U.S. Department of Justice Office of Justice Programs Bureau of Justice Statistics ******************************************************** This file is text only without graphics and many of the tables. A Zip archive of the tables in this report in spreadsheet format (.csv) and the full report including tables and graphics in .pdf format are available on BJS website at: http://bjs.gov/index.cfm?ty=pbdetail&iid=5368 ********************************************************* Bulletin Criminal Appeals in State Courts Nicole L. Waters, Ph.D., and Anne Gallegos, M.P.A., National Center for State Courts James Green, M.A., and Martha Rozsi, Westat In 2010, an estimated 69,348 criminal appeals were resolved in the 143 appellate courts with criminal jurisdiction in the United States. Nearly two-thirds (63%) of appeals were reviewed on the merits of the case, and a majority (81%) of these appeals upheld or affirmed the trial court decision (figure 1). Overall, more than half (52%) of all appeals resulted in an affirmation of the trial court decision. Appellate decisions were only considered affirmed when the appellate court upheld the entire trial court decision. In 12% of appeals, appellate courts reversed, remanded, or modified a component of the trial court decision. More than a third (37%) of appeals were not reviewed on the merits. Appellate courts with discretionary jurisdiction can decide not to grant a petition to review a case or to dismiss the appeal (30% of appeals filed). The appellate court can dismiss a case or deny a review of a case on its merits if the court does not have jurisdiction over a case, if a procedural error prevents the review of a case (e.g., untimely filing of documents), or if the notice of appeal does not allege any reviewable errors. An estimated 7% of appeals were withdrawn by the petitioner or appellant, and less than 1% were transferred or certified to another appellate court. A transfer or certification to the court of last resort (COLR) occurs when an intermediate appellate court (IAC) appeal is reviewed by the COLR, which addresses errors or outcomes with important questions of law or to secure uniformity in case law. ********************************************************** ************** HIGHLIGHTS ************** * In 2010, an estimated 69,348 criminal appeals were resolved in the 143 state appellate courts with criminal jurisdiction. *In 2010, 38 states had both intermediate appellate courts and courts of last resort, while 12 states had courts of last resort only. * Of the cases reviewed on the merits, 81% were affirmed. More than half (52%) of all appeals, the appellate court upheld the trial court decision. * About 2% of all intermediate appellate court appeals were subsequently reviewed by the court of last resort. * The most common legal issue appellate courts addressed was sufficiency of evidence (15%), followed by excessive or inconsistent sentencing (10%). * Reversal rates were highest for aggravating or mitigating factors that might affect the severity of the offense (20%). * In 2010, appellate courts reversed 23% of appeals for drug trafficking, 23% for larceny or theft, 21% for sex offenses, and 17% for murder. ********************************************************** State appellate courts were created to detect and correct errors in trial court decisions and provide fair, consistent, and timely resolutions to all appeals. This report presents data on state court criminal appeals, resolution of appeals, and time to resolution. Data are from the Bureau of Justice Statistics’ (BJS) 2010 Survey of State Court Criminal Appeals (SSCCA). The outcomes were analyzed by type of court, level of court review, and appellate court structure. SSCCA data were collected from a nationally representative sample of all criminal appeals disposed in all 143 state appellate courts in 2010. This report uses those data to provide a national picture of appeals disposed in 2010, including rates of reversal and length of time to appeal resolution. ******************************** State appellate court structure ******************************** The appellate court structure used to process and resolve appeals varies considerably from state to state. In response to the needs of the judiciary and state-specific characteristics, some states created a two-tiered structure, adding IACs to accommodate high workloads. In 2010, 38 states had IACs and COLRs, and 12 states had COLRs only (map 1). In states with a two-tiered system, some IACs were established with geographic divisions to increase access and distribute workload. To selectively reduce caseloads, some states have established strategies (such as discretionary jurisdiction) and implemented rules to assign some cases to panel review (a smaller subset of judges) rather than an en banc review (all judges of a court). See Terms and definitions for more information. Establishing a two-tiered structure has also affected the roles and responsibilities of appellate courts. Of the 38 states with IACs in 2010, 27 used a two-tiered system of appellate review that includes one or more IACs with primarily mandatory review and a COLR with primarily discretionary review. In such systems, the IAC is mainly responsible for correcting potential errors of the trial court and the COLR is responsible for the interpretation and development of case law. However, for IACs that review appeals by permission (discretionary review) as opposed to by right (mandatory review), the IAC was responsible for the definition and development of the law, in addition to its responsibility to correct errors. Understanding the organizational structure of the appellate court is necessary for comparing appellate cases