Criminology Police

8/12/2019 Criminology Police

1/8

Citation: 53 J. Crim. L. Criminology & Police Sci. 333 1962

Content downloaded/printed from

HeinOnline (http://heinonline.org)

Tue Apr 1 13:08:51 2014

-- Your use of this HeinOnline PDF indicates your acceptance

of HeinOnline's Terms and Conditions of the license

agreement available at http://heinonline.org/HOL/License

-- The search text of this PDF is generated from

uncorrected OCR text.

-- To obtain permission to use this article beyond the scope

of your HeinOnline license, please use:

https://www.copyright.com/ccc/basicSearch.do?

&operation=go&searchType=0

&lastSearch=simple&all=on&titleOrStdNo=0091-4169


2/8

JUDICIAL BACKGROUNDS AND CRIMINAL CASESSTUART S. NAGEL*

Various scholars of the judicial process have com-piled data on differences in the backgrounds ofAmerican judges, but they have not shown thatthese background characteristics correlate withdifferences in the decisions of the judges analyzed.'Various other scholars of the judicial process havecompiled data on the different decisional tenden-cies of American judges, but they similarly havenot shown that these decisional propensities corre-late with differences in the backgrounds of thejudiciary.2 It is the purpose of this paper to explorethe empirical relationships between various back-ground and attitudinal characteristics of judgesand their decisions in criminal cases.3

THE RESEARCH DESIGN4The judges analyzed consist of the 313 state

and federal supreme court judges listed in the 1955 Professor Nagel, a member of the Illinois bar, is

Assistant Professor in the Department of Political Sci-ence of the University of Illinois.The writer gratefully thanks the Political Theoryand Legal Philosophy Committee of the Social ScienceResearch Council for providing funds for conducting astudy of relations between judicial characteristics andjudicial decision-making of which study this paper is apart. Thanks are also owed to the members of the 1960SSRC Summer Research Institute on the Administra-tion of Criminal Justice for the suggestions which theymade relevant to the completion of this paper.1E.g. Schmidhauser, The Justices of the SupremeCourt: A Collective Portrait,3 M.W.J. or PoL SCL 1(1958); EwING, THE JUDGES OF THE SUPREME COURT1789-1937 (1938); Mott, Judicial Personnel 167 ANNALS 143 (1933).2E.g. Gaudet, Individual Differences in SentencingTendencies of Judges, 32 ARcmiEs oF PSYcHOLOGY 1(1938); Iverson, Human Element in Justice, 10 J.Cm. L. C. 90 1919); PITCHETT, THE ROOSEVELTCOURT (1948); SCnUERr, QUANTITATIvE ANALYsIs OrJUDICIAL BEHAVIOR 1959).Non-statistical speculations on the relations be-tween judicial characteristics and judicial decisionsinclude Haines, General Observations on the Effects ofPersonal Political and Economic Influences on theDecisions of Judges 17 ILL. L. REv. 96 1922); FRANx,COURTS ON TRIAL 146-56, 165-85 1950); CARR, THESUPREME COURT AND JUDICIAL REVIEW 231-57 (1942);and Hall, Determinationof Methodsfor Ascertaining theFactors that Influence Judicial Decisions in Cases In-volving Due Processof Law 20 Am. POL. Sci. REv. 127 1926).

4 For a more detailed analysis and justification of thejudges, the cases, and the analysis used in the researchdesign than can be given here see Nagel, Testing Rela-tions Between Judicial Characteristics and JudicialDecision-Making (a mimeographed paper presented atthe 1961 Midwest Conference of Political Scientists).Copies of this paper are available on request from thewriter.

Directory of American Judges;5 15 judges so listedleft the bench before the end of the year, however,and as a result were included in only certain por-tions of the study.8 The background characteris-tics of the judges were determined by consultingthe Directory of American Judges, Who s Who inAmerica,7 the Martindale-Hubbell Law Directory,8and the governmental directories published bymany of the states. The attitudinal characteristicsof the judges were determined on the basis of theiranswers to mailed questionnaires.

