘i'1 the British Isles. In addition, they all had received law training D"'ior to their appointment and many had attended Ivy League schools where they completed their university and law degrees. This i nforma- 'tion needs to be regarded as descriptive and perhaps even a bit tentative since it was extracted from biographical expositions on 10 these men. No mention was made of exactly which sources served as references, nor of how the data was collected, or who collected it. The significance of Schmidhauser's (1959) report must not be underestimated because his findings enabled theorists who were con- cerned with sentencing decisions and disparity to speculate on the probable environments judges were exposed to, and thus the attitudes ‘they might have been expected to develop. The next step was to see udiether or not these values affected the decisional propensities of the judges. One of the most important studies in this area was conducted by Nagel (1962). The sample consisted of 313 state and [p]... federal supreme court judges listed in the 1955 Director of American Judges. Background characteristics were determined by consulting four sources: The Director of American Judges, Who's Who In America the Martindale-Hubbell Law Directory, and the governmental directories published by the state. The factors extracted from these references Were political party affiliation, pressure group affiliation, education, ,age, geographical location, religion, and pre-judicial occupations. Judges were given a decision score which represented the proportion of tiInes he voted for the defense out of all the decisions he made on the full court criminal cases he heard in 1955. This score was then '“etched with the attributes of each judge. Thus the major findings were that judges with higher decision scores tended to be Catholics, Democrats, and unaffiliated with the Anerican Bar Association. Based on these three factors, the author concluded that "there will probably always be some correlation between judicial characteristics and judicial decision-making." (p. 339). Once again this statement must be regarded as conditional because of the numerous methodological 11 shortcomings. Like in the previous study by Schmidhauser (1959), there was no mention of how the data was collected, or by whom, or if any reliability measures were determined. In other words, how reliable was the information gathered on each subject? From a more computational perspective, a major flaw was that each variable was considered in isolation. Thus if education on it's own did not lead to a low decision score, it was assumed that it did not affect the judicial outcome. However, it's effect may have been masked or con- founded by a third extraneous variable. One possible solution'to this problem would have been a multivariate predictive technique. Such an approach was assumed by Bowen (1965) who replicated most of Nagel's (1 962) results and then subjected them to a multiple regression analy- sis - It was discovered that the maximum amount of variance among 3 udges accounted for by any single background factor was 16%. Judicial Prejudices. Previous researchers considered the back- Q‘F‘Ound characteristics of judges from a descriptive viewpoint (e.g., S(Ihmidhauser, 1959) and/or from a decision-making perspective (Goldman, 1 965; Nagel, 1962; Schmidhauser, 1961). Some studies have narrowed the focus of their investigation by considering the effect of social- ‘i Zation on one aspect of the judge's personality, namely his prejudices. This is an area of concern, and of controversy, because theorists such as Chambliss (1969), Sutherland and Cressey (1970), and Burke and Turk (1975) have claimed that socially disadvantaged persons, defined as those having low SES or a minority racial or ethnic member- ship, were more likely to be severely penalized upon conviction. If this were so, then proponents of the existence of disparity would have some support for their case. I". '— ; In I; 1- // . 12 Martin (1934) studied the relationship between the social traits of 10% of the felony cases disposed of in the district courts of Texas in 1930 and the judical outcome. He found that the courts favored native Americans Over Negroes, Mexicans, and others, those engaged in trade over those in lower-grade occupational categories such as mech- anical and domestic services, property owners over nonproperty owners, I ”7 married men over single men, and fathers over childless men. Age, sex, education, and presence or absence of parents did not affect the sentence. Martin's conclusion that outcome was biased by racial pre- judice was not substantiated by his data. Besides for the serious ; If’law that no precautions were taken to insure that his data was rpealiably collected, no attention was paid to the legal factors such its; the circumstances of the crime or the criminal history of the (leafendant. Since members of the ethnic minority groups in Martin's Sample comitted the more serious offenses, the fact that they also received the harsher sentences does not reflect racial prejudice. Fr'ilnally, the limited number of factors considered by the study as possible determinants of the outcone also restricted the impact of 1t11ea conclusions. These same shortcomings surfaced in Lemert and Rosbergs (1948) investigation of the differences of penalties dealt ()Lrt to white, Negroa. and Mexican offenders in the Superior Court of Los Angeles, in 1948. The authors concluded that the judges were triased because whites received milder sentences than either of the (Ither racial groups. Once again, no controls existed for the degree 13f recidivism. The researchers themselves demonstrated the serious- ness of this particular omission in a later study, where, when a control was imposed for previous offenses, the differences in penalties l3 turned out to be nonsignificant. It must be stated that other cone trols, such as for the circumstances of the offense, defendant's characteristics, and court-related procedures, should have been enforced. Bullock (1961),in 1958, gathered information on inmates in the Texas State Prison in Huntsville from the Prison Classification and F'--‘ i- Identification Department. The sample consisted of 3,644 white and black inmates who had been convicted of burglary, rape, and murder. The six variables which were considered were: race, type of offense, riumber of previous felonies, disposition mode, county from which he teas committed, and the sentence. He found that the correlation between reace and length of sentence remained strong regardless of any of the Other variables studied. Type of offense, guilty plea, and area of "Eesidence also had a significant effect on the length of the disposi- ti on. It must be noted that although Bullock is conmonly referenced \nlfien discussing studies concerned with racial prejudice, the decision- maker was a jury and not a judge. In this way, it is unlike it's F>raedecessors. However, this work is like its predecessors in that it cii«d not specify how the sample was selected, who recorded the data, Elrrd how it was recorded. In addition, by simply constructing dicho- tomous tables and chi-squares on the data, he could not study the iliteraction of the variables on the dependent one. Once again, a rnultivariate approach would have accomplished this. The three studies just presented by Martin (1934), Lemert and Rosberg (1948), and Bullock (1961) all concluded that racial prejudice ‘was a significant factor affecting the severity of sentence. ’However Green's (1964) research which controlled for some of the items 14 neglected by the previous three experimenters, found no differences in the length of sentences according to racial denomination. Green selected 118 cases of robbery and 291 cases of burglary from a possible sample of 1437 consecutive cases disposed of by conviction in a criminal court in Philadelphia. He noticed a definite discrepancy between the type of crimes, number of previous offenses, and seriousness of prior record for Negro and white offenders. When these differences were controlled, by holding a particular variable constant and looking at only those criminals who possessed that variable, the seemingly dis- criminatory sentencing pattern disappeared. For example, if Negro and evrfite defendants convicted of burglary and having no prior felony (:liarges were compared, then distribution