200 JUDICATURE Volume 90, Number 5 March-April 2007

In November 2006, voters in several states faced ballotmeasures that would have crippled the ability of statecourts to do the work we expect of them. The “JAIL 4

Judges” referendum would have subjected SouthDakota’s judges to civil and criminal penalties for decid-ing cases in ways that offended a small minority of citi-zens. A proposal in Montana threatened judges withspecial recall elections if they made unpopular decisions.In Colorado, a ballot initiative sought to penalize judicialexperience by imposing retroactive term limits for appel-late judges, a measure that would have ousted nearly halfthe sitting appellate bench. And Oregon voters consid-ered whether to elect their appellate judges by geo-graphic district, apparently intending to tie judicialcandidates more closely to the values of a particularregion of the state.

Each of these initiatives was couched as an effort to holdjudges more accountable, “accountability” being defined(implicitly or explicitly) by their sponsors as adherence tothe will of the majority. The chief architect of the Colorado

initiative, for example, argued that term limits would makethe judiciary as a whole “more responsive to the sovereignwill of the people.” Similarly, in Montana, proponentsargued that recall of individual judges would “be a power-ful tool for judicial accountability and democratic over-sight of a branch of government that for too long has beentoo removed from the will of the people.”

Although none of these initiatives was ultimately suc-cessful, those who want an effective, impartial judiciarycan ill-afford to be complacent about the conditions thatfueled their placement on the ballot. The public isincreasingly being asked to hold judges accountable forthe outcomes of specific cases, rather than the appropri-ateness of the process used to reach those outcomes. This

The authors would like to thank Institute for the Advancement of theAmerican Legal System board members Russell Wheeler of the GovernanceInstitute and Brookings Institution, and Lynn Mather of the Baldy Center forLaw and Social Policy at the University of Buffalo, for their invaluable assis-tance in preparing this article.

1. The study, entitled Shared Expectations: Judicial Accountability in Context, isavailable through the Institute’s website, www.du.edu/legalinstitute.

Using judicial performanceevaluations to promotejudicial accountability

by REBECCA LOVE KOURLIS and JORDAN M. SINGER

ESTELLE CAROL

www.ajs.org JUDICATURE 201

was not always the case. The time isripe to return “judicial accountabil-ity” to its traditional role: a necessarypartner, along with judicial inde-pendence, in ensuring an effectivejudicial branch.

This article summarizes the resultsof a recent comprehensive study of anexisting but underutilized approachto process-oriented judicial accounta-bility: judicial performance evalua-tion (JPE). The study, undertaken bythe Institute for the Advancement ofthe American Legal System at theUniversity of Denver, concluded thatif properly designed and executed,JPE can be an effective means ofbuilding appropriate, shared expecta-tions about the proper role of thejudiciary, and could be implementedin every American jurisdiction.1

A primer on JPENineteen states, plus the District ofColumbia and Puerto Rico, employsome form of a judicial perform-ance evaluation program (see“Overview of official judicial per-formance evaluation programs,”page 204). These programs vary intheir specifics—for example, theymay use slightly different criteria formeasuring judges’ performance, orseek information from somewhat dif-ferent sources, or share informationwith the public in different ways—but as a general rule all focus onwhether judges are managing casesefficiently, deciding them on thebasis of established facts and applica-ble law, explaining their decisionsclearly, and exhibiting proper court-room demeanor. In addition, regard-less of the differences in theirformats, JPE programs are uniformlyprocess-oriented, not outcome-ori-ented: what matters is whether thejudge handled a case in a balanced,fair, and efficient manner—notwhether the ultimate decision in thecase provoked limited or even wide-spread opposition.

Each judge is typically evaluatedby an independent commission con-sisting both of attorneys and non-attorneys. The commission providessurveys to attorneys, jurors, and oth-ers who have interacted with thejudge in a professional setting, ask-ing for anonymous responses toquestions about the judge’s profes-sional skills. In more comprehen-sive programs, the commission alsoreviews the judge’s case manage-ment statistics and written opinions,solicits public comments on thejudge’s performance, and conductsone or more interviews with thejudge. The commission then usesthe collected information to meas-ure each judge’s performanceagainst predetermined criteria.Because appellate judges typicallywork more collectively and have dif-ferent roles in the judicial system,they generally are subject to differ-ent criteria than trial judges or mag-istrates.

