The Trail and the Bench: Elections and Their Effect on Opinion Writing in the North Carolina Court of Appeals. Adam Chase Parker

Transcription

1 The Trail and the Bench: Elections and Their Effect on Opinion Writing in the North Carolina Court of Appeals By Adam Chase Parker A paper submitted to the faculty of The University of North Carolina at Chapel Hill In partial fulfillment of the requirements for the degree Master of Public Administration Spring 2010 This paper represents work done by a UNC-Chapel Hill Master of Public Administration student. It is not a formal report of the School of Government, nor is it the work of School of Government faculty. Executive Summary This study examines opinion writing in the North Carolina Court of Appeals from 2005 to 2008 to determine whether judicial productivity is affected in election years. The study contains numerous findings, including that judges running for re-election to the court produce fewer opinions than judges not facing re-election. The majority of this decrease is found in published opinion writing (opinions that may be cited as precedent), while unpublished opinions (opinions concerning settled law) do not show a significant decrease. Significant differences were not seen in page lengths, dissents, case withdrawals, or case dismissals. Clearance rates fell for the entire court in election years and non-election ( safe ) judges opinion writing outputs remained consistent between election years and non-election years. Times from docket to opinion and oral argument to opinion showed an annual decrease, election year or not. In the conclusion, potential options for policymakers are discussed to address these productivity effects.

2 Introduction The establishment of the North Carolina Court of Appeals in 1965 was the most significant modification to North Carolina s appellate court system since the Constitution of The Court began hearing appeals in 1967 with six judges, and since expanded to fifteen. Cases are heard and decided in panels of three, unlike North Carolina s Supreme Court (which hears cases before the entire court). In the past decade, reform primarily concerned North Carolina s judicial selection process. Restrictions on judicial candidates speech changed 2 in response to Republican Party of Minnesota v. White. 3 Campaign financing and voter education were central to the Judicial Campaign Reform Act of 2002, which included public financing of appellate judicial campaigns, publication of a voter guide profiling candidates, and campaign contribution limits of $1,000. This legislation resulted in America s first publicly funded judicial election program. 4 These reforms and numerous others concentrate on what candidates may say and spend, hoping to mitigate the effect of election pressures on judicial impartiality. A consideration not as often debated is whether election cycles affect court productivity. North Carolina Supreme Court Justice James Exum noted this widely held assumption over twenty years ago. As it was, the 1986 Fall Term of the North Carolina Supreme Court was severely disrupted. Incumbent justices were forced to spend almost as much time campaigning as they did doing the business of the Court. 5 Appellate judges campaign statewide while tending to their court workloads. If Justice Exum s statement were correct, court productivity in election years versus non-election years would be markedly different. This exploratory study aims to determine whether judicial productivity is quantitatively affected in election years. The study uses four years of data from the North Carolina Court of Appeals ( ) and compares key judicial performance indicators across groups of judges, across years, and within each year. After analysis, the study offers suggestions to address election pressures on court performance. Court Caseloads and Opinion Writing The National Center for State Courts operates the Court Statistics Project, which provides court productivity measures for every state. The Court Statistics Project offers the number of cases per judge, which indicates that North Carolina is providing its intermediate appellate court a reasonable caseload compared to other states. In fiscal year , North Carolina Court of Appeals judges heard 166 cases per judge (including appealed cases and petitions). This value compared favorably with the national median, 161 cases per judge. 6 The National Center for State Courts provides suggested performance measures for courts, including clearance rate, time to disposition, and age of active pending caseload. Clearance rate is defined as the number of outgoing cases as a percentage of the number of incoming cases. 7 Time to disposition is the percentage of cases resolved within established timeframes. Age of active pending caseload shows the age of current cases from filing until the time of measurement. 8 Further measures are provided, which consider juror use, cost per case, employee satisfaction, case file integrity, trial date certainty, and access and fairness considerations. The Court Statistics Project does not consider opinion writing in detail. Opinions are the primary output of appellate courts. Opinions are time-consuming, multi-faceted court decisions penned by judges to decide a case. Opinions typically require oral argument, where judges explore issues with the parties before opinions are written. Opinions in North Carolina can be published or unpublished, where published opinions establish binding precedent and unpublished opinions apply existing law to the appealed case (but cannot usually be applied as precedent). The type of opinion, published or unpublished, may indicate opinion difficulty. Further, dissents in the North Carolina Court of Appeals automatically engender review by the North Carolina Supreme Court, but represent additional work for Court of Appeals judges. 1