of sentence severity was fitegligable (p > .80). Similarly, when the number of prior convictions ‘VVEare controlled for other crimes, the results showed no traces of "Eicial discrimination in sentencing. The overall conclusion reached was that this particular court did not differentiate the seriousness (>1F ‘the crime according to the race of the offender. This work had the Siiilne methodological faults as all the other ones in that the data collection procedures were ignored. Only three independent variables Were considered which were type of robbery and burglary (armed vs unarmed), number of bills of indictment, and prior convictions. F=i1ially, the only statistical analysis used was percentages. No cflii-square or measure of association was done to determine whether the differences were statistically significant. To review, three out of four studies delineated concluded that One of the causes of disparity was racial prejudice (see Table l for a Comparison of the feur studies). However they all shared similar 15 mcowuop>coo cowcg .ucmsuuwuc_ to m__ec eeeepoeep_;a .mmcmemo xngmczn pmm mwmopcmocmo P_< to waxy muzmpcmm xgonnoc mFF Aemmpv :mmcw woos aucaou.muoe :o_ppmoamec cowu_moamwv .xucaou .mmpcopme cowumwuommm .mmwcopme msow>mcq to tenuous maow>mca .maxp mmcmmeo montage :omwca .mcczcmiwsu .mazp mmcmeeo .muem mucoucmm mmxmp eem.m Apmmpv xooppzm mmpmmc< mod Amempv economiesu mcoz moan mocmucmm momma ¢_m pm pm “swam; .cmcupmgo mo conga: .mgcmcon to mucmmmcg .zowacusum .mspoum ”mavens .cowpeazouo ommp .moxmh monogamocmo ocoz .wmm .mumm mucmacmm momma mum Aemmpv septa: mmmapoc< mmpaewse> mopncwce> mpnmwcm> mPQEem Pcuwumwumpm cmpposucou pcmucmamccH pcmucmawa agoewco xvsuw mcwocmpcmm perupuao ou.emv=meeo ms» co mama one acruepmm mmpusum N E: 16 nethodological and statistical problems which seriously question the accuracy of their results. Briefly, these were: 1) unreliable data since no reliability checks were done to insure that there was consis- tent interpretation of the data; 2) poor data sources; 3) no, or limited, consideration of the effects of other variables on the dependent variable and on the independent variable of interest; and 4) no consideration of F”? the effect of the predictor variable on the criterion in isolation of other confounding factors. Thus far, two possible contributors to disparity were examined. ' lliese were the background characteristics of the judge and judicial Prejudices. However, disparity may be perpetuated by other factors, £3l1<3h as individualized sentencing, plea bargaining, and the presentence "Eezaort. Each of these will be briefly considered within the context 01’ disparity. The latter two will then be explored in more depth in the last two sections of the review which are devoted to the topics of '31 ea bargaining and the presentence report. Other Precipitants of Disparity. It is unfair to place the onus (5“: 'the disparity problem on the judge alone, since often he is forced 15(3 Ideal with inconsistencies in the procedural aspects of sentencing. Eind in the decisions made by other court administrators. For example, ‘1 punpular notion in some areas is individualized sentencing which dic- tates that the judge sentence according to the circumstances of the particular case and the defendant's history. Since this demands that the judge use his own discretion, and since no two individuals think alike, it is to be expected that a certain amount of guesswork and ambiguity will evolve. The disparity resulting is a function of the SVStem and not the magistrate. 17 Plea bargaining is also felt to contribute to inconsistent out- comes because this procedure involves a negotiation between the two attornies which is devoid of any judicial input. The judge is then practically (though not legally) required to adhere to the decision if the utility of this procedure is to materialize. This suggests that in some circumstances the sentences decreed by judges are not their own, thus accounting for some discrepancies in decisions. There is some research evidence from both England and the United States that the courts tend to "follow" the recommendations of the pro- bation officer in the presentence report.(carter andWilkins, 1967;. Hood, ‘15966). In addition, many judges have publicly admitted that the pre- sseelitence report is an essential document without which an adequate <1€3<2ision would be impossible (Hogarth, 1971). These two facts imply t:hat if the presentence report is not prepared properly, or if the Officer makes a hasty suggestion as to the severity of sentence an <3”f’1Fender deserves, there may be variations in the decisions formed on the part of the judge. Finally, before leaving the issue of disparity, it is necessary 13¢) Ido two things: 1) examine those studies which consider disparity, Eil1cl 2) consider what effect any of the possible contributors to disparity Suggested earlier, such as judicial personalities or plea bargaining, Would have on the sentence if they were all included in the design. Perhaps one of the earliest studies addressing the first goal was Everson's (1919) comparison of the sentencing records of 42 New ‘Vork Magistrates sitting in rotation in 28 courts in the year 1914. 'The Committee on Criminal Courts went over the records of approximately 155.000 cases of summary violations or local ordinances and the results 18 collected led Everson to conclude that "justice is a very personal thing." (p. 98). This result was unfounded for many reasons. For one, there was no verification of the types of offenders each magistrate saw. ' Thus one may have had a disproportionate number of offenders who had committed serious or nonserious crimes thus explaining the unequal dis- tribution of sentences. Secondly, no information was supplied about Team the Committee on Criminal Courts nor the methodology utilized. In fact, ; the report was so general that the exact number of cases recorded was ‘ not specified. Finally, the outcomes were considered without attending 3 to any other variables such as the previous criminal history of the . subject which conceivably might have affected the sentence. Thus it LMI vvcauld seem that very little could be concluded from this article. Interestingly, one of the most widely cited and influential of ii?! 1 American studies of disparity sentencing was Gaudet's research of 1 938 (according to Green, 1961) which was also susceptible to the same 'fialults as Everson's (1919) earlier work. The data was extracted by a 1 aw student from the records of the Court of Common Pleas of one county 1' n New Jersey over an anonymous lO-year period. The information col- 1 ected on each subject was the disposition mode (jury vs. nonjury), the name of the sentencing judge, the charge, the plea, the sentencing date, and the sentence imposed. The prisoners who were to be sentenced were assigned to judges by the prosecutor on a rotation basis. All offenses Vuere divided into four categories--sex crimes, property with violence, Firoperty, and violation of state liquor laws. Gaudet insured that there \Nas proportionately equal distribution of these four groups among the Six judges, and then made the assumption that the cases heard by the magistrates were of similar gravity. Based upon this reasoning, he y / 19 then concluded that the discrepancies in sentences which appeared were caused by the personalities of the judges. Gaudet made a number of serious methodological errors and certainly overstated his case. For one, he assumed that the caseloads of the judges were about the same with respect to the proportion of serious and minor crimes and the proportions of first offenders and recidivists, an assumption which was not verified nor likely to have transpired. For another, the time factor greatly confounded the results. There was no indication of how long each judge served nor what proportion of the cases seen corresponded to each year served. The number of individuals sentenced by four of the judges was approximately '3-5 times greater than itilose sentenced by the remaining two judges. This is important because Si gnificance of results varies with the size of the sample. The impact <>”f’ the temporal factor was suggested in the study itself, because when 1tfice type of offense and year were held constant, the sentencing patterns (>1F’ the judges were no longer distinguishable. The fact that this was 31 