The cases analyzed consist of the full-courtcriminal cases which these judges heard in 1955.By full-court cases is meant cases heard by alljudges in the sample from the court involved. Bycriminal cases is meant cases in which one p rtywas charged with an act subject to fine, imprison-ment, or other penalty owed to the collectivityrather than to the individuals who may have beenparticularly harmed. The cases involved both ap-peals and habeas corpus proceedings, and theycentered around questions of guilt, punishment, orprocedure. Tax and business regulation cases wereexcluded because they were analyzed separatelyin the larger study of which this paper is a p rt

Each judge was given a decision score represent-ing the proportion of times he voted for the defenseout of all the times he voted in the criminal cases.For example, in the 21 Pennsylvania criminalcases, 9 Justice Arnold voted 3 times for the de-

DIRECTORY or AMERICAN JUDGES (Liebman ed.1955). This source is sometimes referred to herein as theDIRECTORY.6The term supreme court is used as a synonymfor highest court or court of last resort. Statementsin this paper describing the background characteristicsof the judges are based on all 313 judges, except forthose who did not supply the relevant background in-formation. Statements describing the relations betweenbackgrounds and decisional tendencies, however, arebased on 298 judges 313 minus the 15 judges who leftthe bench before the end of the year), except for thosewho did not supply the relevant background informa-tion, who sat only on unanimous cases, or who sat oncourts that were homogeneous as to the relevant back-ground characteristics.726 27, 28 WHO S WHO IN AMERICA (Sammons ed.1954, 1956, 1958).

8MARTINDALE-HUBBELL INC., MARTINDALE-HUB-BELL LAW DIRECTORY 1955).Commonwealth v. Burdell, 38 Pa. 43, 110 A.2d193 (1955); Commonwealth v. Edwards, 38 Pa. 52,

110 A.2d 216 (1955); Commonwealth v. Mackley, 38Pa. 70, 110 A.2d 172 (1955); Commonwealth v. Grays,38 Pa. 77, 110 A.2d 422 (1955); Commonwealth ex.


3/8

COMMENTS AND RESEARCH REPORTSfense, resulting in a decision score of .14. Sometimesa judge did not vote clearly for either the prosecu-tion or the defense, as where he concurred in partand dissented in part. In one Pennsylvania case, 0Justice Musmanno cast such a half-way vote, re-sulting in a decision score of Y out of 21 or .55.

The data were analyzed to determine what judi-cial characteristics, if any, were associated with adecision score above the average for a given court.For example, the question was posed whether beinga Democrat rather than a Republican is associatedwith a decision score above the average for one'scourt. The answers to this and similar questionsfor the data used are presented in the Tablewhich accompanies this paper.

The Table includes data for only those courtson which both groups compared are present, be-cause if, for example, there are all Democrats or allRepublicans on a court, then comparisons withinthe court between judges from the two parties can-not be made. The Table also includes data for onlynon-unanimous cases, because the unanimouscases have no bearing on who is above or below theaverage decision score of a court.

The probability scores listed in the far rightcolumn represent the probability of the observeddifferences occurring merely by chance, given thenumber of judges involved in each group. Wherethe scores are less than .05 (i.e., less than 5 out of100), the differences have been regarded as statis-tically significant and not merely attributable tochance, in accordance with conventional statis-tical procedures. In the discussion which follows,each row will be discussed separately, and an illus-rel Dunn v. Ruch, 380 Pa. 152, 110 A.2d 240 (1955);Commonwealth ex rel. Lane v. Baldi, 380 Pa. 201, 110A.2d 409 (1955); Commonwealth v. Chaitt, 380 Pa.532, 112 A.2d 379 (1955); Commonwealth v. LaRue,381 Pa. 113, 112 A.2d 362 (1955); Commonwealth v.Lane, 381 Pa. 293, 113 A.2d 290 (1955); Common-wealth v. Thompson, 381 Pa. 299, 113 A.2d 274 (1955);Commonwealth v. Mason, 381 Pa. 309, 112 A.2d 174(1955); Commonwealth v. Cisneros 381 Pa. 447 113A.2d 293 (1955); Commonwealth v. Bolish, 381 Pa.500, 113 A.2d 464 (1955); Commonwealth ex rd. Mat-thews v. Day, 381 Pa. 617, 114 A.2d 122 (1955); Com-monwealth v. Farrow. 382 Pa. 61, 114 A.2d 170 (1955);Commonwealth v. Capps, 382 Pa. 72, 114 A.2d 338(1955); Commonwealth v. Wable, 382 Pa. 80, 114 A.2d334 (1955); Commonwealth ex rt. Taylor v. Superin-tendent of the County Prison, 382 Pa. 181, 114 A.2d343 (1955); Commonwealth ex rt. Bishop v. Maroney,382 Pa. 324, 114 A.2d 906 (1955); Commonwealth v.Thomas, 382 Pa. 639, 117 A.2d 204 (1955); Common-wealth v. Moon, 383 Pa. 18, 117 A.2d 96 (1955).1 Commonwealth v. Edwards, 380 Pa. 52, 110 A.2d216 (1955). YUKxR, A GUIDE TO STATISTICAL CALCULATIONS