3 Court statistics also do not control for whether a judge is facing external pressures that could affect work performance, such as elections. This study begins filling these gaps in court statistics study through analysis and comparison of judges opinion writing activity in election years and non-election years. Methodology and Findings Data Source This study used a database of docket sheets obtained from the Information Technology Department of the North Carolina Supreme Court and Court of Appeals. This database contained 8,313 entries of Court of Appeals docket sheets from 2004 to Between 2005 and 2008, 6,611 cases were docketed, 5,965 opinions were written, 430 cases were withdrawn, and 554 cases were dismissed (several cases remained unresolved from the 2008 docket). Each docket sheet provided a docket number, docket date, case title, opinion date, opinion author, dissents, concurrences, the number of pages in the opinion, oral argument date, whether the opinion was published or unpublished, withdrawn or dismissed date, and whether the case is considered complete. Not every variable is present in each docket sheet (for example, not every opinion has a concurrence or dissent). The data was sorted and analyzed to examine trends. Data Groupings Candidate and Safe s: Groupings were created to compare the productivity of judges seeking election (hereafter candidate" judges) and those who were not (hereafter safe judges) within the fouryear period of In 2006, two judges sought re-election to the North Carolina Court of Appeals, and three judges sought election to the North Carolina Supreme Court. Assuming election pressures of a Supreme Court seat are comparable, these three judges were considered candidates. This brought the total number of candidates in 2006 to five. In 2008, five judges ran for re-election to the Court of Appeals, and no other offices were sought and 2008 were combined to create ten candidates. Safe judges were totaled at eight judges in 2005, ten in 2006, seven in 2007, and ten in Candidates in Non-Election Years: To compare candidates activity to the same judges activity in nonelection years, judges were flagged as candidates in non-election years. In 2005, seven judges who ran for election at one point in the four-year study period sat on the court. In 2007, eight judges who ran for election in the four-year study period sat on the court. Election and Non-Election Years: Two years (2006 and 2008) of docket information were used to create the election years category and 2007 were used for non-election years. Units of Analysis: To standardize comparisons, average outputs were used. For example, the population of candidates in election years is ten (five judges in 2006 and five in 2008), and in election years, they wrote 279 published opinions. This creates an average of 27.9 published opinions per judge in election years. Additional statistical significance tests were performed on page lengths, docket to opinion days, and oral argument to opinion days. Court Panels: Each case is decided in a panel of three judges, although each case panel was not provided in the data set. Additionally, cases without an opinion were not identified with a particular judge, meaning withdrawn and dismissed cases could only be analyzed at the entire court level. Entire Court Comparisons (Full tables, graphs provided in Appendix A). Clearance rates are reduced in election years, but trend upward over time. Clearance rates were calculated for the entire court only in this study (each case isn t assigned to a single judge). Courts often identify 100 percent clearance rates as a goal (or the ratio of dispositions to new cases is equal). In election years, the court s clearance rate is 6.24% less than non-election years. A dip in the clearance rate occurs from July to September in all years. This decrease coincides with a court recess 2