'longitudinal study also necessitated that certain Confounding external \Iiir~iables be controlled, or at least documented. For example, senten- <2firlg may have become more or less lenient within certain periods over the lO—year time span due to public reactions or historical events. Essentially, this points to the importance of studying the effect of FILnnerous influences on the outcome and of not assuming a restrictive lunidimensional perspective. Gaudet did record certain legalistic \Iariables but did not include them in his analysis. Finally, there was ‘no connection‘whatever between the data and the conclusion that person- ality variables were accounting for the discrepant outcomes. He failed to show that disparity existed and he had no measures of the judge's 20 characteristics or background. The statistical analyses employed also did not allow for causation conclusions because Gaudet simply compared the percentage of sentence types chosen by each judge. Two of the more recent studies (Baab and Furgeson, 1967; Green, 1961) addressed the second issue of what effect do the extra-legal characteristics of the offender, the legalities of the case, and the r... other factors mentioned such as the presentence report, have on the outcome? These last studies were of particular interest because they assumed a more multidimensional approach which was also followed by this research. For this reason, they were examined in considerable A. depth. In Green's (1961) study, the sample consisted of 1,437 convic- t‘ions recorded in a non-jury prison court of the Philadelphia Court (>1F’ Quarter Sessions, within a 17-month period during the years 1956-57. The data was derived from court and police records of the city of F’li'iladelphia. The variables looked at were divided into three sets: 'Iegal factors, legally irrelevant factors, and factors in criminal F>Ietasecution. The first group consisted of 1) the type of crime <2<>nmnitted, 2) the number of current charges, 3) the prior criminal history, 4) the recomendations of auxiliary agencies of the court. Legally irrelevant factors were composed of demographic characteristics Such as 1) sex, 2) age, 3) race, and 4) place of birth. The last area concerned the personnel participating in the trial--i.e., the judge and prosecuting attorneyuand the type of plea entered by the defendant. 'The statistical analyses used to test the hypothesis that two or more groups differed significantly. with respect to the distribution of Penalties imposed was the chi-square test. The variables of the first 21 group tended to have a significant effect on the sentence imposed. The nunber of current charges and especially prior convictions of felonies had a striking effect on the severity of sentences (p < .001). However, as the cases mounted in severity, the effect of the previous criminal history declined and the offense itself became the. foremost determinant of the judicial outcome. In the non-legal factor category, he found that youthful offenders were favored over older ones and whites over Negroes. Finally, neither the court officials nor the disposition mode had any significant effect on the sentence. Green concluded that J udges tried to comply with the mandates of the law while accomodating the various factors which they regarded as important. Although this WOrk was a significant improvement over the earlier ones in scope and cOmprehensiveness, it was restricted to looking at single variables or Variables with one other factor controlled. No attenpt was made to See how §_l_l_ the variables affected the dependent one or which one(s) aQcounted for the most variance. Again, no information was provided Q<>ncerning the data collection technique thus rendering the accuracy of the information questionable. The last study which was reviewed in depth was the one by Baab and Furgeson (1967) because it also assumed a plural approach. The ‘3 nformation was collected from 27 courts from July-September of 1966. The courts were located in counties of various sizes which represented di fferent median income levels, racial and ethnic mixtures, and economic capacities. The dispositions were recorded in each chosen district court every other month and the final sample consisted of 1,720 felony cases. The factors that represented the elements in the administration of the criminal justice system were pretrial freedom, type of defense 22 counsel, and disposition mode. Of these, only the first two proved statistically significant implying that offenders who were granted bond and who had retained defense attornies were more likely to receive lenient sentences. The second set of factors related to the individual characteristics of the offenders. These included the type of offense, previous felony and misdemeanor convictions, age, marital status, edu- cational level, sex, and race. It was found that only the severity of the offense and the number of prior felony convictions influenced the Judicial outcome. The methodology used in this study was multiple regression, thus making it one of the first pieces of sentencing- V‘e1ated research to use it. However, no analytical results, such as regression equations were supplied. The authors specified .02 as the s ‘3 gnificance level, but they did not delineate what this applied to. Di (1 they use a step-wise or a direct regression procedure? What were the R-square values? No mention was made of which variables accounted For the most variance and, in fact, the results were presented as if a uh ‘i‘variate method, rather than a multivariate one, had been employed. Fi nally, the same nethodological flaws existed in this work since questions regarding the source of the data and the collection proce- tiure were left unanswered. For example, no information was provided about the specifics of the communities where the data was gathered, S'«Ich as the approximate size, location, ethnic mixture, and income 1 evel. Thus it would be impossible to replicate or evaluate this Study. .- To quickly review the issue of disparity, studies have explored the possibility that background variables and prejudices of judges, court administration elements, and extra-legal factors of the offender 23 might act as predictors of the severity of sentence. It was believed that this multitude of factors impinging on the judge would lead to discrepant sentencing patterns, especially if individualized sentencing was operative in that court. It was also pointed out that many of the conclusions developed by these works could not be accepted conclusively because of the numerous methodological and statistical problems. To repeat, the conmon flaws were: a failure to adequately describe the data source and the data collection techniques; a restricted number of independent variables; a failure to consider the impact of a number of variables on the criterion; and a failure to partial out the effect of extraneous variables. The review will now turn to a consideration of plea bargaining. I 12 must be mentioned from the outstart that no research is cited in th is section or the preceeding one on the presentence repOrt. This is be cause no studies, other than the few just reviewed, have discussed these t0pics from the sentencing perspective. p\1ea Bargaining Plea bargaining has become, in recent years, an indispensible l3art of the justice system. An indication of its relative importance was the fact that in New York County, in the late 19505, literally thousands of pleas of guilty, or compromises were effected each year On felony indictments. Ninety to 95% of the cases were disposed of by a guilty plea or by some other sort of compromise in order to avoid a completely unmanageable backing-up of the case load (Fay, 1968). The three types of plea arrangenents currently being used by the courts are: l) a recommendation in which the prosecutor 24 suggests a term of years, which has been negotiated with, and accepted by, the defendant, to the judge; 2) the dismissal of certain criminal allegations in the charging papers; and 3) a suggestion that the court accept a guilty plea to a lesser offense included in the offense actu- ally charged (Vertr, 1964). It is the judge's responsibility to insure that the guilty plea has been ascertained according to the general standards articulated in Rule 11 of the Federal Rules of Criminal Pro- cedure in which it states that a plea must be made "voluntarily with an understanding of the nature of the charge." Often, however, this is not done because there is no clear definition of voluntariness by which to judge the nature of the act. Plea bargaining has provoked a lot of controversy because many e>co .3 .co acwpowz gems» nae: weeps; pcvaacum o» he .cmgucammom onmp ._~ mesa mnmp .pm mesa xgacpewpmca .m .cucmmmmc on» c_ mcwuoqwuwucon cw woumwcmpcw mucmvaum v=_e o» Am cmcocmmmmm ommp .op mesa mump .m he: mgoumm awacuom .N .muconmc mucmucmmoca mo xuwgonae a :_ sauce on can cows: mmpnm_ca> macs» gmmamcwumwu oh An . .mnmpionmp soc; mco_uuoccou mo .uawa mueum any an com: manages mucmucmmmeg mo guano; may new: ngFVEam oeoumn oh Au cosucmmmom onmp .NN .amu onF .NN .aom xvaum pope; ._ opnrmcoqmom copumpneou eo xuw>wuu< we «women; Amvpoauw>mucfi mama kuumaxm canon oEFP xmuh usage meek m u4m new: meONe. Auweumesxmv o mxmgmsom AuPcpmsE>m mxgwsmgu 65 o eeeeepeeemem .eoeeeee co meeemee as see: NpoeN.meN eeescm-peu 3am No.oop N. m. P.F m. N.m o.N N.m N.m F.NP e._e _._ upeeee NNN _ _ a N NF eN Np em me mNN a caepoo NN.N_ ea _ o N N _F ON e e m N_ F gamete Nm.N NN o o o o o F e __ P __ o Fame am.op Pena ecu mm o o _ o P _ _ N N a, o eeeeeeeea Nm.Nm m_N o F o o o _ o op NN _N_ N compeeeea NN.N - . NN o o o o o N _ _ N m, o 6:62 t. 8mm mpepop meme» mcem> mcem> mcem> memo» mcew> cem> wave use mucmucmmmca mo gem op N m a m N _ Fame eeeeeeeee eeaeeeeea we? eoeeeeeeeseeem _FANBN5 mucmpcmm eo m_neh zucmmcwpcou 66 time there was a definite deviation in this pattern was when the officer felt the defendant should be jailed. In this case, only 39% received this sentence whereas 43% were placed on probation and 18% were imprisoned. The second contingency table investigated the relationship between the maximum sentence and the recommendation of the presentence report. Previous research (Green, 1961) had suggested that the seri- ousness of the charge, which in this case was indicated by the statua- tory maximum sentence, was the major determinant of the sentence. Yet the findings uncovered did not support this conclusion. Thus it seemed necessary to examine the correspondance between the probation officer's impression of the severity of the crime and the legislature's impression. If the two did not concur, then the officers were using their own guidelines for making a recommendation which the judges seemed to find satisfactory. Table 12 indicates that approximately 43% of the probation suggestions were for maximum sentences of two years or less and 34% were for maximum sentences of five years. Thus these recommendations seemed quite appropriate. However, in 9% of the cases which had a maximum sentence of two years or less, prison was suggested and in 11% a jail term was believed the correct outcome. Crimes with a maximum sentence of l4years seemed to present some dif- ficulty fer the officers. Whereas 47% recommended probation, 41% suggested that the defendant spend a specific period of time in jail or prison. On a rational level, the results seemed to imply that the recommendation was made without considering the maximum sentence for that crime established by law, or that the officer's judgement of the seriousness of the crime differed from the statuatory one. This was 67 opmm. u eeeeeeeeemem peeeeeeea me ease new; Npeop. mmoFF. mace. u eeeeeeepemwm FoFcpmeeamv a mxmcmeom FoFcpmeezm maeagaeu .eeeeaee co mamemee NN new; ooNNm.mN u eeeeem-meu gem ao.ooF m. F.F w.m c.NF o.m F.F N.mm m. o.mm mFmFOF Nam N e vF we FN v NwF m mmF cssFou NN.FF mm F N a FF v o Fm F NF camwca um.“ mN o o o N N o m o mF Fan RN.oF . FFma use mm o F N m N o wF o .m :oFumaocm nm.mm mFN F F m NN NF a Nu N Fm :onmnosm mm.n RN 9 o o m F o NF o m mcoz mmmb om ON mF ¢F OF N m e co N peonwm mFeucF mucmpcommca mo 30m chemFV mucmucmm Ezeerz :oneucmEEoumm aboawm mucmucmmwcg asp mo conmucweeoumm An mucmpcmm Ezeerz mo NF m4m=w memwo.- vomoo.- o_ovo.- mopmo.- om_.o.- mommo.- ~mmmo.r mkmqo. moqvo.r .owmnpwr mumocc< coco~Ememrx o__:u>:w ovoop.- momm_. vwmm_.- cummo. mpmmo. mmmmo.i o-~_. mmemo.- «ammo. aomnc.- cmucmco ucom ~_mpo. wmcv_. mocxo.i ~_mmo. MQOMmM- Noeop.- momwo.i oooeo.- momno.- Nommo.i cope xu—wao mmm_o. ammxp.i _-o_. mnemo. mwdmm. eoomw. a.mm_. _Fo_o. mmwN_. “mom~.- acoum_z ”cupcoz-amc< um~_ouwa “mo: _Fvoo. mmnmo. nvmoo. moo—F.- mmwv~.- opvmo.- mmooo.- 06:00.- «mmmo.- omnm_.- :o_uau:uw-ewc< uwpmmuwo “mo: m_omo.- mmmco.- mmmmo.- mm_mo. mpmc_.- Nme__.- mmmoo. «mmco.- oooo_.- ~m_-.- zeoum_x a__Eud-omc< uw__oumo “mo: Ncoko. ommvo. _mo~o.i mooNo.- maem..- mmomo.- omomo.- m_va. mmooo.- uawdm41. xuoum_x _mc_E.Lu-emL< um_+auma “mo: “mme_. omkoo.- «Fmoo.r _omoo. wmxmo. Numoo.- mmmmp.- oemoo. emc~_.- voNNP.- cmu_eco an cozucocqa< o_aoma Macao.i ommmo.- ccmmo. Nmomo.- _m~mo.- mmomo. vom~o.- mmkmm. Rocco. meNo. .¢.a mo cowuoucmesoumm ovum". smuko.r moo_o.- mmkmp. smouo.r mfinup. coca..- Mmo_o.- mmooP.i moomm. acoumw: mace .Noc_.- ommON. _cmoo. «make. «mQNF. mmoMm. P_oo_. mmmwo.- _moo~.i v~_~_. momma .ozou_< owmoo. mmvom. mmmmo.- mam—N. Nmmm_. mwowm. mmvmo. mm_mo. mmmm—.- mkcom. muuw>com ucm_uoqu:o new: omkmo. ommv_. mmovo. -_oo.. pmmmo.- _mmmwii “mmmo. mm_mo.- m~m_o.- mmqmo. cowuauwumcfi _~ucw: ou umuu_fieou mNmm_.i m~v_~.- “come. mmmoo.- Npmmo.i —ommo. o_P__.- mo~oo.- mm_no. omm—_.i mauu_ucoz _ou_mxza we: owceo. mmomo. Nameo.- wmceo.- mammo. commo.. wwwmui v~m_o.- FmNcF. ofimmo. mu_:u< ace cucuprnu ;u_3 m:.>p4 mmmMMM. amp—o.- mmomo.- om_mo. N-o_. mow—P.- Nmmmm.- mom—0.- omom_.- com—o. mupzv< cu.) m:_>_4 oomo~.- mammo.r mnemo.- m_mmo. commo.- ¢opmo. _vmw_.- NNm_o. mmNNo.- Fo_mo. ocop< mc_>_4 mmmop. momm~.- ammoo. .memv.- oo_o_. omoko.- omm__. commo. ammom. mmmo_. mmc__n_m co amass: Nvmmo.- vvmoo.- Nmmxo. mwm0v.- momvm. mkmmm.- omomF.- ~mmm_.- omom~.- Nmmeo.- voyage: mucwcea momnp. onNO. mmvmo.- fiompw.- mmmoo.- mm_~o. ommpo.- cmmmo. mn~o~.- mooeo. o>_—< cacao: ommm_.- ammoo. momma. vamo. mmmv_. mmmm—.- mm_Po.- “ammo. mcho.r wvmmp.- u>_.< conga» mvomo. moNNo.i mo~oc.- m__mo. erm_. NNFmo.- Pomom. Nmmoo. mmbkm. ovmno.- ucaucocmo :u_x mc_>_4 cwcup_zu ommwo. No_mo. ~mmmo.- mmomo. mvov_. oNFuo.- ommm_. mmm_o.- Mace“. «ammo. ~_ use m. cmozumm cwcu_wzu m_omo. Noomp.- “moao.- ov_oo. mmkmp. Pmovo.- mwoc_. Nmmmo. oflwqq. emmm_. N_ can o comxuwm cwcu_wzu oN~__. cmomp.- mmmeo. «mama. npmo_. MMNNO. ommmo. wepmo. mmmw_.- moemo. memo» m Lona: cage—wcu o~m__. m¢_No. mm_vo.- o~c__.- mmmmm. mmmo_.- vamo. N_Nvo. nmopo. mmkmo.- vm_gcex m_meo. mmbbbw. ommmo. mvvoo.- oo¢m~.- F_mmo.- .om_p.- .moog. m~eq~.- cmomo.- o_mcwm ommoo.i mMWMMM ommmo. No_oo.- mvN_o. mmwoo.- mxovo.- vmv_o.- mmmmo. mama..- noumocc< can: um»o_aeu ommmo. .mmao. mammo.- oo~mo. mmooo. mmmvp. mammo.- “mmoo.- _mv_o.- Naomo.- cowuoazuuo mpmop.i ~m_qm. ofiwvo.- ooewm.- mmmpp.- omopo.- mo_mo.- oo__~.- opovm.i mmmmo.- _u>m4 Focoeuauauu mmmmo. mommo.- «memo.r NMQOF. mvmwo. “momo.- amo~o.- Pomco.- oom~o.- MNVwO. .m.a mo Loo» smooo. .momo. ~m0_o.- Pmmko.- Nmmm_. “mmNo. omwN_.i eke—o. mmomo.- vo_om.- coma .mmm_.i mommo. m—voo. om_~_. mokwm. mmmvo.- ommNo. vao_.- NNNmN. mvmpp. om< m~_eo. “Fmoo.- mpmmmm. mmmmo. smNmo.r Pmomo. _mmoo. Rumoo. emcee. mmnmo. us_cu e_»uw> enomF. mmmmp.- mmmmm.r mmmnw. mm~¢o.- «Powo. scene. Fmpeo. mmvom. omen”. oucowvo aucuaoga m~o-.u mono. _N_No. _mmo_. mmmk_. mmmco.- mammo. owmoc.- p.mme. Nuomm.- umoLL< ”need an om< mmomo. _~m__.- _mo~o.- omcmo. mo_oo.r mmcop.- emwoo.- opomp.. ocm_m.r Mbm_o. magnum acaou mo_vo.- mevmo.- o-vo.- _m~_o. ovnoo.i mmwmo. om_oo.i Romp”. Nmmvp. bmmkmw cowuwmoamwo maovcom awe: “Nooo.i mkomo.- .mo__.- F_eoF. omom_. cocoa. m0mo_.- mango. m_mm~. “memo. mumacc< m:o_>mca _mov_. MFNoo.- eocm_. “come. Romeo. 00mm..- om~mo.i mmFoo.- ommw_.i mmom_. mmmcmgu unoccau co guess: oo_mo. “coco.- wvomm. «NN__. Nommo.- o~_¢o.- ooomo.- Nmmmo. Pawn—.r _mmeo. E:E_xo: agouoauoum ~m_mo. mvmco.- exec—.1 coomo.- mm__o. mmpmo.- ~emoo.- vp_o~. mmmmo.r memwp.- muu__asouu< mo consaz omoo_. momoo.i Noomm.i __omo.- oomoo.- mmmeo. emmpo. mvo_o. _mmoo. mmooo.- wow—on nap—cu Lov_mu=o commo. mmmwp.- .mmmm. Fomoo. eofic_.- mmpop. mmomo.- momop.- ommmo. oeovp. mow—on um_qu eruuw> .meoo.- okovo. «wow—.1 wmmoo. owmoo. mmopo.- omomo. _m¢eo.- ommmo.- ummmo. mac: :o—uuouwo cacao o. gouuou a genome m gouge“ N Louuma w.cOuuau m Louumu e couuom m Louuou N Locum; _ Lagoon mmc_uo04 League mopno_cu> opnavgo> acoucmqmo ecu mo: mucoucom co xuvco>om cog: mw_nowgu> co mmcvuooa Lagoon mp m4m=a mmse. Nooo_.- moomo.- ewooo. Daemo. a_.oo. m_meo.- oomoo.