across jurisdictions and drawing accurate conclusions about key case and processing characteristics. The nation’s state judiciary has seven basic appellate court structures. ********************************************************** ******************************** Appellate court structures ******************************** COLR mandatory--This structure is typically found in states with comparatively small volumes of appeals. There is only one appellate court, a court of last resort (COLR), and the court hears appeals by right. COLR discretionary--This structure is typically found in states with relatively small volumes of appeals. There is only one appellate court, a COLR, and the court hears appeals by permission. Deflective structure--In states with this structure, appeals are filed and may be fully briefed and submitted with the COLR, which then retains select appeals and transfers others to the intermediate appellate court (IAC). COLR discretionary/IAC mandatory—This is the most common appellate court structure. The COLR hears appeals mostly by permission, and the IAC hears appeals mostly by right. COLR and IAC discretionary--In states with this structure, both the COLR and IAC hear appeals mostly by permission. IAC by subject matter--States with this structure have more than one IAC that is distinguished by subject matter. For example, one IAC is specialized for criminal appeals and the other IAC reviews civil appeals. COLR by subject matter--States with this structure have more than one COLR that is distinguished by subject matter. For example, one COLR is specialized for criminal appeals and the other COLR reviews civil appeals. ********************************************************** Courts of last resort dismissed more appeals than intermediate appellate courts ********************************************* Examining how appeals are resolved by various case characteristics provides additional insight into the national picture of criminal appeals. COLRs primarily interpret and develop case law rather than correct errors in individual cases. In 2010, COLRs primarily reviewed appeals by permission (82%), while IACs primarily reviewed appeals by right (93%). COLRs dismissed more appeals (76%) than IACs (13%), in part due to courts’ discretionary jurisdiction (table 1). Overall reversal rate for intermediate appellate courts was higher than for courts of last resort ************************************************* In 2010, the overall reversal rate for COLRs was 7% of all appeals. Petitions requesting a COLR to review a case more often came from the defendant (94%) than the state (4%). When the state petitioned the COLR, the court more often reviewed the appeal on the merits (49%) than when the defendant petitioned the COLR (18%). Also, the state won a reversal of the trial court decision more often (40%) than the defendant (5%). Death penalty appeals had a higher reversal rate (19%) than both felony (8%) and misdemeanor (8%) appeals. Resolutions from an IAC are different than those from a COLR, primarily because one purpose of the IAC is to relieve the workload of the COLR and serve as a first review of potential trial court error. IACs reviewed appeals on the merits in 80% of appeals (table 2). The overall reversal rate for IACs was 14% of all appeals. IAC reversal rates were higher for state-initiated appeals (38%) than for defendant-initiated appeals (13%). However, IACs were nearly as likely to review on the merits a state-initiated appeal (75%) as a defendant-initiated appeal (80%). Ten death penalty appeals were heard by IACs, all of which were reviewed on the merits. Of these 10 appeals, 3 were reversed. The reversal rate for nondeath penalty appeals was 15%. ***Footnote 1 in alabama and tennessee, the IAC has jurisdication over cases in which the death penalty has been imposed.*** About 2% of all intermediate appellate court appeals were subsequently reviewed by the court of last resort *********************************************************** After a review by the IAC, appellants may seek subsequent review by the COLR. An estimated 29% of IAC appellants sought additional review by the state’s COLR. Of those, the COLR granted additional review for 6%, or 853 appeals. Nearly three-quarters (73%) of those accepted for subsequent review by the COLR arose out of a decision by the IAC to affirm the trial court decision (figure 2). Twenty-three percent sought COLR review following a reversal of the trial court decision and 4% following a dismissal by the IAC. A reversal may have resulted in only a partial reversal and may therefore be appealed. Nearly all of the appeals in which the COLR granted subsequent review were initiated by the defendant (93%) and were reviewed on the merits (96%) at the IAC. For appeals reviewed on the merits, nearly all opinions addressed six or fewer legal issues ******************************************************** On appeal, the appellants or petitioners present appellate courts with alleged trial court errors through the briefing process and, if applicable, during oral arguments. In response, appellate courts prepare a dispositive opinion when the appeal is reviewed on the merits. In that opinion, the appellate court will address the legal issues presented and, at times, address additional and relevant legal issues. For appeals reviewed on the merits, a majority (97%) of the opinions addressed six or fewer legal issues. The most common legal issue the appellate courts addressed was sufficiency of evidence (15%), or whether the prosecution met the burden of providing facts sufficient to satisfy the burden of proof for each element of the crime charged. The second most common legal issue was excessive