JPE programs have been mostcommonly used in states employing

A recent study concludesthat if properly designed

and executed, judicialperformance evaluations can be an effective meansof building appropriate,

shared expectations aboutthe proper role of thejudiciary, and could be

implemented in every jurisdiction.

Providing the results of judicialevaluations to the electoratethrough, for example, voterguides, builds trust andconfidence in the judiciary.

LEAVE YOUR PRINT

202 JUDICATURE Volume 90, Number 5 March-April 2007

the “Missouri Plan” or “NonpartisanCourt Plan,” so called because it wasfirst adopted by Missouri voters in1940. Under the Missouri Plan,judges are initially appointed to thebench through a merit selectionprocess and subsequently partici-pate in periodic retention elections.In a retention election, the judgeruns uncontested, and the onlyquestion on the ballot is “ShouldJudge X be retained in office?” Inmost cases, if the sitting judge wins asimple majority of affirmative votes,he or she will continue to hold theoffice for another term.2 In MissouriPlan states, JPE commissions gener-ally make the results of each judicialevaluation available to the publicwell in advance of the judge’s reten-tion election. Where judges do notface retention elections, individualevaluations are generally kept confi-dential.

The need for JPEJudicial performance evaluations arelikely to promote judicial accounta-bility in three ways. First, JPE pro-grams can provide a valuable sourceof information to voters about theirjudges and judicial candidates. Inmany states where judges face elec-tions to hold or remain in office, JPEprograms may provide the only sub-stantive, neutral source of informa-tion about judges on the ballot. Thisinformation is critically important.Without adequate, accurate informa-tion, voters generally decline to castballots in judicial elections.3 Further-more, many of those who do votechoose to elect or retain a judgebased not on the judge’s perform-ance, but on the judge’s ethnicity,gender, name, party affiliation, orlength of time on the bench, or evenfor no articulable reason whatsoever.4

Surveys in 2004—one of anational sample, another in NewYork—suggest that voters wouldmuch prefer to make informedchoices about their judges. Morethan two-thirds of respondents inthe national survey agreed that“receiving a nonpartisan voter guidecontaining background informationon judicial candidates would make

them more likely to vote in judicialelections.”5 A report on the NewYork survey said that 88 percent ofthe respondents “believe that voterguides are a useful way to educatethe public about judicial elections.”6

The inclusion of performance evalu-ations in voter guides can help fillthe information gap.

The limited empirical research onvoters’ use of performance evalua-tion information suggests that theinformation substantially informsvoting choices. In one study of fourmajor metropolitan areas in statesusing JPE, 66-76 percent of the vot-ers surveyed responded that the offi-cial evaluation information eitherhelped their voting decision orserved as the basis for that decision.7

A majority of respondents in each ofthe four cities who said they werefamiliar with the evaluation reportsalso stated that “the informationinfluenced their voting decisions,added confidence to their votingchoices, [and] made them morelikely to vote in judicial elections.”8

JPE programs can also buildshared expectations about the judici-ary by educating the public aboutthe specific qualities that make agood judge. If popular commentaryis any indication, the most funda-mental threat to judicial independ-ence today is pressure on voters to

“hold judges accountable” for politi-cally unpopular outcomes in specificcases, or to vote for judicial candi-dates based on those candidates’ per-sonal opinions on hot-buttonpolitical issues. That pressure reflectsclaims that too many judges aremerely “legislating from the bench,”and that judicial opinions are exam-ples of policymaking rather thanapplication of existing law.

For example, judicial candidatesin Iowa’s November 2006 electionwere asked by one coalition to com-plete a six-page questionnaire indi-cating their personal positions on(among other things) abortion, gaymarriage and civil unions, assistedsuicide, homosexual relationships,and the display of the Ten Com-mandments in public buildings andschools.9 Widespread use of JPE pro-grams can dilute this threat to judi-cial independence by shifting publicfocus away from political positions orparticular case outcomes and towardthe process of adjudication. JPE pro-grams can measure the characteris-tics expected from an independent,knowledgeable judge: impartiality,temperance, knowledge of the law,fair application of the law, and effi-ciency. The voter who thinks of ajudge in these terms, rather than as arobed policymaker, is arguably morelikely to vote carefully and objectively

74 JUDICATURE 271, 271 (1991); Anthony Cham-pagne, Tort Reform and Judicial Selection, 38 LOY.L.A. L. REV. 1483, 1492 (2005) (discussing “‘partysweeps’ in which popular top-of-the-ticket candi-dates have swept judges of the opposing party outof office and elected judges of a popular candi-date’s party for no other reason than that thejudges shared the popular candidate’s party affili-ation”); Griffin & Horan, supra n. 3, at 74; JonaGoldschmidt, Selection and Retention of Judges: IsFlorida’s Present System Still the Best Compromise?, 49U. MIAMI L. REV. 1, 6 (1994).