- “amoeba xeopaa apecasaa m_m~. «QNOO. mm_oo.- moveo.- maaoo. omooo.- ONQNO. smooo. mamacce cocooeoumez a_,=6>=a swam. “somN. cameo. amps“. NmmoF. e_eep. sommp. «Fame. absence econ «mam. ommoo.- _o-o.- momma.- weave. nam_o. «ammo. oofieo. aa_a se__=o coke. Dunno. ooAmo. mo_-.- Cameo. m~m_F. FA__~.- mmam0.- museum .eu_eaz-eae< ua,aabao use: Ammo. Nww_h. Renae. Romeo. Nepmo.- m_m_o.- wwomo.- omomo. coeuauaau-e62< capeobao “so: mean. mmmmo.- o~A~_.- No.6o. somehpu mmmmo. mmmao. Nmmeo.- accuse: »_.Ee.-aac< uapeebao “no: seek. omowm.- .pmmpppi. Nooe_. _maam. «Nemo.- emcee. omwoo.- xuobmex Fo=,s_.u-eaea oa_wauao use: mmoo. smoo_.. naeoo.- Famop. Nowao.- moomo. cameo. m_~mo. Loa_ceo an vagueocaa< a_aoaa comm. mwomo.- maomo.- upmmp. sce__. ~A__F.- mameo.- _meo. .¢.a co cocbaocaesouaa «mam. Nmaoo. mmemo. cmmwm. «meow. oee0n.- _eeeo.- cm~e_.- accuse: mace ammo. msmm_.- Nmm¢_. mavmo.- momam.- A~vm_.- Navmm. .wmaw. amen: _o;ou_< some. macaw. e~_5~.- awkwo. AN,_o.- _eo~o. momma. Nmumo. mao_>eam beanaapzo can: 42,“. anmo.- Namoo.- thoN. mmamo. _6opo.- aweeo. _Nmep. =o_b=acumc_ .aoca: ob uabbeeeou muwo. _mom_.- e_PNo. .eawo.- cameo. ~_omo. Noemo.- oom~_.- ma~6_ucaz Pacemxga an: Pack. o-w0.- mnemo. acmao. oeo_0.- mehmo.- Nmm_o.- omoo_. mo~=u< ace caeup_;u gee: ac.>,b m_-. NNNmo. mmmmo. _A~mo.- cacao. m_mFo.- mmawo. mo_m0.- mo.=u< gee: m=P>.b amok. «memo. om_mo. -c~0.- Pmemo.- .oowo.- mwmeo.- a_ooo.- acop< a=_>,b cmmm. mammo. meamo. o_MNF. Newm~.- 40060.- m_mmo. moeeo.- mm=__n_m .6 260532 ammo. ”omen. N_mmo.- aOOmP. ekeop. m__op. oewmo.- _Amm_. uaeeca: manages Noam. mecca. oe_oo. e-eo.- cammo. ~_~NN. “.65o.- m_omo.- u>F_< Lasso: Nome. emem_. mmooo. «ammo. moAN_. m~m~_.- Amoco. comma. 6>F_< Lessee Name. Rosco. mmmmo. Npmao.- moomo.- oemm_.- Nmmwo. m-~o.- oceacocaa ;o_x mecseb =62c_esu NNMo. mammo. mm_mo. wmwno. oa_~o. mmowo.- meeoo.- ~N_mo. N_ was m_ asexuam =6.u_,;u mmom. mmoom. momma. momm_.- oemhw. oNoo_.- amN__. Neeoo. NP age a cooxpam coeu_ezu meNo. 0mmm0.- “Fweo. _N_eo.- mammo. Nqayo. ~_mefi.- mmN.F.- mama» m Laue: eacupegu ammo. Aoom_.- e_m_o. omomo. weave. ooeoo.- wm_mo. ae_oo.- usages: meme. mammo. Cameo. cmo_o.- mammo. comao.- weeno. Novmo. a.mccm ammo. a.meo.- Reese. om0mo.- maowo.- ooNNN. mmamo.- oammo.- uaumbcc< cog: case—as“ ommm. Nmawo. mnemo.- amm—_.- No.0,. «mac..- m~o~o.- mwmco.- cowueaaouo Nmmm. sap—N. ONooO.- moomo.- emoo_. moF~_. mFe~_.- momwo.- Pesos Fecoeuauaum New“. m~__o.- momom. Amp__. momma. oom~0.- ammmo. mwmmo.- .¢.a cc 2.6» some. mow—o. mammo.- meopo.- oeepp.- memPN. Nmooo. mam_o.- boom mm_m. mo__o.- cacao. Nemmo. ~mmoc. mmamo.- oeowo. .~_mo. 66¢ mamm. moses. enema. Poemo.- mwemo.- momma. .mmmo. “memo. «zeta a_bu_> _mmm. omamo. ~_moo.- oaoeo. eaaao. Pe~m_. mmaeo. oomm_. omcocco aucoaoaa 6mm“. swamp.-. _mo~o. ommmo.- «memp. a.mmp. .ooeo.- _m_m_. emoee< 632.; be ua< memo. m_~mo.- mo_o_. Pomeo.- "moon. 046mm. mammo. cosmo. «spasm 82:69 NFNA. cacao.- .mmm.. mooMo.- aaomp. ommmo.- oofioo. oo~e0.- coeuemoam_o maoeeom use: “wee. oems_. am_ofi. movop. mmmeo.- Neopm.- onoo.- Noeoo.- mumuge< maoesoea wmnm. moeo_. Nmamp. mwmmo. mee,~.- macaw. mmmo_.- moowo.- «outage ueuacau co Longs: comm. .oweo.. ~ae~_.- eomem. mqmm_. a_a~o. N-~o. m.wmo. s=e_xex xeoueauaum newm. o-__.- Poewo. emcee. .Noeo.- weave. oo_ao. mammo. mau_.ae°uu< co cones: Am_m. mwmmo.- e_~_o. Romeo. meomo.- mamwo.- mmepm.- «swam. ouegoa ua__.u Lou_mu=o came. woomo. ~m_o,.- mommo.- enema. m_wpo. mamoo. ommmr.- 66_Foa ua__~u 5.86,) mwam. eNeoF.- «mo~_. mwemo. owomo.- o~_~o.- _Na_~.- movao.- «cox =o_uuobao Lasso wwwuw—Mfisficou up LOuucu o— Louuou m— LOuumu v_. .2506; MP Louuou Np LOQUQm _.— Lauua... mmcvvood Louuou mm_aowco> coacvucou -m_ u4mee mcnmo.- m__oo.l _vm_o.i Foopo.l mnemo.i ommmo.i .oomo.i mowwm. Nevvo. mwmvo.i umec< Ace—mu o—wcm>:n emeee.- eeeee.- meeee.- eeeee.- eeeee. seeme.- ewwme.- eeeee. eeeee. emeee.- emeeee Loeeeeeem_: e_eee>ee memee.- eeem_. eNAeF.- Nee_e. emeee. meeee.- nemep. ~e-e.- sees..- memee. eaeeeee eeee mem_e. eee~_. meeee.- eeeme. m_mme.- emee_.- eeewe.- Neeee.- pe__e. emeNe.- ee_e ee.e=e mmeee.- enme~.- eamee. eeeee.- _emee. ee~e_. meee_. eseee.- meeee.- .PNe.. xteun_= .eeeeez-eee< ee__eeae one: mNeme. ememe. eee_e. emcee. meee~.- ewe—e.- eeee..- e_eee.- «Neee.- ~_mee.- ee_eeeaew-eee< ee__eeee one: memw..- memee.- eeeee.- ~_~ee. meme..- eeee_.- .eewe. m_e~e.- Nqunin Nen__.- ecoen_: »__Eea-eee< ee_eeede one: _~ee_. e_mee. memme.- eemee.- _m_m_.- eeeee.- eeeee.- eeee_. mlmae4. enmee.- ageam_: .ee_e.ee-eee< ee__eeee one: memew. eemme.- eemee. eNmee. eme_e. eaeme. me~_~.- emNee. maemniu mem~_.- end—ece an eoedeoeae< e_eoea eeee_. eNFM..- meeee.- e~m~_.- .eehe.- .emNN. neeep.- .Fm_e. eeeeN. nee__.- accede: eeee eeeem.- e_e~_. .eeee. eNmMN. eNeee. enmemh. e~e~_. eeee_.- _meee.- me~e_.- ae_eeea .eeou_< meemp. .eme_. -_No. m_oeo. mwvoo. mmmmshl o_o_o.- mmmno. «mseo.i .omwo.i mou_>cmm ucwpueauao new: NN_ee.- ememe. meeme. meeee. eemme.- eeeenhi eeeee. em~ee.- meeee. eee_e.- ee_u=e_eme_ .eeee: e co eeee_25ee Peeem.- mmemm.- ee~_e. e-m~.- emeee. ee_ee. eeemepn. «a.me.- .APee. eemme. maee_eeez .66—neea ea: meeme. ~P~e_. e_~ee.- eeee_. Ameee. Npeee.- eeeee. meeme.- _-~e.-. .ee~_. nu_ee< eee eeee_eee gee: eeesee meme". eeeme.- eeeme.- eeeme.- nee~_. ee~ee.- eemem.- nem~e.- ee_ee. .eee_.- ne_=e< as.ueeae=m gee: ee_>_e mNeeN.- meeee.- Neeee. eeeee.- emaee.- e~m~e.- ee~e~.- e_e~e. Reece.- eNeme.- nee—e eee>_e e_eee. eNeFN.- eemee. eeeme.- ee~e_. Naee_.- eewme. eeeme. Navee. ~e_~_. mee__e_m co emcee: emeee. Nemee. eem__. ee~m_. e_ee~. Nem.~.- come..- meemp.- mewee. ceme~.- ee_eeez mueaeea eeme_. eeeee. eweee.- ee.ee.- e_~ee.- eNeee. eeeee.- “meme. mmeNP. nuewuqu. a>._< emcee: eeeN_.. Nemee. mee_e. meeee. Neee_. eeee_.- mmmpepu meeme. eeeee.- ewameln e>e_< eeeeea eeeee. emeee.- “meme.- meaee.- ee-_. Ameee.- eeeem. e_eee. eeeee.- mmnew. eeeeeocae no.3 ee_>_e eeee__ee mesme. emeee. eemee.- Neeee. eeae_. eeeae.- seem.. eemwe.- eem~e. mmaqu. a. eee e_ eeeaeee eeee__ee Newmo. ooem_.i _vepo.l “mmoo. mo_v_. ~Nvmo.r o~m_—. _m_mo. emcew. awmumd. N. can w coozuom cocup_zo emem_. MANe~.- emeee. m~_e_.- emme_. NNeNe. mmmeeq. meaep. «mmee. eaea_.- neeo» e emcee eeeepeee eeeee. ameee. eemee.- mNNee. n_em~_ Neeme.- eamee. e_eme. me_ee.- ~_eee. eeeeee: eee_e. .mmwwmwu Nmeee. eee_e. m_~m~.- mm_ee.- Neeee.- .eeee. eemme. meme~.- e_ee_m ememe.- .emmmmm. ee_me. me_ee.- eeeme. eeeee.- eee_e.- «meme.- eeNne.- Neeee. eeenuee< eoex ee»e_eae maeme. Neeee. eeeme.- mmee_.- e_-e. eeepm. mew—e.- ee__..- ee_ee.- Nee_e.- eo_eee=eee _ooeo.- oemmm. wee—o.i .mm~o.- meomo.- mmepo. mmmwp.r camo~.i opwmo. camem.i po>04 peeoeuoueeu meem..- eeeee.- mm~e_.- meme_.- “Neee. ee___.- Fweee. eaee_.- ~_eem. -n__.- .e.a co and» mmeee. mme__. eNNee.- meeee.- eem~_. eeeme. _e~e~.- meaee. meme..- eeeee.- «one ~_me_.- eeeee. e_eee. m_m~e. «ween. .meme.- mamee. meme..- .eee_. “ewe“. gee eee~e.- eemne.- eee_e. __N_e. ee.ee.- eepee. Nee—e. meeee. .Neme. eeeee. useee s.eu_> .AMeN. eaeee.- eeemm.- “meme. Kmpmp.- eeeee. e~e~e.- eeeae. «mem_. eemm_. «acne aeeeeeea eke~_.- Neeee. eekee. mee__. «seeN. eeee..- eee~_. emamein. e_eee.- merme. uaeee< once. ea de< N_Ne_. ~_eee.- Nmmme. memee. ekee_.- ee-e.- ee~e_.- e—eme.- nNmmuq. ee_me. neeeum “Lace eeeme.- eeeee. m_ame.- “meme.- mmeee.- eeeep. newee.- ePee_. emqmai. ._ee_. ee.u.neenee nee_eem one: eemee.- memee. ”Nee..- _-ee.- mamae. ee_ep. _eee_.- memm_. eeemm. ~e_em. newnee< neemaeaa eeeee. eme_e. em~_e. .meee.- emcee. m_eee.- eewee.- ememe.- Famee. eeeee.- neeeee co cease: eeee_. eNNee.- .Nmee. emeee. eemme.- eke—e. eeeee.- m~ee_. e_e~e. na~m_.- seaexex eeeeeeeeem Keene. ekeee.- .Fem_.- eeeme. ee_me. eeeee.- emeee.- .eee_. memee.- eaeee.- neu__eeeuu< .6 Lease: ememe. eeeee.- Neee~.- Nemee. mem_e. ~N_ee. Neeee. neNee. ~em_e.- eem~e.- no._ea ee__ee eeeeneee eeeee. emee_.- ememm. eewee.- emmep.- emee_. e_~e..- “meke.- makee. Neme_. eu__ea ee__ee eeeu_> eemme.- h_eme. mmee~.- Nemee.- eeeae. .mhee.- .meee. eeeme.- emcee. eeme_.- nee: eeeuuaeee Leeee o. coeueu a Louueu m Louueu n gouge“ o Louuou m ecuuem v Louuom n couuou N Locum; p cocoon maceveOA gouge» mo—no_co> .opeeece> “sausages one we: ucoeaa mucaucomace mo coeueecossouom can: ma_noece> a— m4mea .amn. moo—o. ommpo. mampa. owaao.- comma.i aaaea.i um¢cc< x:o_wu w_ecm>:a e_mN. paema.i .aNma.- NnNPa. «mafia. mm_aa. oae__.- unoceq Locooeacm_z o—eee>ea MNmm. P_waN. NeNmm.- Nm_a_. maNNm. ace—N. FaeNa.- eoueeea eeoa acme. ameam.- Naa__.- .avea. ampNa.- Namma. aaaNa.