4 from the second week in June to the first week in August. The table below provides a visual representation of the clearance rate decrease in election years as the year progresses. Entire Court Clearance Rate The court writes fewer opinions in election years, while page lengths appear similar. The court wrote 4.33 fewer opinions per judge in election years than in non-election years (or 4.26 percent fewer opinions). Election year opinions are slightly longer (.21 more pages, P-value of.08, t-value of 1.36). Unpublished opinions change little, while published opinions decrease. The gap in opinions per judge is seen mostly in the number of published opinions. In election years, judges wrote 4.24 fewer published opinions than in non-election years (or 11.2 percent less), but only.37 fewer unpublished opinions in election years than non-election years. Dissents increase slightly, while withdrawals and dismissals stay consistent. Dissents were greater in non-election years, with 1.36 more dissents per judge. Withdrawn cases were greater in election years per judge (1.13 more in election years), while dismissals were greater in non-election years than in election years (1.86 more per judge). Evidence does not prove that opinions are delivered faster in election years than non-election years. Cases in non-election years take longer from docket to opinion (21.66 days longer than election years, or 7.4 percent longer, P-value of 4.22E-23, t-value of 9.9), while time from oral argument to opinion was remarkably similar (.06 more days, P-value of.97, t-value of.034). However, docket to opinion time also decreased each consecutive year, suggesting a different cause than election cycles. Total docketed cases varied slightly across years. In 2005, 1,654 cases were docketed, 2006 saw 1,735 cases, 2007 showed 1,642 cases, and 2008 had 1,580 docketed cases. Candidates (Full tables, graphs provided in Appendix B) Candidates write fewer opinions in election years than safe judges. Candidates wrote, on average, 11.7 fewer opinions than safe judges in election years (or 11.5 percent fewer opinions) and 7.6 fewer than the entire court average (7.8 percent fewer opinions). Candidates wrote 7.03 fewer opinions than the same judges wrote in non-election years (7.26 percent fewer opinions). Published opinions show the greatest decrease in the candidates group. The major difference is seen in published opinions. Candidate judges wrote 9.7 fewer opinions on average than safe judges in election years (25.79 percent fewer published opinions). Candidates wrote 7.23 fewer published opinions per judge than the same judges wrote in non-election years (20.6 percent fewer). Docket to opinion and argument to opinion show the same tendencies as the entire court. In cases written by candidates, docket to opinion days fell by an average of in election years compared to the same group in non-election years (P-value of 5.99E-19, t-value of -8.98). Candidates opinions were completed an average of days faster than the entire court average and days faster 3

5 than the safe judges in election years. A decrease of 2.42 days occurred in the time from oral argument to opinion (P-value of.324, t-value of -.986). Candidates wrote fewer opinions as the May primary and November general elections drew closer. As the following graph shows, candidates wrote fewer opinions as the May primary and November general election draw closer. Candidates non-election year opinions stayed relatively constant, with increases in April and August. Average Opinions Per Month Candidate s (10 total in , 15 total in ) Safe s (Full tables, graphs provided in Appendix C) As a group, there was a less pronounced difference in the performance of safe judges between election years and non-election years. Safe judges wrote 4.0 more opinions in election years than non-election years. Their opinions were.23 pages longer in non-election years (P-value of.62, t-value of.48), and they wrote.4 fewer dissents in non-election years. Docket to opinion length shows a similar decrease as other groupings (22.05 more days in non-election years, P-value of 7.76E-14, t- value of -7.5). Oral argument to opinion took 3.93 days longer in election years than non-election years (P-value of.32, t-value of -.981). Safe judges write 8.7 more opinions than candidates in non-election years. This may indicate a population difference between candidate judges and safe judges, spillover of election pressures into non-election years, or another unanticipated factor. Safe judges do not exhibit the May primary or November general election drop-off. As the graph below illustrates, safe judges showed no sign of a May primary decrease, nor as severe a decrease in November when compared to non-election years. Average Opinions Per Month Safe s (20 total in , 30 total in ) Limitations This study is limited to a quantitative assessment of productivity, which is only one element of a successful court. Other potential considerations of the North Carolina Court of Appeals (complexity of cases, etc.) are not addressed and are important measures of productivity. Qualitative study should be 4