or inconsistent sentencing (10%), which includes consecutive or concurrent sentencing decisions, mandatory minimums, and calculations of time served. The third most commonly addressed legal issue was suppression of evidence (8%) related to statements, identification, coerced confessions, illegal search and seizure, and the denial of other pretrial motions requesting the inclusion or exclusion of evidence from trial. When the appellate court addressed aggravating or mitigating factors that might affect the severity of the offense, the court found reversible error in 20% of the appeals (figure 3). The reversal rate for excessive or inconsistent sentences was 17%, and the reversal rate for guilty plea challenges was 13%. The court was less likely to reverse legal issues on relevancy or prejudicial evidence (3%) or ineffective counsel (4%). Reversal rates of convictions varied by the most serious offense in underlying trial case ******************************************************** Appeals are classified by convictions of the most serious offense in the underlying trial case. In 2010, murder convictions were most often appealed and resulted in a reversal rate of 17%. Appellate courts reversed 23% of appeals for drug trafficking, 23% for larceny or theft, 21% for sex offenses, 17% for drug possession, 16% for burglary, and 16% for aggravated and simple assault (figure 4). The top 10 most serious offenses made up 82% of known offense types for all cases brought on appeal. The median time for an appellate court to resolve an appeal was approximately 10 months ****************************************************** In 2010, the median time for appellate courts to resolve an appeal was 297 days (figure 5). Appeals heard by permission underwent an additional step in the appellate process—petitioning the court to grant or deny review. The median time for the court to grant or deny a review was 92 days. Once an appeal was initiated, the median time for completing the record preparation was 69 days. The next step was for the petitioner or appellant to brief and fully submit the appeal; the median time for this step was 81 days. The final step, and the longest, was for the appellate court to issue a decision. The median time for the court to issue a decision after the appeal was fully briefed and submitted (or after oral argument, if held) was 107 days. Case processing times for cases disposed in 2010 should be used with caution when attempting to estimate the case processing times of either cases disposed in other years or cases filed in a particular year. Case processing times are affected by many factors, including existing court rules and guiding legislation, which differ across time periods and vary over the lifespan of a case. Court types, court structures, and proportion of caseloads undergoing reviews on the merits all contributed to time variations *********************************************************** When examining the total duration of cases, time on appeal varied depending on whether or not the court reviewed the case on the merits. Excluding death penalty appeals, the median time from the start of an appeal to the final resolution was more than 1 year (421 days) for appeals reviewed on the merits and about 3.5 months (104 days) for those without a review on the merits (figure 6). Time on appeal also varied by type of court (i.e., COLR or IAC). For COLR appeals reviewed on the merits, 50% were resolved in about 1 year (360 days). For IACs, 50% of the appeals reviewed on the merits were resolved within 1.2 years (426 days). Differences in the appellate court structures in 2010 defined how courts processed appeals. In Iowa, under a deflective appellate court structure, briefing was consolidated at the COLR before the appeal was assigned to the IAC. In Virginia, both the IAC and the COLR had discretionary jurisdiction over the majority of their caseloads, and fewer appeals were granted review. The most common appellate structure in 2010 occurred when the IAC had mostly mandatory review of its caseload and the COLR had mostly discretionary review. The median time--—or the time needed to resolve half (50%) of the courts’ appeals—did not vary considerably when comparing appellate court structures (figure 7). However, at the 75th and 95th percentiles, courts with mostly discretionary review resolved most appeals in a shorter time than courts with mandatory jurisdiction. In states with mandatory review at the COLR and without an IAC, a small percentage of appeals (5%) remained unresolved until nearly 3 years (1,071 days) after the appeal began. Additionally, court structures with larger percentages of caseloads resolved on the merits took longer to resolve appeals. Courts without an IAC and a mandatory review by the COLR reviewed more than three-quarters (78%) of its caseload on the merits. Courts with both an IAC and a COLR with discretionary review granted fewer cases overall, and a lower percentage (25%) of its caseload was reviewed on the merits (table 3). A total of 134 death penalty appeals were resolved in 2010 ****************************************** Appellate courts are expected to review appeals proportionally with regard to the complexities of the case. Death penalty appeals merit relatively more scrutiny and more time than nondeath penalty appeals. In 2010, a total of 22 courts resolved 134 appeals of death penalty cases. Another 14 appellate courts had jurisdiction over death penalty appeals, but did not dispose of any such appeals in 2010. While the majority of death penalty appeals were heard by COLRs, Tennessee and Alabama had specialized criminal jurisdiction IACs that reviewed these appeals. Florida’s COLR resolved the most death