- ee_e xe__ea emaa. mmmaN.- aoomp.i puma". FNNaa. emeaF.- paeea. apogee: _~ueeexreuc< eopeeuma ewe: veNo. _mmao.i eNoma. mmvNo.i ameN. FNNoa.- _vm_a. :o_uou:auiowc< um__euma awe: .mao. mo_va. oNNvN. Namm—. NNNao.i NaNOP. Nwmmm. xcoum_x x__EeulomL< um~wuuma umoz aamN. Fmae_. mNavp.- aN_aN.- aN_am.- ameaa. N_NeN.- Neoumex Fecesecaieee< ew__euoa ewe: mama. cameo. aamva.- amaea. NmmNN.- Fm_Na. maema.- cou_eea Ne eoeueoeee< e_eome eeem. ampaN. aNaaF.- ma—me.i mpapa. mNNaa.- peNma. meoeme: mega MNma. mamma.r mkmm_. amen..- Noam..- epkmm. maaNF. sm_eoea .oeou_< NnNa. N_mma.- ammNa.- mmmaa.i aeaNa. maaaa. aeaao. muu_>com eemeueeuea can: a_aN. F_amp. Neaaa.- apaea. .Naea.- «NeNa. NeNm_.- eeeueueumc_ _eece: e o» amputeeoa mean. FNFma.- aaa__.- oNaNN. aNaNa.- aNaNa.- mama..- meeu_eeez Peu_mxce no: meme. Nmaua. a_.Na. ama—a.- Naama.- mNmaa. mamea.- mu_=u< use cage—eza sue: ace>ea _eea. mm_a_.- a_aNa. Naoma.i Neaaa. m_maa. m_mm_. mu_eu< o>eucoeeem ewe: aee>ea a.me. maeaa.- N_Paa.- aa_ma.r _Nmpa. ammo_.- N_vma.- acoP< ac_>_a See. 28.. dermal ewe—e. wees. Reeve. 22:; 853.; .3 has: Kmaa. comm—. meaN. aoam_. aNamm. Pamea.i NeamN.- eweceez mueoeee Nana. aaaNa.l va_am. emaNm. aaea_. aN_ma.i spamm.- o>e_< cacao: _~ao. avMNa. mmeNN.- oaaeP.- Naaa_. NNmaa. Naeaa. o>e_< emcee; amNN. —avaa.- avava. aeaN~.- aaaaa. _aMNa. eamaa.- acoeceeea eu.z mc_>_e cage—ecu maNo. amNNa. Paeea. Nonma.r a_ema. Nm_aa. aaNma.- N_ van m. :aoxuoa eoce__eu meam. a.mep.- Na_ma.- aMN_N.- mNam_. mNeNa. map__.r N. ace 0 awesome cage—ecu aNPa. ampea.- aNN_a. aaepa.i aaaaa.r mapmp.i ameoa. «Lee» a Love: cone—ego eNNa. Naaaa. mNNNa. pea—a. Na_e_.- aaeNa. m_aeo.- eoeccex eeae. camaa.- ~aaaa.l NNm_a.l am_a.. aeema. aemaa.- o_ae_m Naaa. meaNa.- aaava.l amaaN. mac—0.- mmaNa.- _Fema. umueuge< cog: no»o_esa aN_m. eeae_.- NaeNN.- NNNm_.- aNa_p. eaaaa.- mNNaa.r eoeueeeuua maNm. mN_a~.- Nmpaa.- aaam_. FNFQN. amaep.- mamaN.i Fuse; Fecoeueueuu «mac. PNme_. aaaoa.- cea~a.- aema_. aaam_. mae_m. .¢.a so cum» maNa. Fme_a. poaaa. .aNPN. NmaNa. Nemma. makmp. eue¢ m~_a. aNmmo. mmmNa.- amm__.i aNMNa.- meNaa. Naamo.- ma< N_Nm. Np_va.- ee_ma. mmmea. meNma. comma. eN_ap. ueeea eeuue> .mee. NaNma. aa~aa.- ape—a.i avaa.- Naoea. NvaaN. oeeea mucuaoce aNMN. Pmmaa. mamaN.- aNNaN. Nmaaa.- N_e_a.- Namma.- unocc< umc_m uo om< eaNm. a_ema.- MANNF. maeaa. aNamF.- aeNea.i maaa_. neueum acaou Name. enema.- NmmNm. anemN.- e__NN.- mmpaa.- momma. coeuemoawea nee—com awe: aamN. Nekma. mNNmm. m_a_m.- Nma_a.- amaaa.i a~_N_. memucc< m=o_>oce emam. moo—a. mmaea.- mNeaa. oNNNa. maeaa.- aa—aN. muceoa co geese: mmmm. am__m. mean..- _m_ma.- Nera.- mmema.- aNNma. e:s_xex apogeeueum aeao. Noaaa. aaNma.- aNNa_. N_mma.- me_NN. oomva.- meow—escou< eo_.oce=z «Mae. Nampo. Faaaa. ma_ca.- amaNa.- mmmmwwn NaNNa.- ou__oa em__ea coeemueo oaNa. amNN_.i momma. emoea.- ammaa.- emekm. NNpNa.i woe—ea eo__~a segue> a.mm. Naaaa. aemma. enema.i Nmeaa.- aaNaN.- emaNa. eve: eo_»umuoa eozuo 3.522358 e. at: 2 238a 2 e368 2 .338 N. at: : 28.62 maceueoe Cayuga . mo_no_ee> emaceucoa.mp uaacuu aco.uonu=o u.cue.zuan emu: 5.. 3353...... .355 e 3 33.5.3 vaN. Nu..:a no: mo eo.e a. canoe. mEo.noLa .a:OmLme m gouge; 3e. 85;; .3 2352 m. .36: m3. 3...... 2.28%“ 5.: 2:: eez . a o a a: «we. memo» m eous: coca..ea mmm.- manomewemaouwawu Na.” weaueoeaa 0;» cu.) oc.>.. come.»zu m... .ou.c2. mung .va «p.3ue use cuau..:u no.3 a. .4 denote 3.32 2.. e225 9:5. 5... e5): e .36: . -i. . cam. manages magnum o..=m>:a aNN =o.uuueuu a a mo coco uo..euoe one: m. Louuea new. numogee zeo.oe o..co>=a cam. ou..oa ao..au s.uu.> «ca mumucee coeoueocmte u..co>:a .NN. moo..asouua mo cones: mumoce< o..cu>=a m segue. oa..oe .. 8:8 533:3”...282 .6 goes: N. .38: “mm. 8:3 8:88.. 3 eafiefleuecwfiuwu .. Nu m .aN. moocezu ecouceu .0 Loans: .. gouge. Nam a..e~.ii.¢.a mo coca ue..ouueowmwn . . r . ago m mac. mu.eee o>.ucoeaem cu.x m:.>.a aNa .ec.s.cu--.x.e co ease an..euoe New“ am. eco.~ ae.>.. uoz ace. co.u.moem.u m:o.eum awe: 3.33 93.8% 5... as»: 2 e862 8.... 253m :38 . mam. co.uea=uua aboum.: .ec.s.ga N gouge. «as. eoumcgco cog: v6.6.asm Nam. o>..o no: segue: uoao.a5m a Louueu awe. uuumugco ung.» cos: oa< mac. N. age a coozuen coca..=u .Nm.- «e.gu xueoaoge ova. u>..e no: Logan. cam. oucaucom_cse.xea .uoauaueum .v.. N. van a. coozuoa coca..zu a... oe.eu e.gu.> N».. can penance oaeana mo amoemeo.com a Lagoon manage; aeogu.: manage: eoe.o . gouge. mac.eo04 nae—coo. gouge. mo.ae.eo> canoe; cocoa. mo.ao.go> Louuoa o.ne.co> acoucoqua as» me: peace: mucoucomoea ecu no co.ueu:uasouo¢ can: mm=.voo. canon. o>—»uonno¢ 5.0:. can mgouue. a. 6:» mo zoom :. uoe.uu:oa m~.ao.ee> ON mam.e aeNN.. .mmaa. mma.m. camam. N... NNmNa.. mummec< m..cm>ea aaN.a.- cacao. wea.m. mmNam. me.. ..am..N msm.eoce .ecomcwa mammN. eaeaa. aeN.m. maamm. we.» aNaa..N .uopm.: mega .aaaa. Nmmaa. amaam. ammmm. aaa. a.NNa.N mp.:a< m>.ecoeaem su.3 a:.>.a Nmaaa.- amaaa. eaNam. amamm. maa. maaNe.m .m.e .0 auewuum .aNN.. meaaa. emaaN. .meem. aaa. ma.am.m maceca .0 mmmemeo.emm mama..- weaaa. .amaN. meamm. ama. aNaae.e ew.e .a..=a emNN..- ama.a. aa.aN. amamm. .Na. aaaae.m .o>ma .eeo.ueueam a.mea.- .aa.a. NaaNN. amaNm. aNa. aNaNe.m mcmcpee ueoeu.3 memeuoz cme.a aNNNN.- a.m.a. mNaaN. m.a.m. aaa. aem.a.a mung aaae.. NemNa. amaeN. aeaae. .aa. ama.a... mmo..esouo< _ mo L2.532333; am..ea E.pu.> aNN.¢. mamNN. maMNN. aNNNe. aaa. Naamm.Na. «comma muempcmmmce mo coepeecmseooom ace .uoum.= .ec.s.ca macega m>osmm co m m.ee.m meeecm-m mceeam-m a m.e.p.ez moeeu.w.:a.m cmpcm o» a mm.ee.ge> m.ee.ee> “emacmema we» we: moempcmm .o xu.cm>mm ems: :o.ue.om :o.mmmgawm m.e.e.=: .eu.e.esm we» NN m4m.a mnmea. mmNaa. ammam. .mmmm. NmN. .Nmm¢.. ms.ca mnemeoca mm.ma. amNaa. memae. mmmmm. NaN. NmNmm.. mew.eoca .ecomcme m..m..- a.maa. Nm.ae. mmmma. aNa. mmmm..m mp.ee< ace amen..;a cu.z m:.>.a NNmm..- mamaa. Nemmm. mmmNm. ama. NmmNa.m Empmmm pcoeeem .e.oom momma. emmaa. .m.mm. mmmNa. mma. Na.mm.m co.mowpma mo maoz emcee mmm.m. ammaa. m.amm. ae.Nm. .aa. Nmmmm.m mcomm... .e:.E.La-emc< am..euma “we: mmma..- aa.aa. mmmem. mam.m. .ma. meemm.e mp..ea uoz .0 em.e mmNm.. aN.Na. m.NNm. maa.m. aaa. .m.mm.N. mo..oe am..ea E.uu.> mmNee. NNNNa. mmamm. meNmm. aaa. NNmem.m. myopm.: .ec.e.ca mmmmm. mNmNm. mNmNm. mmmmm. aaa. mm.ma.mN. ucoemm moemuemmmce we» mo co.peacwseoomm emceea m>oEmm co m m.ee.m mceemm-m mceemm-m m m.q.p.ez mo:eu...cm.m cmuem cu m m.ee.ce> o.em.ce> Hemacmema we» MN m4mmm can: eowpe.om :o.mmmcmmm a.m.».ez .ece.pem wee 90 than the empirical one because it could account for 39%, instead of only 29%, of the variation in judicial outcomes. There was also some consistency in determinants across the methods because previous crim- inal history, recommendation of presentence report, plea bargaining, and mode of detection surfaced in both equations (Table 24). TABLE 24 A Comparison of the Variables Forming the TWo Regression Equations when Severity of Sentence was the Dependent Variable Empirical Solution Rationa1 Solution Criminal History and the Recom— Recommendation of the P.R. mendation of the P.R. Number of Accomplices/Victim Number of Accomplices/Victim Called Police Called Police Race (Black) Criminal History Older Mothers Without Fathers Most Detailed Area of P.R.-- Criminal History Educational Level Plea of Not Guilty Plea of Not Guilty This suggests that certain variables were strong enough pre- dictors that they would enter the equation regardless of the data reduction technique employed. Criminal history and the recommendation of the presentence report, the first entry in the empirical solution, accounted for 22% of the variance whereas the first entry in the rational equation, the recommendation by itself, accounted for 32%, or an increase of 10%, of the variance. This implies that the first factor diluted the power of the recommendation variable. Perhaps 91 this blanketing effect accounted for the fact that the rational equation could explain nearly 10% more of the deviations. In fact, the recommendation of the presentence report alone explained 32% of the variance, whereas the entire equation had a predictive ability of only .39. In other words, the subsequent four entries increased the Brsquare by only .07. Initially, it appeared that the equation formed when using the rational factors was better able to predict the severity of sentence. To insure that this was so, the adjusted firsquares of the equations were compared by using the following fbrmula which tested for differ- ences in population correlation