5. See Zogby International Survey of 1,204American voters, commissioned by Justice at Stakeand conducted March 17-19, 2004, cited in RandallT. Shepard, Electing Judges and the Impact on JudicialIndependence, 42-JUN TENN. B.J. 23, 25 & n.16(2006).

6. Feerick Commission Report, supra n. 3, at 39.7. Kevin M. Esterling and Kathleen M. Samp-

son, JUDICIAL RETENTION EVALUATION PROGRAMS INFOUR STATES: A REPORT WITH RECOMMENDATIONS 39(Chicago: American Judicature Society, 1998).

8. Id. at 40.9. See “Iowans Concerned About Judges, 2006

Judicial Voters’ Guide Questionnaire for JudicialCandidates” (on file with authors), available athttp://www.iowansconcernedaboutjudges.org/doc/Survey.pdf. Similar questionnaires greetedjudicial candidates in other states.

2. Two states, Illinois and New Mexico, requirea supermajority to secure retention.

3. See “Report of the Commission to PromotePublic Confidence in Judicial Elections” 38(2004) (on file with authors), available athttp://www.courts.state.ny.us/reports/Judicial-ElectionsReport.pdf (hereinafter “Feerick Com-mission Report”) (58% of New York voters saidthey did not vote in judicial elections because theylacked candidate information); see also Kenyon N.Griffin & Michael J. Horan, Patterns of VotingBehavior in Judicial Retention Elections for SupremeCourt Justices in Wyoming, 67 JUDICATURE 68, 72(1983) (finding a high rate of abstentions amongvoters who had no information on the judges fac-ing retention).

4. See Larry Aspin et al., Thirty Years of JudicialElections: An Update, 37 SOC. SCI. J. 1, 3 (2000); seealso Susannah A. Nesmith, 16 Judge Seats Draw 35Candidates, MIAMI HERALD, Sep. 1, 2006, at 6B(noting the electoral advantage of Hispanic andJewish-sounding surnames in Florida judicial elec-tions); Marie Hojnacki & Lawrence Baum, Choos-ing Judicial Candidates: How Voters Explain TheirDecisions, 75 JUDICATURE 300, 308-09 (1992) (not-ing voter reliance on low-information cues,including the gender of the candidates, in elec-tions for associate justices of the Ohio SupremeCourt in 1986 and 1988); Anthony Champagne &Greg Theilemann, Awareness of Trial Court Judges,

www.ajs.org JUDICATURE 203

in a judicial election.Finally, judges themselves stand to

benefit from the formal feedback ofan evaluation. Each evaluated judgereceives concrete information aboutthe strengths and weakness of his orher performance, creating individu-alized opportunities for professionalself-improvement. JPE programs canprovide judges with feedback thatsimply could not, or would not, becaptured through any othermedium. This is particularly true forinterpersonal performance issuessuch as courtroom demeanor, whicha judge cannot truly evaluate forhim- or herself and which lawyers,jurors, and litigants are unlikely tocomment upon except through for-mal, anonymous evaluations.10

Judicial support for JPEDespite the natural human aversionto being reviewed, sitting judges whohave participated in performanceevaluation programs have expressedsupport for them. A 1998 survey ofjudges in the four states with themost developed JPE programs(Alaska, Arizona, Colorado andUtah) found that:

• A very high percentage of judgesin all four states agreed that the eval-uations provided useful feedback ontheir performance;

• A significant majority of judgesin each state agreed that appropriatecriteria are used to evaluate theirperformance;

• Nearly all judges in each statefelt that evaluation commissioners

are fair;• Large percentages in each state

said that commission membersunderstand their role as judges;

• Majorities in each state agreedthat commission members under-stand the importance of judicialindependence; and