64 66 (1958); SIEGEL, NoN-PARAMeTRIC STATISTICSFOR THE BEHAVIORAL SCIENCES 13-14 (1956).

trative court will be described for each relationshipwhich falls below the .05 level of chance proba-bility.

THE RESEARCH FINDINGSPolitical artyAffiliation

The first row of the Table tends to indicate thaton the bipartisan supreme courts Democratic andRepublican judges do differ from one another indeciding criminal cases. In 1955, 15 bipartisan stateand federal supreme courts decided at least onenon-unanimous criminal case on which all theirjudges sat. These courts were comprised of 85judges who gave a party affiliation in the sourcesconsulted. Fifty-five per cent of the 40 Democratswere above the average of their respective courtson the decision score, whereas only 31 per centof the 45 Republicans were above the average oftheir respective courts on the decision score.

The California Supreme Court illustrates thisstatistically significafit difference. Two of the Cali-fornia judges, Justices Carter and Traynor, de -clared themselves as Democrats in the Directoryand 2, Justices Shenk and Spence, declared them-selves as Republicans. Justices Edmonds, Gibson,and Schauer did not indicate party affiliation. It isunusual for so many judges on a supreme court notto give party affiliation. The California SupremeCourt, however, is not an elected court. Partly toeliminate partisan influence, the judges are ap-pointed initially by the governor with the approvalof a Commission on Qualifications, and they appearon the ballot for voter approval every 12 yearsthereafter. In 1955 only Missouri had a similarsystem of judicial election. In spite of this attemptto eliminate partisan divisions, Democrats Carterand Traynor were on opposite sides of RepublicansShenk and Spence in a large number of cases ofdifferent types. All 7 judges of the court heard 14criminal cases together in which non-unanimousdecisions were reached. 2 The 2 Democrats had an

12 Lyons v. Superior Court, 43 Cal. 2d 755, 278 P.2d681 (1955); In re Chessman, 44 Cal. 2d 1 279 P.2d 24(1955); People v. Sykes, 44 Cal. 2d 166, 280 P.2d 769(1955); Bompensiero v. Superior Court, 44 Cal. 2d 178,281 P.2d 250 (1955); In re Bartges, 44 Cal. 2d 241, 282P.2d 47 (1955); People v. Cavanaugh, 44 Cal. 2d 252,282 P.2d 53 (1955); People v. Terry, 44 Cal. 2d 371,282 P.2d 19 (1955); People v. Cahan, 44 Cal. 2d 434,282 P.2d 905 (1955); People v. Berger, 44 Cal. 2d 459,282 P.2d 509 (1955); People v. Jackson, 44 Cal. 2d511, 282 P.2d 898 (1955); In re Hess, 45 Cal. 2d 171,288 P.2d 5 (1955); People v. Acosta, 45 Cal. 2d 538,290 P.2d 1 (1955); People v. Tarantino, 45 Cal. 2d590, 290 P.2d 505 (1955); Calhoun v. Superior Court,46 Cal. 2d 18, 291 P.2d 474 (1955).

[Vol. 53


4/8

COMMENTS ND RESF RCH REPORTS

TABLE owJUDGES OF DIFFERING BACKGROUNDS AND ATTITUDES DIFFER IN TiHEIR CRIMINAL C SE DEcisIONs

Based on the non-unanimous cases of the state and federal supreme courts of 1955 on which both groups beingcompared are present.I of - Number of Group Group 2Judges In- bove bove