coefficients (p01 - p02) when the values were based on dependent samples, i.e., when the 3:5quare values were correlated: z _ "‘— [("01"'02) ' (901‘ 902)] /fl 2 2 2 2 2 _ _ _ 2_ 2 12* (1"01 ) * (l‘roz ) ' 2r12 + (Zrlz r01 ”02“1 r01 r02 'r12 ) (Olkin, 1967). The adjusted, rather than the unadjusted firsquares, were employed to take into account the fact that the equations did not contain the same number of variables. The empirical adjusted R:square was .278 whereas the rational one was .378. Since the size of the firsquare value varied with the number of items, Rysquare may have been arti— fically high in some of the equations. This statistic controlled for the amount of variables included in the equation and it made the standard error larger, thus producing a more conservative test. In order to utilize the formula, the correlations between the adjusted 92 Rfsquare values were calculated. The correlation between the empirical and rational predictors (9) was .778. The z_score was then determined using the adjusted R:square values for rm2 and r022 in the above formula, and the square of the correlation between these for the r122 (Table 25). TABLE 25 A Comparison of the Adjusted BySquare Values for the TWo Multiple Regression Equations when Severity of Sentence was the Dependent Variable Factor Type Adj us ted ; R-Square Empirical .2784 Not Calculable Rationa1 .3778 No standard score for the comparison of the rational and empirical adjusted Rfsquares was determined because the formula was unable to calculate one. It seemed that whenever the two y's of the respective equations had a high correlation (i.e., one that was greater than .7000), the denominator became undefinable. Thus the conclusion reached was that this equation was not robust for high inter-y correlations. Nevertheless, simply from an intuitive perspective, the difference between the rational and empirical adjusted Brsquares appeared significant because the former solution was able to account fbr approximately 10% more of the variance. 93 Recommendation of the Presentence Report as the Dependent Variable. The same rule which guided the inclusion of variables in an equation when severity of sentence was the dependent variable, was used. Only items with an f_to enter or to remove which had a signi- ficance level of .05 or better were considered. Thus, in the case of the empirical regression solution, only six of the 19 entries fulfilled this requirement. They accounted for 28% of the variance in comparison to 30% when all items were included (Table 26). The factors composing the equation were: criminal history (most serious previous disposition, on pro- bation/parole when arrested, year of presentence report, number of people approached by probation officer, most detailed areas of presentence report were previous criminal history and family history); juvenile arrests (age at time of first arrest, previous juvenile misdemeanor arrests, previous juvenile felony arrests, previous juvenile status arrests); white offenders with few_previous arrests (number of previous arrests, race, drug history); persona1_problems (committed Una mental institution, used psychiatric outpatient services, alcohol consumption); and plea of not guilty. Thus, if a women was not released on bond, had not plea bargained, had personal problems, a juvenile record, and any of the variables included in the factor criminal history, she was more likely to get a harsher recommendation. This model involved 18 of the possible 48 variables and left over 70% of the variance in the recommendations undefined. The equation which would be used to predict the severity of sentence was: 94 .m.ma. m.aaa. «mean. mm.mm. mm.. .emma. mme..e.m mo Lmeesz mamma.- .Naaa. .mmam. Nm.mm. m¢.. mmma.. ew.ccez mucmcee mNNNa. m.aaa. ameam. ea.mm. em.. aemma. emu..esoou< we L35:58.3“. em..ea s.po.> .emNa. aNaaa. ..eam. aa.aa. m¢.. maea.. mm=e< .onou.< mm.m.. .maaa. .mmam. mN.mm. m.e. macaw. emcesa .o mememeo.cmm maam..- ...aa. mmNam. mmamm. mam. NNm.m. co.peoeeu . -.m.e .o emc< em..euma pmoz mNm.a.- Nm.aa. Nm.am. mmmem. mae. mNmma. mmmceea pemccea mo L3:52 mamma.- maNaa. .maam. m.mem. mam. m.ama.. ma.ee< ace eeee..ee eeee. ee.z ee.>.e mmmaa.- amNaa. .emmN. mmmem. m.N. N.¢m... mpcmcee peoep.2 mcmceoz cme.a aamNa.- m.aaa. ..mmN. NNemm. we.. .aNe..N mu.ee< m>.pcoae=m :u.z mc.>.4 ama.a.- m.aaa. maNmN. .eaem. mm.. m¢.Nm.. .m>ma .eeo.peoeem .mam..- mmmaa. .mmmN. mmmmm. mm.. NN.mm.. ammo.e2m Nmmm..- .mmaa. mmemN. emmmm. .ma. NMNmm.m em.ccez emNN..- mm.aa. mm..N. mmmNm. mma. Nm.m..m m»..:a poz mo em.e Nmm... Nmmaa. .Na.N. .mm.m. mma. .m.Nm.¢ mam.eoce .eeomcme ma.mN. NmmNa. mm.mN. mm..m. aaa. NmamN.e. mummcs< meo.>mee 2m. an.2 mcmaemwea mu.;: a.maN. mmNma. mmmmN. momma. aaa. ma....m. mummcg< m..:m>=a .amNm.- aa.aa. .maaN. ammee. a aa¢m¢.mN umpeeea aeom .a¢.m. mmmm.. mmmm.. .a¢.m. aaa. mamem.am mcopm.: .ee.s.ca wmceca m>osmm co m m.as.m mememm-m memecm-m m a.m.».ez mueeu.m.cm.m empam on m mm.ee.ce> m.ae.ce> acmeemema we“ we: peoemm muawpemmmge we» we co.peaemesoumm ewe: eo.p:.om co.mmwgmmm a.m.».ez .eu.c.aEm 0N m4m mmN.a.- .m.aa. .memm. m.mmm. Nmm. am.m.. m:o.< m:.>.a mamma.- aa.aa. ammmm. emamm. amm. eemmm. mm3.988.. mo ceasez mmNNN.- ...aa. a.Nmm. mammm. mNm. mammm. um~gg< aaa.. we mm< momma. mm.aa. mmamm. mm.mm. .mN. aN.ma.. emummcc< can: mm< a.me..- mmNaa. eemmm. mNamm. aNN. e..m¢.. t23...... .ee.cez-emc< aw..euwa “we: mean..- mm.aa. m.aaa. mammm. mmN. aNama.. eo.peueum-ewc< em..euma “no: maaa..- emNaa. mmmem. mmmmm. mmN. .m.N¢.. cmo...a eo.ueeoca . me emeueocae< m.aome ..em..- ..Naa. .m.em. NNemm. m.N. a.m.m.. Emammm ecoeeem .e.uom am.ma.- mamaa. ammmm. mm.mm. .m.. .mm.... mmcmwea one .o mememeo.cmm emNN..- ..eaa. Nmmmm. eNm.m. .m.. mmamN.N mu..:a go: we em.e N.ma..- mmmaa. .e.mm. mam.m. .aa. amNNm.m eo.uomuma mo mac: cmnua .Nmam. aN..a. mamNm. e.a.m. Naa. mm¢.m.m meoum.: .e=.s.ca-emc< em..epma “no: a.mm.. mmMNa. m..am. m.aaa. aaa. aNmm..N. mso.eoge .eeomcme .ammm. a.mmN. a.mmN. .ammm. aaa. mam.e.me. xgopm.: .e:.e.ea emcega .m>osmm co m m.ee.m weeecmum oceemmum a a.m.».ez ouceo...:m.m empem om . mo.ee.ge> m.ae.ee> ucwacmewa .N meme. one we: pcoewm muemucmmmee one .o =o.pee=mEEoomm can: eo.u=.om eo.mmmemmm a.m.“.ez .eco..em 97 TABLE 28 A Comparison of the Variables Forming the Two Regression Equations when Recommendation of the Presentence Report was the Dependent Variable Empirical Solution Rationa1 Solution Criminal History Criminal History No Bond Granted Personal Problems Juvenile Arrests Most Detailed Area of P.R.- Criminal History White Offenders with Few Previous Arrests Personal Problems Plea of Not Guilty misdemeanor arrests, previous juvenile felony arrests, previous juvenile status arrests, drug history, the most detailed area of the presentence report was previous criminal history, committed to a mental institution, used psychiatric outpatient services, and alcohol abuse. I In order to judge the predictive ability of the two equations, the same fOrmula devised by Olkin (1967) and used previously when severity of sentence was the criterion, was employed. 'l"_[ (rOl'rOZ) ' (om-602)] _Z_= The adjusted firsquare for the empirical solution was .266 whereas the adjusted Bysquare for the rational one was .3196. The correlation 98 between the adjusted Rfsquares of the two equations was calculated to be .6532. The solution generated by the formula was 3.154, which was significant at greater than the .01 level (Table 29). This implied that the rational result was significantly more predictive than the empirical one. TABLE 29 A Comparison of the Adjusted R:Square Values for the Two Multiple Regression Equations when Recommendation of the Presentence Report was the Dependent Variable Factor Type Adjusted 3:5quare ;_ Empirical .2658 3.154** Rationa1 .3196 **p < .01 At this point, it is necessary to stop and look at the predictive equations developed using the two alternative dependent variables, i.e., severity of sentence and recommendation of the presentence report. Each of the equations derived for the two different dependent variables were compared as to content. The two results based upon the empirical factors were not all that alike although they both began and concluded with similar entries. The four items inbetween did not have much overlap. This might lead one to presume that the judges and the probation officers considered different criteria when they were forming their decisions. In contrast with this, the equations resulting from the rational 99 factors showed a fair amount of resemblance. They both contained the variables criminal history and the most detailed area of the P.R.