• Majorities of judges in each statesaid that the evaluation processmakes them appropriately account-able for their job performance.11

Similarly, judges who took part ina 2001 pilot study in WashingtonState had “predominantly positive”comments about the experience,and reported that the informationthey received was useful, had notbeen previously available, and couldnot have been transmitted through ameans other than anonymous sur-veys.12 Federal judges who partici-pated in a 1991 pilot program alsoassessed the value of judicial per-formance evaluation as “overwhelm-ingly positive.”13 As one federal judgeput it, “this project is extremelyworthwhile to me. Although I wouldfeel distressed if the responses werecritical and unfavorable, I still wantto know.”14

Judges have occasionally opposedJPE programs in the abstract, basedon fears that evaluations lie in thehands of unreliable, and potentiallypartisan, evaluation commissions.15

But the risk of commission bias canbe reduced through appropriatesafeguards, such as partisan balance,requiring formal training for com-mission members, and formal

approval of commission members bythe legislature or another body.Moreover, in reality judicial electionshave been politicized far more fre-quently when the public is not able torely on performance reviews as asource of information. In the last 20years, the most notorious examplesof campaigns to remove judges fromthe bench—Rose Bird, JosephGrodin and Cruz Reynoso in Califor-nia, David Lanphier in Nebraska,and Penny White in Tennessee—occurred in states where official, for-mal evaluations of each judge’sperformance were not available tovoters at the time of the election. Asa result, one or two controversialissues became the focus of the cam-paign. It is telling that some of thestrongest advocates of JPE are indi-viduals whose time on the bench wascut short or threatened by segmentsof the public demanding particularoutcomes in individual cases.16

Variation and innovationsAs noted above, JPE programs havevaried somewhat in their implementa-tion. Several western states haveadopted quite comprehensive JPEprograms, which feature predeter-mined standards for judicial perform-ance, thorough collection ofinformation, and widespread dissemi-nation of results. Other states havebeen more limited in their data col-lection, or have debated whether, andthe extent to which, evaluation resultsshould be kept confidential. In addi-tion, different jurisdictions havesought information from differentsources. These program variationsreveal a number of innovations to theevaluation process, which other juris-dictions may want to consider.

The first innovation is broadeningthe pool of survey participantsbeyond attorneys. For example, anincreasing number of states now sur-vey jurors on the judge’s clarity,demeanor, and level of preparednessat trial. New Mexico has broadenedthe pool even more significantly, pro-viding surveys for appellate courtjudges to court staff, other appellatejudges, trial court judges whose caseshave been appealed, the judge’s cur-

10. See, e.g., Editorial, The Judicial Survey, 155N.J.L.J. 748 (Feb. 15, 1999) (noting that “[t]hetradition of deference may serve to conceal thatinformation [on courtroom demeanor] from thevery person who needs it most, particularly if thejudge’s problem is a lack of audience-sense or ofthe ability to put himself in the shoes of anotherperson.”).

11. Supra n. 7, at xvii.12. American Judicature Society – Washington

Chapter, Committee on Judicial PerformanceStandards and Evaluation, FINAL REPORT OF THEWASHINGTON STATE JUDICIAL PERFORMANCE EVALUA-TION PILOT PROJECT 29 (2002).

13. Darlene R. Davis, JUDICIAL EVALUATION PILOTPROJECT OF THE JUDICIAL CONFERENCE COMMITTEEON THE JUDICIAL BRANCH 9 (Federal Judicial Cen-ter, 1991). The pilot project evaluated all districtcourt, magistrate, and bankruptcy court judges inone judicial district, the Central District of Illi-nois. The pilot program limited its informationcollection to anonymous attorney surveyresponses in five areas of judicial performance:

integrity, judicial temperament, legal ability, deci-siveness, and diligence. Id. at 5. Although partici-pation was voluntary, every judge in the districtchose to participate. See id. at 2.

14. Id. at 8.15. See generally, e.g., Jacqueline R. Griffin, Judg-

ing the Judges, 21 LITIGATION 5 (1995).16. See Penny J. White, Judging Judges: Securing

Judicial Independence by Use of Judicial PerformanceEvaluations, 29 FORDHAM URB. L.J. 1053, 1076(2002) (advocating for robust performance evalu-ations and noting that “Undoubtedly, much of thesuccess of those who seek to destroy judicial inde-pendence results from the lack of available infor-mation upon which to base one’s decision injudicial elections.”); Leonard Post, ABA Offers NewWay to Judge the Judges, NAT’L L.J., May 5, 2005, at 4(noting that Virginia’s JPE program was spurredby Justice Barbara Milano Keenan, who facedopposition to her reappointment by the state leg-islature in 2003 because of a dissent she wrote ina 1995 case involving the custody of a child by ahomosexual parent).