Group I Hypothesized Group 2 tltypothesized volve in Their Their Probability of theto be less defense minded) i to be more defense minded) Each Group ourt Court e - sitiveueifferenceAverage Average ie Being D~ueto Chanceon the on the

ston DeciioScore Score*

PARTYRepublicans

PRESSURE GROUPSMembers of a busi-

hess groupMembers of ABA

Members of a na-tivist groupOCCUPATIONS

Former businessmenFormer prosecutors

EDUCATIONAttended high tu-

ition law schoolAG E

Over age 65GEOGRAPHY

Practiced initially insmall townRiLGION AND ANCES-

TRAL NATION-.ALITY

ProtestantsHigh income Prot.

denominationOnly British An-

cestryATTITUDES

Low general liberal-ism score

Low criminal liberal-ism score

DemocratsDid not indicate such

membershipDid not indicate such

membershipDid not indicate suchmembershipDid not indicate such

occupationDid not indicate such

occupationAttended low tuition

law schoolUnder age 60Practiced initially in

large city

CatholicsLow income Prot. de -

nominationPart non-British an-

cestryHigh general liberalism

scoreHigh criminal liberal-

ism score

31C C4737

36

3236

54

4335

314138

2727

Decision Score: proportion of times voting for the defense in criminal cases.

+25

Less than .05.20 to .50Less than .05

.20 to .50

.05 to .20Less than .05

.20 to .50

Negligible tliff

Negligible diff.

Less than .05.05 to .20

+9 1 .05 to .20

+30 1 Less than .05Less than .05

average decision score in these cases of 85 per centfor the defense whereas the 2 Republicans had anaverage decision score in the same cases of only 18per cent for the defense. In the famous Chessmancase of 1955 for example the only 2 dissenters infavor of the defense were Democrats Carter andTraynor.Other supreme courts in which the Democrats

had a higher average decision score for the defensethan did the Republicans include the supremecourts of Colorado Idaho Maryland MichiganMontana New Jersey Rhode Island NorthDakota and Pennsylvania. Of the 15 courts withqualifying criminal cases only the Illinois NewYork Ohio Utah and federal supreme courtsfollowed an opposite pattern although not to as

59621


5/8

COMM3ENTS ND RESEARCH REPORTSgreat extent as the 10 courts which followed thegeneral pattern.PressureGroup Affiliations

Many of the judges were members of pressuregroups which endorse various kinds of legislation.The types of pressure groups most frequently men-tioned were professional groups (e.g, the AmericanBar Association), veterans groups (e.g., theAmerican Legion), business groups (e.g., chambersof commerce), and nativist groups (e.g., the Sonsof the American Revolution). Decision scores ofthe judges who were members of these groups didnot differ to a statistically significant extent fromthe decision scores of the non-member judges, withthe exception of the scores pertaining to member-ship in the American Bar Association. The Tableshows that 52 per cent of the judges who indicated(in the Directory, Who s Who, or Martindale-Hub-bell that they were not members of the A.B.A. haddecision scores above the average for their respec-tive courts, whereas only 37 per cent of the judgeswho indicated that they were members of theABA had such scores.On the United States Supreme Court, for in-stance, Justices Black, Douglas, Frankfurter, andMinton did not indicate ABA membership, whileMr. Chief Justice Warren and Justices Burton,Clark, Harlan, and Reed, did indicate such mem-bership. In the 9 full-court non-unanimous crimi-nal cases of 1955, the 4 non-ABA members hadan average decision score of 70 per cent for the de -fense, whereas the 5 ABA members had only anaverage decision source of 51 per cent. 3 Subsequentto 1955 the Chief Justice withdrew from the ABAon ideological grounds. If Warren were considereda non-ABA member in 1955, then the averagedecision score of the non-ABA group would moveup to 71 per cent for the defense, and the averagedecision score of the ABA group would movedown to 45 per cent for the defense.Pre-JudicialOccupations

Many of the judges indicated that they hadformerly held occupations other than the privatepractice of law. The types of occupations most fre-quently mentioned were prosecuting attorney, 3ell v. United States, 349 U.S. 81 (1955); In theMatter of Murchison, 349 U.S. 133 (1955); Quinn v.United States, 349 U.S. 155 (1955); Emspak v. UnitedStates, 349 U.S. 190 (1955); Bart v. United States, 349U.S. 219 (1955); Williams v. Georgia, 349 U.S. 375(1955); Donaducy v. Pennsylvania, 349 U.S. 913

(1955); United States ex rel. Toth v. Quarles, 350 U.S. (1955); Michel v. Louisiana, 350 U.S. 91 (1955).

legislator, corporation counsel, businessman,teacher, public administrator, attorney general,and regulatory agency attorney. The pre-judicialoccupations thought to have the most relevance todecisional differences among judges in criminalcases were those of businessman and prosecutingattorney. Although judges who were former busi-nessmen tended to have a lower decision score forthe defense in criminal cases than did judges whowere not, the difference was not quite statisticallysignificant. However, 50 per cent of the judgeswho did not indicate being former prosecutors haddecision scores above the average for their respec-tive courts, whereas only 36 per cent of the judgeswho did indicate being former prosecutors hadsuch scores.