6 conducted (including possible judicial performance evaluations conducted by the North Carolina Bar Association) 9. Informal interviews with Court of Appeals judges suggested safe judges tend to aid candidate peers by helping with special panels and other items that do not appear in productivity measures. Qualitative study may bring these additional work processes into focus. The study also does not account for the quality of staff counsel, clerks, and research assistants who aid in case disposition. Conclusion This study shows productivity decreases in election years. In election years, candidate judges write fewer opinions than safe judges and their opinion writing decreases occur around significant election events. Candidate judges opinion writing decreased sharply around March and April, precisely when judges campaign hardest to qualify for public financing. Other indicators show a minute difference (page lengths, dissents), or no clear relationship (length of time between docket/oral argument and opinion). This study cannot explain a gap between candidate judge and safe judge opinion writing in non-election years (a difference of eight opinions, on average). Whether the shown difference is significant enough to warrant further judicial reforms is a decision for North Carolina lawmakers. All study years are after the 2002 Judicial Reform Act in North Carolina. In the years before this reform, disposition rates trended below filing rates. 10 This trend has reversed, with a string of successive years with above 100 percent clearance rates. There is also a trend indicating opinions are being completed faster each successive year. These trends may be attributable to in-house reforms enacted since the current Chief assumed office in This time decrease occurs before oral arguments, which suggests briefs and court procedures are completed faster than in years past. If policymakers choose to further alter judicial selection, judicial appointments will likely be discussed. The 1989 North Carolina Judicial Selection Study Commission argued that judges should be appointed by the Governor and later reconfirmed by the General Assembly. 11 The study posited numerous reasons for an appointive system, including how election pressures compel judges to avoid legally correct but politically unpopular decisions. 12 The study commission offered unsuccessful legislation to move to appointments, which required a constitutional amendment. Numerous public surveys in North Carolina have found a move away from elected judges is unpopular 13, even as voters profess ignorance about the judges they select. 14 A lack of public support and the requirement of popular vote approval make a move to appointment difficult to accomplish. These pressures could be addressed in other ways, including the use of retired judges or mediation. Currently, the Court of Appeals employs four retired judges (also called emergency judges). These judges are used sparingly (for example when a judge is appointed to a higher court and a vacancy exists). Further use of these judges to alleviate election judges caseloads may reduce the impact of elections on Court of Appeals judges. Further use of mediation might alleviate caseloads at the Court of Appeals as well. Mediation is a form of alternative dispute resolution that brings parties together to craft a settlement. Mediation is already performed at the Court of Appeals, 15 and does not require judicial opinions to settle disputes. Mediation may reduce opinion-writing demands of judges, especially unpublished decisions regarding settled law. Further use of mediation would likely require additional resources. Qualitative study of judicial productivity in the North Carolina Court of Appeals is needed. This study shows a quantitative difference in the number of opinions in election years versus non-election years, but does not consider other intangible measures of court productivity. Judicial Performance Evaluations may provide a window into this qualitative component. In-depth interviews with judges, court officials, and other parties involved in the appeals process would further inform how the courts business is affected. Textual analysis of opinions written by election judges may yield valuable insights as well. Regardless of method, qualitative contributions to the study of appellate court productivity would be a welcome addition and help inform policymakers should they pursue further court reforms. 5

7 1 Britt, David. History of the Court of Appeals of North Carolina July 1967-October (1992). p. 2 2 See Order Adopting Amendments to the North Carolina Code of Judicial Conduct, (Apr. 2, 2003), available at [hereinafter 2003 Judicial Code of Amendments] 3 Matthew Eisley, Code Loosesns Grip on s, News and Observer (Raleigh, NC), September 20, 2003, at 1B. 4 Rogers, Philip, Stenberg, Carl, and Waterman, Sarah. Voter Owned Elections in North Carolina: Public Financing of Campaigns. Popular Government. Vol. 74, No. 2, Winter 2009 (p. 36). Available at 5 Chief Justice James G. Exum, Jr. to the Judicial Selection Study Committee, 15 April Judicial Selection and Retention in North Carolina. (page 6). 6 National Center for State Courts. Court Statistics Project: Appellate Courts, 2007 report. Accessed Jan. 18, National Center for State Courts. Courtools: Measure 2: Clearance Rates. Accessed Jan 18, (p.1) 8 National Center for State Courts. Courtools: The Ten Core Measures. Feb. 3, (p.1) 9 JPE Recommendations On Schedule for Accessed Jan 18, North Carolina Courts: 2000 Annual Report. Accessed Jan. 24, (p. 12) 11 JUDICIAL SELECTION STUDY COMMISSION: : Report to the Governor, Chief Justice, Attorney General and the 1989 General Assembly of North Carolina -- Judicial Selection Study Commission (p. 4-5) 12 JUDICIAL SELECTION STUDY COMMISSION: : Report to the Governor, Chief Justice, Attorney General and the 1989 General Assembly of North Carolina -- Judicial Selection Study Commission (p. 5) 13 Heagarty, J. Christopher. Public Opinion and an Elected Judiciary: New Avenues for Reform. 39, Williamette Law Review (2003) (public opinion and judicial selection) 14 Heagarty, J. Christopher. Public Opinion and an Elected Judiciary: New Avenues for Reform. 39, Williamette Law Review (2003) (public opinion and judicial selection) 15 North Carolina Court System: Mediation. Accessed Jan. 24,