204 JUDICATURE Volume 90, Number 5 March-April 2007

Overview of official judicial perfomance evaluation programs

State or Public jurisdiction Participating judges/frequency dissemination?Alaska All judges/Prior to retention election Yes — Included in election pamphlet mailed to every voter;

detailed evaluations posted on website; evaluations printed innewspapers and aired on radio

Arizona All appellate judges; Superior Court judges in Yes — Pre-election reviews are mailed to voters and made Pima and Maricopa Counties/Every two years available at public centers such as libraries, banks and (mid-term and prior to retention election) grocery stores, and are posted on Arizona courts webpage.

Mid-term performance reviews are confidential.

Colorado All judges/Prior to retention election Yes – Blue Book of Ballot Issues (election information) sent to allvoters prior to election; also available on judicial branch websiteand published in newspapers

Connecticut New judicial nominees and incumbent judges Only evaluation criteria and procedural rules are made public.seeking reappointment/Upon seeking Judge may request that hearings concerning his reappointmentreappointment be open to the public.

D.C. Those seeking reappointment or senior status/ NoUpon seeking reappointment or senior status

Florida Voluntary, informal program; appears to vary No – evaluation forms go directly to judge with committeefrom circuit to circuit/No evaluations reviews

Hawaii All full-time judges/As retention and appointment Summary reports are disseminated; individual results are keptdecisions warrant confidential.

Idaho District magistrates only/After initial 18-month Noterm of office

Illinois Voluntary/N/A No – evaluation data is kept strictly confidential

Kansas* All judges/N/A Yes and no – for judges in retention elections, evaluationspublicly available; for judges running in contested elections,evaluations kept confidential

Massachusetts All judges/Judges with four years of experience Annual summary report available to bar members;are evaluated every 12-18 months; judges with no information provided on individual judgesmore than four years of experience are evaluated once every 18-36 months.

Minnesota Voluntary/Varies by judicial district Varies; some districts issue reports or summary information

New Hampshire All Superior Court and District Court judges Annual summary report for entire judiciary is presented to(appellate judges are evaluated collectively)/ Governor and other top state officialsEvery three years, with one-third of judges evaluated each calendar year

New Jersey All judges/Second and fifth year after appointment No – strictly confidential.

New Mexico All sitting judges except those running in a Yes – Retention evaluations are posted on commission’s website,partisan election/Midterm and prior to retention published in newspapers, and made available at county clerkelection offices. Midterm evaluations are confidential.

Puerto Rico N/A/Every 3 years N/A

Rhode Island All judges/Every 2 years No – sent to Chief Justice of Supreme Court and Chief Judge ofeach district court only.

Tennessee Appellate judges seeking retention/Every 8 Yes – final report of less than 600 words per judge is publishedyears, prior to retention election at least 180 days before qualifying deadline in general circulation daily newspaper

www.ajs.org JUDICATURE 205

rent and former law clerks, and lawprofessors (who are asked to evaluatethe clarity and accuracy of the judge’spublished opinions). Trial judge sur-veys in New Mexico are sent tolawyers and jurors, as well as courtstaff, law enforcement personnel,probation officers, psychologists, citi-zen review volunteers, social workers,interpreters, and court-appointedspecial advocates. Courts in Min-nesota have also reached out to anon-traditional resource, sendingdetailed questionnaires to litigants togauge their perception of how theywere treated and their overall satis-faction with the court process.

Another innovation is the use oftrained court observers in Alaska andNew York. In Alaska, independentcourt observers receive approxi-mately 40 hours of advance training,and are assigned to sit in on courtproceedings at unscheduled inter-vals. As many as 15 observers revieweach judge. They observe both crim-inal and civil matters, and reviewcourt proceedings ranging from jurytrials to motion hearings andarraignments. Observers give both

numerical ratings and written com-ments in response to straightforwardquestions about the judge’s behavior,such as “Did the judge pay closeattention to the testimony?” and“Did you understand the judge’sexplanations and decisions, or didyou leave feeling confused?” Allobserver data for each judge arecompiled into a one-page evaluationthat sets out the total number ofhours the judge was observed, thetypes of cases observed, and the aver-age rating the judge received in eachcategory. While New York does nothave an official JPE program, anindependent organization, the Fundfor Modern Courts, sponsors similarregular, public court monitoring(complete with detailed observerreports) throughout the state.