The Pennsylvania Supreme Court exemplifiesthe general pattern found on supreme courts havingsome judges who were former prosecutors. On thiscourt, 3 of the judges (Justices Arnold, Bell, andChidsey indicated tLey had been prosecutors be -fore becoming judges. In the 9 nonunanimouscriminal cases which the full court heard in 1955the other 4 judges in the court (Justices Jones,Musmanno, Stearne, and Stern) had an averagedecision score of 26 per cent for the defense,whereas the 3 judges who were former prosecutorshad an average decision score for the defense of only7 per cent 4 Mr. Justice Musmanno, the famousdefense lawyer in the Sacco-Vanzetti case andother criminal cases, alone had a decision score of94 per cent for the defense in these cases.Education, Age, and Geography

Approximately one-third of the 313 supremecourt judges serving in 1955 went to law schoolswhose annual tuition was under 120 in 1927 (theearliest year for which school-by-school tuitionfigures are available , and approximately one-third went to law schools whose annual tuitionwas over 240.1' A higher percentage of judges who

1 Commonwealth v. Edwards, 380 Pa. 52 110 A.2d216 (1955); Commonwealth ex rel. Dunn v. Ruch, 380Pa. 152, 110 A.2d 240 (1955); Commonwealth ex rel.Lane v. Baldi, 380 Pa. 201, 110 A.2d 409 (1955);Commonwealth v. Chaitt, 380 Pa. 532, 112 A.2d 379(1955); Commonwealth v. LaRue, 381 Pa. 113,A.2d 362 (1955); Commonwealth v. Mason, 381 Pa.309, 112 A.2d 174 (1955); Commonwealth v. Cisneros,381 Pa. 447, 113 A.2d 293 (1955); Commonwealth x rel.Taylor v. Superintendent of the County Prison, 382Pa. 181, 114 A.2d 343 (1955); Commonwealth v.Thomas, 382 Pa. 639, 117 A.2d 204 (1955).1 uition figures for the last law school each judgeattended were taken from REED, REviEw OF LEGALEDUC TION IN THE UNITED ST TES AND C N D FORTHlE YEAR 1928 (1929).

[Vol. 53


6/8

COMM NTS ND RESE RCH REPORTSwent to low tuition law schools (under 120) haddecision scores above the average for their respec-tive courts than did judges who went to high tui-tion law schools (over 240). This difference, how-ever, was not statistically significant. Likewisethere was no statistically significant difference be -tween judges in the bottom third age group (under60) and judges in the top third age group (over 65).

There was also no statistically significant differ-ence between judges who practiced law initially insmall towns with populations under 5,000 (thebottom third among the judges) and judges wh opracticed initially in large cities with populationsover 100,000 (the top third among the judges). 6Because in this study comparisons are made onlywithin courts, comparisons are not made betweenjudges of different regions. If outhern supremecourts were compared with northern supremecourts, however, one might hypothesize that thesouthern courts would have a higher per cent ofjudgments granted to the defense than would thenorthern courts because (1) violence is possiblymore condoned in the south than in the north, ex-cept violence by a Negro against a white, (2) lessefficient southern police are possibly more likelyto bring innocent persons to trial than are northernpolice, and (3) southern lower courts are possiblymore likely to commit reversible error than arenorthern lower courts.Religion and ncestralNationality

Most of the judges with entries in the Directoryof mericanJudgeslisted their religion in responseto the Directoryquestionnaire. There were too fewJewish supreme court judges to make comparisonsbetween Jewish and non-Jewish judges. There were11 supreme courts, however, which had some Cath-olic and some Protestant judges, and which heardsome non-unanimous criminal cases with all judgespresent. These 11 supreme courts had 57 judgeswho indicated they were either Catholics or Protes-tants. 56 per cent of the 18 Catholic judges haddecision scores above the average for their respec-tive courts, whereas only 31 per cent of the 39Protestants had such scores.