-- criminal history. Some of the shared items were entered into the hierarchy at different levels, but, nevertheless, they were still included in the equations. The three variables that were distinct to a particular model were number of accomplices/victim called police, plea of not guilty, and personal problems, with the first two associated with the solution developed when severity of sentence was the depen- dent variable. The results were also compared for predictability, and it was found that the rational equation was the more predictive when recom- mendation of the presentence report was the dependent variable. It was postulated that this did not occur when severity of sentence was the criterion because of the high intercorrelation between the pre- dicted scores (y). However, it was also stated that, from a visual perspective, the rational adjusted grsquare appeared more predictive than the empirical one (.378 versus .278). One final observation was that the rational equation developed when severity of sentence was the criterion was able to account for more variance than the one developed when recommendation of the presentence report was the dependent variable. The adjusted Brsquare of the former was .378 whereas it was only .320 for the latter solution. Discriminant Function Analysis It was unclear which predictive technique would produce the most predictive equation since both had their weaknesses, and thus both multiple regression and discriminant function analyses were 100 employed. This latter method had, as one of its objectives, the ability to discern those variables which could differentiate between two or more groups. This capability was applicable to this study because the dependent variable naturally divided into two factions-- offenders released on probation and those incarcerated either in the county jail or state prison. The technique would determine those variables which best discriminated the members of the two groups. This then, was an alternative mode for discovering those variables which were used by judges when deciding the outcome of a female offender. The same sets of factors and scales used in the multiple regression analysis when severity of sentence was the dependent variable were employed because this technique also required the variables to be as uncorrelated with each other as possible. The discriminating variables were entered into the analysis through a stepwise method which, by selecting the next best discriminator at each step from amongst the remaining variables, attempted to produce the "best" set of discriminating variables. The criterion by which independent variables were selected for inclusion was specified as Wilks lambda. This test considered the difference between the group centroids and the cohesion within the group. Ideally, one desired two distinct groups which were very different from each other yet homogeneous within themselves. As in the earlier analyses done, rules were devised to guide the interpretation of the printout. It was decided that only those variables which caused a change in the Raos V statistic that was greater than or equal to the .05 significance level would be included in the discriminant function. Once again, the amount of variance 101 left unaccounted for by choosing this cut-off point rather than con- sidering the entire equation was negligible. The 17 empirically-derived factors and three singlets were placed in the program and an equation consisting of 20 items was developed by the computer (Table 30). Of these entries, only six had a significance level above the stipulated .05 one. These were: criminal history and recommendation of the presentence report (number of previous offenses, most serious previous disposition, court status, recommendation of presentence report, most detailed area of presentence report was previous criminal history, bond not granted); race; drug history; older mothers without fathers (age at time of first arrest, age at time of arrest, children between 6 and 12 years, children between 13 and 17 years, father not alive); educational level; and plea of not guilty. The predictive function which would be employed according to this model to determine the sentence of the offender was: y = .889 Criminal History and the Recommendation of the Presen- tence Report - .307 Race + .203 Drug History - .166 Older Mothers Without Fathers - .205 Educational Level - .161 Plea of Not Guilty.* *independent variables were in standard score form. The ability of this function to correctly classify individuals as probationers or non-probationers was 80.3% (Table 31). Thus women were more likely to get incarcerated if they had a serious criminal history, were black, had a drug history, were younger, were not well 102 aaa. aN ..me~..¢. ..m.m. a.aa. aa.aa. mmv.e. amaze. meme. .o co.ue.mccoa mucmu.m.cm.m .m.a mememm-.;a mx..3 acmmcmm .eu.eocea me.e>:mm.m m.m. emmNa. ..Naa.... aaa. ..m.m. aN mam.a. ammo.msm mam. aemNa. mmmme.... aaa. aNm.m. m. mmm.a. mmcmga mo mmmcmeo.emm .mm. a.maN. meame.... aaa. mNm.a. m. .mNm.. mummcc< m..cm>ea aaa. .am.N. .NmNN.... aaa. mem.m. .. a.m... mmee< .ocom.< a.m. .mmmN. aNamm.a.. aaa. mmm.m. m. .mmm.. mu..oe em..ma cme.mu:a NNm. .mm.m. mmemm.m.. aaa. a.m.m. m. mm.em. :o.umueam -.m.m mo mmc< um..muma umoz mmN. .qua.. mmm...m.. aaa. mmamm. a. mNN... mu.ee< m>.pcoeq:m cp.z me.>.a mmN. Naamm.. mmNmm.¢.. aaa. mm.mm. m. ..m.m. .m.m c. mgoum.: .me.s.ca mmN. Nmmma.. emNam.m.. aaa. mmmma. N. maaN.. am.ccmz a.N. mam.N.. .mm.N.N.. aaa. ..emm. .. mm..m. m..EE :mxocm .m.. NNaaa.N NNaaa.... aaa. MNmmm. a. aammm.. mam.eogm .meomcmm .N.. meamm.N .amem.mm. aaa. mmmmm. m a...m.. mmu..esoou< mo LmeE:Z\mu...oe am..mu E.pu.> ee.. mmmm..N Nmm.m.mm. aaa. .m.mm. m mamme.. .m.m mo mucmumm ma.. mmNNa.N m...e.em.. aaa. mmmmm. . aa.am.. mu.ea< mam cmca..ca memo. 5a.: me.>.. moo. maNma.¢ m.aaa..m. aaa. mm.mm. a ammmm.N mu..=a uoz mo mm.m a.a. mamNm.m N.Nm...m. aaa. Nmma.. m .m.mm.v .m>ma .m:o.ueu:eu aNa. ammm..m ma.m...m. aaa. aNN... a .mNa..m memeumm uaocu.3 mcmnpoz cme.a aaa. ememN.m mNmem.me. aaa. amm... m mamNa.m mgoum.: mega .aa. mammm.a. .mmma..m. aaa. mmam.. N Namm..m mumm aaa. mmm.m.mN. mmm.m.mN. aaa. mmm¢.. . mmm.m.mN. «comma muemuemmmcm .0 eo.pmeemEEoumm ace mgoum.= .me.e.gu > momm meeEea ame=.mc. m>oEmm Lo mu:mu.m.em.m :. mmcmea > mama mo:mo...cm.m mx..3 amazez Lmucm on m m.am.cm> co.uueem acme.s.cum.a .mu.c.esm me» e. mm.em.ce> may go. mmcoum > moem nee cease; mx..3 om u4memm.m ..m. m.aaa. m.am..mmN aaa. aNemm. NN mmmNa. m.mc.m Nmm. aNmea. .a.¢..mmN aaa. mNemm. .N aamNa. .goum.: .mu.cmz-mmc< am..mmma pee: am.. aNm.a. Nmmma.mmN aaa. mNemm. aN mmmea. emu...a co.umeocm me umzumoeea< m.aomm .m.. ...... Na..a.mmN aaa. mmemm. m. Nmmma. mmcmwma muemeocm .mm. .a.MN. emmmm.mmN aaa. .eemm. m. NmmN.. “ceaemmma cu.z m:.>.a cm:e..ea «mm. meaam. ammam.mmN aaa. mmemm. .. mNmm.. m8.3583. .o Lmesez a.a. Naamm. mmmmm.mmN aaa. .mmemm. m. aNNmm. mu..om am..ma cme.muea N.m. .amm.. mmmmm.¢mN aaa. mmmmm. m. .Nmee. meo.< mc.>.a ..m. mm.Na.. mNNmm.mmN aaa. .mmmm. a. mmNmm. mem.eocm .meomcme m.m. Nmm.a.. amemm.NmN aaa. aN.mm. m. .mm.m. mmmcc< aaa.. mo ms.. um mm< Nma. aNmmm.N .m.em..mN aaa. m.mmm. N. mmeNm.. xgoum.: x..Emm-emc< am..muma awe: mma. a.aam.m m.e.a.mmN aaa. Nmamm. .. mm.ma.N amummcc< ems: mm< Nma. .Nmmm.N .me.e.mmN aaa. aNemm. a. mmmmm.. co.uem:mm-mmc< em..muma umoz .ma. .m.ea.m mem.m.NmN aaa. mmamm. m mmmm... emummm utoeeem .e.uom mma. aN.mm.N m.emm.meN aaa. .mmmm. m .N.a... meopm.: .e:.s.ca-mmc< um..mema “we: mma. emmmm.m mm.¢m.meN aaa. amNam. . ammm..N mu.ea< use :mce..za eu.3 m:.>.4 N.a. mm.mm.m N.mmm.NeN aaa. ..mam. a Nmeem.m :o.uumuma .o mac: cmepa .aa. ammmN.. m.NNm.mmN aaa. mmN.m. m mmmme.e mu..oe am..ma e.uu.> aaa. mm..e.m mNmmm.mNN aaa. mmm.m. a meNmN.m mu..=a uoz mo em.m maa. .mmmm.m mmamm.aNN aaa. N.mNm. m mamNm.e mmcmmma me» we mmmcmeo.cmm aaa. .m.mN.e. NmaNa.e.N aaa. momma. N aem.N.m mgoum.: .me.s.cu aaa. a.mm..mm. a.mm..mm. aaa. «m.aa. . a.mm..mm. “comma mucmucmmmcm .o =o.umvcmesoumm > momm mange; uma=.uc. m>oEmm to mo:mu...cm.m :. mmemca > momm muemu...cm.m mx..z cmeEez cmuem o» m mm.am.em> :o.umc:m ucmc.E.LUm.a .mco.umm me» :. mm.am.cm> meg co. mmcoum > momm ace cease; mx..3 Nm m4m