A third innovation is the develop-ment of benchmarks for judicial per-formance. Established benchmarksfor (among other things) a judge’scase management skills and perform-ance in survey responses provide aclear guide for judges and the publicas to expected standards for the judi-ciary, and give evaluation commis-sions a framework for assessing eachjudge’s performance. Benchmarksalso reduce the opportunity for mis-chief by an evaluation commissionthat might be inclined (for whateverreason) to recommend retention of asubpar judge or against retention ofan excellent one.

Utah has taken the lead in settingbright-line rules, instructing its evalu-ation commissions to base their rec-ommendations on the judge’s abilityto meet six predetermined bench-marks. These include: (1) a favorablerating by at least 70 percent of therespondents on at least 75 percent ofthe attorney survey questions; (2) fortrial judges, a favorable rating by atleast 70 percent of the respondentson at least 75 percent of the juror sur-vey questions; (3) compliance withrigid timing requirements for disposi-tion of cases; (4) at least 30 hours ofjudicial education per year; (5) sub-stantial compliance with the Code ofJudicial Conduct; and (6) physicaland mental fitness for office.17

A fourth development involvesmentoring of evaluated judges. Ari-zona has established three-member“conference teams” for each evalu-ated judge, consisting of anotherjudge, a member of the state bar, anda member of the public. The confer-ence team meets with the judge toformulate a written self-improvementplan based upon the judge’s self-eval-uation, public comments, and surveyresults. Conference teams work sepa-rately from evaluation commissions,and are prohibited by rule from “par-ticipat[ing] in formulating any find-ing as to whether a judge or justicemeets judicial performance stan-dards.”18 Justices on the New Hamp-shire Supreme Court also engage in

Overview of official judicial perfomance evaluation programs

State or Public jurisdiction Participating judges/frequency dissemination?Utah All judges/Every 2 years Yes – published in voter information pamphlet and posted on

governor’s website.

Vermont Judges seeking retention/Prior to retention Report for each judge seeking retention presented to the elections General Assembly for consideration

Virginia All judges/Three times per term No – first two evaluations of each term are confidential; third sentonly to relevant members of state legislature

Note: This chart reflects official judicial performance evaluation programs only. State and/or local bars conduct independent judicial evaluations in Georgia, Illinois, Kentucky, Maine,Missouri, Nebraska, Ohio, Pennsylvania, South Carolina, Texas, Washington, West Virginia, and Wyoming. In Nevada, performance evaluations are conducted by a newspaper, theLas Vegas Review-Journal.

*The Kansas program is brand new and has not had the opportunity to conduct any evaluations.

17. See, e.g., UTAH VOTER INFORMATIONPAMPHLET (General Election Nov. 5, 2002) at 60(on file with authors), available athttp://elections.utah.gov/GOV_election_pamphleWEB.pdf.

18. R. P. JUD. PERF. REV. ARIZ. 6(f)(2), quoted inA. John Pelander, Judicial Performance Review inArizona: Goals, Practical Effects and Concerns, 30ARIZ. ST. L.J. 643, 690-93 (1998) (alteration inPelander).

206 JUDICATURE Volume 90, Number 5 March-April 2007

peer review, completing individualself-evaluations and then collectivelydiscussing them to identify personalstrengths and weaknesses.

A fifth innovation is the develop-ment of formal processes to appealthe commission’s evaluation and rec-ommendation. In Colorado, an eval-uated judge is permitted to reviewthe commission’s draft profile andrecommendation before it is madeavailable to the public; if the judgedisagrees with the evaluation, he orshe may request an additional inter-view with the commission or, whereretention is not recommended,attach a statement of his or her ownposition to the profile when it is sentto the public. Similarly, in Arizonathe judge is permitted to review theevaluative report and submit com-ments before the report is dissemi-nated to the public.