This statistically significant difference is illus-trated by the New Jersey Supreme Court. JusticesBrennan and Heher indicated they were Catholics;

cPopulation figures for the place of initial law prac-tice of each judge were taken from 1 U.S. BmEAuoF THE CENsUs, DFP'T oF CoNeRcE, CENSUS oFPOPULATION 178-320 (1920). The year 1920 was thecensus year nearest to the year when the average judgeamong the 313 judges began practicing law.

Justices Oliphant, Vanderbilt, and Wachenfeldindicated they were Protestants; and JusticesBurling and Jacobs gave no religious affiliation. Inthe 12 non-unanimous criminal cases which thefull New Jersey Supreme Court heard in 1955, the2 Catholics had an average decision score for thedefense of 52 per cent, whereas the 3 Protestantshad an average decision score for the defense ofonly 28 per cent.17

The members of certain Protestant denomina-tions are traditionally thought to have a higheraverage income than that of members of otherProtestant denominations. The relatively highincome denominations are the Congregationalist,Episcopalian, Presbyterian, and Unitarian; andthe relatively low income denominations are theBaptist, Lutheran, and Methodist. 8 When judgesfrom each of the two groups sitting on the samesupreme court criminal cases were compared, thejudges from the relatively low income denomina-tions tended to have i higher decision score for thedefense than did the judges from the relativelyhigh income denominations. The difference, how-ever, was not quite statistically significant.

The ancestral nationality of each judge can beroughly determined by taking each judge's pa-ternal and maternal family names or their com-ponent parts to Elsdon C. Smith's Dictionary ofFamily Names 1956), which in dictionary formgives the nationality origin of over 10,000 familynames. If one compares judges whose ancestralnationality is exclusively British (which includesEnglish, Scotch, or Welsh) with judges on thesame court whose ancestral nationality is at leastpartly non-British (to the extent determinable inthe Smith book), one finds that the judges of par-tially non-British derivation tend to have higherdecision scores for the defense than do the judgesof wholly British derivation on the same courts.This difference, however, is not quite statisticallysignificant.

7State v. Schmelz, 17 N.J. 227, A.2d SO (1955);State v. Newton, 17 N.J. 271, A.2d 272 (1955);State v. Low, 18 N.J. 179, 113 A.2d 169 (1955); Statev. Cianoi, 18 N.J. 191, 13 A.2d 176 (1955); Johnson v.State, 18 N.J. 422, 114 A.2d (1955); In the Matter ofWhite, 18 N.J. 449, 114 A.2d 261 (1955); State v.Haines, 18 N.J. 550, 115 A.2d 24 (1955); State v. Wise,19 N.J. 59, 115 A.2d 62 (1955); State v. Fary, 19 N.J.431, 117 A.2d 499 (1955); State v. D Ippolita, 19 N.J.540, 117 A.2d 592 (1955); State v. DeMeo, 20 N.J.118 A.2d (1955); State v. Kociolek, 20 N.J. 92, 8A.2d 812 (1955). 8Allinsmith, Religious Affiliation and Politico-Eco-nomic Attitudes 12 PUBLIC OPINION Q 377 (1948).

96


7/8

COMMENTS ND RESE RCH REPORTS

Off the Bench ttitudesIn the spring of 1960 a mailed questionnaire was

sent to each of the 313 state and federal supremecourt judges of 1955 in order to determine theirattitudes on various issues. One hundred and nine-teen of the judges returned answered question-naires. The questiqnnaire represented a condensedand revised version of a liberalism-conservatismquestionnaire written by H. J. Eysenck.19 Thejudges were asked to indicate whether on thewhole they agreed a lot, agreed a little, neitheragreed nor disagreed, disagreed a little, or dis-agreed a lot with a list of statements. The ques-tionnaire was primarily designed to measure thedegree of sympathy a respondent has for lessprivileged groups and the degree of acceptance hehas toward long run social change. These two com-ponents make up what is generally referred to asliberalism.20 The questionnaire was scored in sucha way that a respondent could receive a liberalismscore ranging from 41 to 195. The median liberal-ism score actually received was 109. Seventeensupreme courts that heard full-court non-unani-mous criminal cases had at least one judge with ascore over 109 and at least one judge with a scoreat or below 109. Fifty-seven per cent of the judgesscoring above 109 were above the average of therespondents of their respective courts on the de-cision score, whereas only 27 per cent of the judgesscoring at or below 109 were above the average ofthe respondents of their respective courts on thedecision score. This difference is statistically sig-nificant, but a specific example cannot be givenbecause the judges were promised anonymity ifthey responded to the questionnaire.