8 Appendix A: Entire Court

9 Clearance Rate Opinions per Average Opinion Page Length Published per Unpublished per #Dissents per judge #Withdrawn per judge #Dismissed per judge Average Time of Docket to Opinion Average Time Oral Argument to Opinion Entire Court Year(s) % % % % Election Years (06+08) Non-Election Years (05+07) Difference (05+07) net (06+08) Non-Elec. net Election P-Value (Two-Tail Test) T-Value (Two-Tail Test) % % % E

10 Entire Court (# Opinions/Month) Election Non-Election 50 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Election Non-Election

11 Entire Court Clearance Rate 160% 150% 140% 130% 120% 110% 100% All Years Election Year Non-Election Year 90% 80% 70% 60% Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec All Years 87.90% 93.78% 94.60% % % % % 97.74% 82.40% % % % Election Year % % 83.06% % % 98.14% % 84.47% 70.88% 98.95% % % Non-Election Year 69.35% 80.72% % % 88.31% % % % 93.77% % % %

12 Appendix B: Candidate s

13 Opinions per Average Opinion Page Length Published per Unpublished per #Dissents per judge Average Time of Docket to Opinion Average Time Oral Argument to Opinion s Facing Re-Election or Seeking Supreme Court Seat (Candidates) Year(s) 2005 (7) (5) (8) (5) Non- Election Years (05+07) (15) Election Years (06+08) Difference (05+07) net (06+08) Non-Elec. net Election P-Value (Two-Tail Test) T-Value (Two-Tail Test) E

14 Candidate s (# Opinions/Month) Per Election Year Non-Election Year Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Election Year Non-Election Year (Adjusted for # of s)

15 Appendix C: Safe s

16 Opinions per Average Opinion Page Length Published per Unpublished per #Dissents per judge Average Time of Docket to Opinion Average Time Oral Argument to Opinion s Not Facing Re-Election or Seeking Supreme Court Seats (Safe s) Year(s) Election Years (06+08) Non-Election Years (05+07) Difference (05+07) net (06+08) Non-Elec. net Election P-Value (Two-Tail Test) T-Value (Two-Tail Test) E

17 Safe s (# Opinions/Month) Election Year Non-Election Year Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Election Year Non-Election Year (Adjusted for # of s)

18 Appendix D: Comparison Tables

19 Opinions per Average Opinion Page Length Published per Unpublished per #Dissents per judge Average Time of Docket to Opinion Average Time Oral Argument to Opinion Comparing Groups in Election Years ( ) Group Entire Court Candidate s Difference (Entire Court Candidate) Entire Court Safe s Difference (Entire Court Safe) Safe s Candidate s Difference (Safe net Candidate) P-Value (Two Tail Test) E E-08 T-Value (Two Tail Test)

20 Opinions per Average Opinion Page Length Published per Unpublished per #Dissents per judge Average Time of Docket to Opinion Average Time Oral Argument to Opinion Comparing Groups in Non-Election Years ( ) Group Entire Court Candidate s Difference (Entire Court Candidate) Safe s Candidate s Difference (Safe - Candidate) P-Value (Two-Tail Test) E E-08 T-Value (Two-Tail Test) Entire Court Safe s Difference (Entire Court Safe)