A sixth innovation is partnershipbetween the state judiciary and stateand local bar associations to developJPE programs.19 Bar polls are fre-quently used to evaluate judicial per-formance, but alone they cannotaccount for important measuressuch as case management statisticsand courtroom observation. Further-more, bar polls have been criticizedby some as unscientific or haphaz-ard. By working directly with thejudiciary to convert existing bar pollsinto more comprehensive programs,bar associations can build trust withthe judiciary and provide a more use-ful product for the public.

A final development is the growthof creative efforts to disseminateresults, including the increasinglysophisticated use of performancecommission websites. One obstacleto comprehensive programs is thecost of disseminating evaluationresults, particularly if evaluations areconducted frequently. Costs, how-ever, can be reduced significantly byposting evaluation results online,provided that the public is madeaware of the website and the infor-mation is easy to access.

Indeed, the value of JPE programsis tied directly to the extent to whichthe results are shared with the public.Providing the results of individual

judicial evaluations to the electorate(both directly and through the newsmedia) in a manner that is easilyunderstood builds trust and confi-dence in the judiciary by identifyingjudges with outstanding performanceand identifying those who needimprovement. The broad public dis-cussion of judicial performance stan-dards and results reinforces theexpectation that judicial accountabil-ity should be process-oriented ratherthan outcome-oriented, and increasesthe profile of the evaluation commis-sion, which encourages greater partic-ipation in the JPE process. Lack oftransparency, by contrast, tends topromote suspicion about the evalua-tion process (from both judges andthe public), causes the public tobecome disinterested or apatheticabout its judges, and invites the cre-ation of informal judicial rankingsand polls to fill the information gap.

JPE has yet to gain widespreadapplication, but the innovationsdescribed above are working in statesthat have adopted them. Whetherimproving an existing JPE program orstarting one from scratch, evaluationcommissions would benefit greatlyfrom frequent sharing of ideas withtheir peers in other jurisdictions.

Nationwide valueJudicial performance evaluation hasvalue regardless of how a statechooses its judges. JPE has been usedmost frequently, and in its mostrobust form, in Missouri Plan states.This is not surprising, since voters inretention elections are natural con-sumers of information on the per-formance of sitting judges. However,many of the benefits of JPE also trans-late to jurisdictions where judges arechosen through contested elections.Limited JPE programs in Washingtonand New York, among other states,strongly suggest that comprehensiveperformance evaluation and wide-spread public dissemination of evalu-ation results would have a positiveeffect on judicial elections, by inform-ing voters about the performance oftheir judges and judicial candidates,and reducing the need for voters torely on expensive and politically

charged campaign advertisements.The most significant obstacle to

the broad implementation of JPEprograms in contested election statesis the concern that a candidate whois not currently on the bench cannotbe evaluated in the same way as a sit-ting judge. This concern is well-taken, but it is not insurmountable.Even candidates who have not previ-ously held judicial office can be eval-uated on the skills they would expectto use on the bench. Judicial candi-dates are almost always attorneys orjudges on lower courts, and would beexpected to have skills and knowl-edge that are measurable in muchthe same way as the skills and knowl-edge of an incumbent judge. Forexample, an attorney could be evalu-ated on his or her disposition, timeli-ness, responsiveness, fairness innegotiating with opposing counsel,use of facts and appropriate sourcesof law in briefs, and the like. Sourcesof information on the attorney can-didate’s performance might include:

• Surveys of members of the bar,especially attorneys who have workedwith and against the candidate inrecent cases;

• Surveys of non-attorneys whohave interacted with the attorney incourtroom, mediation, or depositionsettings, including judges, media-tors, arbitrators, court staff, stenogra-phers, and perhaps jurors orwitnesses;

• Surveys and/or consultationwith sitting judges (allowing for par-tisan balance among the judges con-sulted if desired);

• Review of selected submissions tothe court, including a variety ofmotions and briefs; and

• Management of cases for whichthe candidate was the lead attorney,looking for compliance with courttime frames and other rules and the

19. See, e.g., Hawaii State Bar Association,Standing Committee on Judicial Administration,Report: Regarding a Judicial Evaluation Program, 3-DEC HAW. B.J. 9, 9 (1999) (describing efforts towork with the state judiciary to implement a JPEprogram); Press Release, Missouri Bar Associa-tion, Judicial Evaluations Available Online to thePublic (on file with author), available athttp://www.showmecourts.org (noting that forthe first time, the Missouri bar would be surveyingjurors as part of its bar poll for the 2006 election).

www.ajs.org JUDICATURE 207

number of times the candidaterequested extensions or continu-ances.