There was one particularly relevant statementto which the judges were asked to respond by indi-cating whether and how much they agreed or dis-agreed. This statement read Our treatment ofcriminals is too harsh; we should try to cure not topunish them. Twenty-four of the 119 respondingjudges indicated they strongly disagreed on thewhole with the statement, 48 disagreed but no tstrongly, 22 neither disagreed nor agreed, 20 agreedon the whole but not strongly, and 5 agreedstrongly. The average responding judge was thusin between disagreeing mildly and being neutral.Fifty-nine per cent of those who were neutral orwho agreed with the statement were above the 9YsENc PSYCHOLOGY OF POLrTICS 122-24 1954).2 MACIVER, Tr WEB OF GOVERNMENT 215-171951).

average of the respondents of their respectivecourts on the decision score, whereas only 27 percent of those who disagreed with the statementwere above the average of the respondents of theirrespective courts on the decision score. This differ-ence is statistically significant.

REASONS AND REMEDIESHow might one account for the relationships be-tween judicial characteristics and judicial decision-

making that have been described? Some of therelationships found are easily attributable tochance. Others, however, are not. The latter arethose where the odds are more than 95 to 5, or 19to 1, of obtaining the differences purely by chancegiven the size of the differences and the size ofthe groups. They include the differences betweenDemocratic arid Republican judges, non-ABAmembers and ABA members, non-former prose-cutors and former prosecutors, Catholic :udgesand Protestant judges, and relatively liberal judgesand relatively conservative judges as measured bytheir off-the-bench attitudes.

To some extent a criminal case represents a con-flict of social groups, in that the defendant gener-ally tends to be a member of the lower-middle orworking class (particularly if tax and businessregulation cases are analyzed separately),21 andthe prosecutor tends to be a member of the upper-middle or upper class, enforcing laws promulgatedby upper-middle and upper class legislators andjudges.n Mass data show that persons holdingcertain positions e.g., being a Democrat or aCatholic) with respect to background characteris-tics e.g., party or religion) tend to have greatersympathy for lower economic and social groupsthan do persons holding obverse positions.2 Giventhe nature of the average criminal case, judgesholding such positions with respect to backgroundcharacteristics are therefore likely to have a higherdecision score for the defense than do judges hold-ing obverse background positions. The correlationbetween a judge's position on background char-acteristics and his relative degree of sympathy forlower economic and social groups may account forthe differences found concerning party, pressuregroups, religion, and liberal-conservative attitudes.

2 TAFT, RIMINOLOGY 131-33 1950).2 2MATTHEwS THE SOCIAL BACKGROUND OF POLITI-CAL DECISION-MAKERS 23-30 1954).

2 CAMPBELL, GURIN MILLER, Tl VOTER DE-cr s 1954); TURNER PARTY AND CONSTITUENCY:PRESSURES ON CONGRESS (1951).

JVol. 53


8/8

CO IE VTS AND RESE RCH REPORTS

However it probably does not account for thedifferences found between former prosecutors andtheir opposite number; judges who are formerprosecutors are probably on the average not sub-stantially more or less ideologically liberal thanjudges who have not been prosecutors. Theirlower decision scores for the defense are possiblymore attributable to a relatively pro-prosecutionframe of reference which caused them to becomeprosecutors or which they acquired or had rein-forced when they served as prosecutors.

Many devices are available for minimizing theinfluence of judicial backgrounds including the

availability to defendants of easy appeals the re-quirement that judges write opinions to justifytheir decisions the use of multi-judge courts withjudges of diversified backgrounds and the draftingof clearer and more detailed substantive statutesthereby limiting the area of judicial discretion.Because criminal cases frequently involve value-oriented controversies however and because d iffer-ent background and attitudinal positioPs tend tocorrespond to different value orientations therewill probably always be some correlation betweenjudicial characteristics and judicial decision-mak-ing in criminal cases.

Documents

Criminology Police