This form of evaluation, while notidentical to judicial evaluation, wouldprovide a reasonably fair and accuratebasis for comparison between thecandidates. More importantly, itwould frame the comparison in terms

of objective, process-directed criteriaexpected of any judge, helping votersto cast an informed ballot. The evalu-ation commission obviously shouldnot endorse a particular candidate ina contested election, but it can statewhether each candidate has met thepredetermined benchmarks to beconsidered qualified for office.

JPE programs can also be adaptedto jurisdictions where judges areappointed and do not have to facethe voters directly. For example,approximately half the federal judici-ary serves terms of office, and per-formance evaluations could assistwith reappointment decisions. Evenfor judges with life tenure and noexpectation of changing jobs, per-formance evaluation serves as anincentive to identify areas for self-improvement, and to confirmstrengths on the bench. For the pub-lic, regular and frequent dissemina-tion of evaluation results allowscitizens to observe growth in judicialperformance, enhances public trustand confidence, and reinforces theexpectation that the proper criteriafor judicial accountability focus onthe adjudicative process, not particu-lar case outcomes.

RecommendationsThe Institute’s report details a signifi-cant number of recommendations forimplementing a comprehensive, well-functioning JPE program. These

detailed recommendations fall withinsix general categories, and may besummarized as follows:

• Conduct evaluations regularly. Eachsitting judge should be evaluated on aregular schedule, at least twice duringeach term of office or, if there is noset term, at least once every threeyears. Regular evaluations help judges

improve more quickly, and help thepublic accept neutral measures ofjudicial performance more readily.

• Choose neutral criteria. Evaluationsshould emphasize apolitical metricsof judicial performance, and shouldbe based primarily on performanceagainst predetermined benchmarks.Where judges and the public under-stand and accept the goals and sub-stance of performance benchmarks,shared expectations about judicialperformance are apt to developmore easily.

• Cast a wide net for collecting informa-tion. An evaluation commissionshould gather a broad and deep set ofinformation on the judge’s perform-ance, seeking information that istimely and based on personal knowl-edge when applicable. Such informa-tion should include survey data,review of case management skills andwritten opinions, courtroom observa-tion, and information gained frominterviews with the judge. The com-mission should issue a report con-cerning each judge’s performancebased on the collected information.Evaluation criteria should be as com-prehensive as possible, and any reportor recommendation should representa thorough analysis of the judge’s per-formance.

• Create trustworthy evaluation com-missions. Each evaluation commis-sion should be independent andmore or less balanced between attor-

neys and non-attorneys and alongpartisan lines. Depending on the sizeof the commission, gender and geo-graphic balance may also be appro-priate. The less opportunity for biasin the commission, the more likelythe public will receive its evaluationspositively.

• Be open. The evaluation processshould be transparent both to thejudge being evaluated and to thepublic. Judges and citizens shouldknow exactly why the commissionmade the recommendation or evalu-ation it did. Those who do notunderstand the process are unlikelyto give it proper credence.

• Share the results. Evaluation resultsshould be widely disseminated to thepublic through voter guides, newspa-pers of general circulation, and theInternet. No matter how comprehen-sive the evaluation, shared expecta-tions cannot be developed if thepublic is unaware of, or unable toaccess, the results.

****Rarely does a process that has been inuse for three decades qualify as animportant “discovery,” but for themajority of state and federal courts,JPE is exactly that. It is an importantcomponent to balancing judicialaccountability and judicial independ-ence. It identifies the proper criteriaby which to review a judge, withoutinvading the province of judicial inde-pendence so critical to our democ-racy. And it serves as a valuableeducational tool both for judges andthe public they serve. For every courtsystem in the United States, judicialperformance evaluation is an ideawhose moment has come. g

REBECCA LOVE KOURLIS is executive director of the Institute forthe Advancement of the AmericanLegal System.

JORDAN M. SINGER is director of research at the Institutefor the Advancement of the AmericanLegal System. (jordan.singer@du.edu)

Even candidates who have notpreviously held judicial office canbe evaluated on the skills theywould expect to use on the bench.