penalty appeals (23 appeals), followed by California’s COLR (20 appeals). Connecticut, Utah, and Virginia COLRs each resolved 1 death penalty appeal in 2010 (table 4). Median time to resolve death penalty appeals was three times that of nondeath penalty appeals ********************************************************* Case processing time in a year should be interpreted with caution when attempting to estimate the case processing times of either cases disposed in other years or cases filed in a particular year. Because death penalty cases take substantially longer than other types of cases, death penalty cases that are disposed of in a year are initiated across a longer time and therefore have a greater potential of being processed under different rules than other cases. As such, case processing time for death penalty cases should be interpreted with caution. Death penalty appeals received greater scrutiny by the state’s appellate courts than nondeath penalty cases and had different case processing rules than those for nondeath penalty appeals. For example, many states automatically provide the notice of appeal for death penalty cases to the appellate court through the trial court clerk instead of requiring the appellant to file a notice of appeal. States may also give the petitioner or appellant additional time to file documents for death penalty appeals. Florida allowed 60 days to file transcripts for a nondeath penalty case, but allowed 110 days for a death penalty case. Oklahoma allowed 90 days after judgment for transcripts to be filed in a nondeath penalty appeal, compared to 6 months after sentence for a death penalty appeal. Half (50%) of death penalty appeals were resolved in about 2.4 years (888 days) and three-quarters (75%) were resolved in 3.4 years (1,224 days). In 5% of appeals, the appellate court resolved the appeal in more than 13.9 years or 5,058 days (figure 8). The duration of death penalty appeals (13.9 years) was much longer than that of nondeath penalty appeals (2.3 years for misdemeanors and 2.4 years for nondeath penalty felonies) at the 95th percentile. Methodology ************** Data in this report are from the Bureau of Justice Statistics’ (BJS) 2010 Survey of State Court Criminal Appeals (SSCCA). The National Center for State Courts, in collaboration with Westat, conducted the study. Criminal appeals disposed in 2010 were sampled from all 143 appellate courts in the United States that have criminal jurisdiction, which included 51 courts of last resort (COLR) and 92 intermediate appellate courts (IACs) (appendix table 1). The final sample of 5,045 appeals was composed of-- * a nationally representative probability sample of all criminal appeals in 2010 * a complete census of all death penalty appeals in 2010 *a supplemental oversample of COLR appeals resolved in 2010 identified in the national sample that were also reviewed by the corresponding IAC. The first component was a sample that included 4,539 criminal, nondeath penalty appeals disposed in 2010. The second component was a full census (134 appeals) of all death penalty appeals in the nation. Combined, these data included 4,673 appeals from 51 COLRs and 90 IACs, which were the main study data for this report. The SSCCA also included a supplemental oversample of 372 appeals that were heard by both the IAC and COLR. While there was no court-level nonresponse, two of the IACs did not provide any data for the SSCCA. North Dakota’s COLR had the authority to transfer cases to an IAC but had not done so in 2010. North Dakota’s IAC was authorized as a temporary court of appeals that was activated when the COLR reached a certain threshold of appeals during the previous year. The threshold was not reached in 2009; therefore, its temporary IAC was not active in 2010. Two cases were sampled in one of Illinois’ IACs (appellate court, fifth district), but were excluded because they did not meet the definitional criteria for inclusion in the SSCCA (i.e., they were study ineligible). Because these appeals were ineligible, appeal-level replacement was not appropriate and was not used. Data represent appeals to the appellate court by right (usually called an appeal and initiated by a notice of appeal) or by permission (usually called a petition and initiated by a petition for leave to appeal or a petition for certiorari). Inclusion criteria were applied to all appeals. The appeals met two basic criteria: (1) they arose from a final judgment or verdict or sentence in a criminal case, and (2) they were disposed by the appellate court in 2010. Appeals were generally resolved in calendar year 2010. Exceptions include courts that used a court term or fiscal year calendar to draw the sample or courts that used closure dates other than the date of disposition (e.g., Idaho’s sample was drawn from the “closed date,” which is the date the court issued a remittitur). Regardless of the dates used, all courts represent a full 12 months of dispositions. Alternatively, an appeal was ineligible if it involved defendants tried as juveniles, was post-conviction in nature, occurred pretrial, was interlocutory, contested a probation revocation judgment, or contested a civil commitment. Determining the eligibility of an appeal often required a full review of the issues on appeal. For example, postconviction issues (e.g., an application for habeas corpus, making a claim of ineffective counsel, or making a claim that the death penalty is unconstitutional) filed after an initial direct appeal had been decided by the court were excluded. Writs, such as prohibition and mandamus (with the exception of writs of certiorari, if the court used such a writ to initiate an appeal by permission), were also excluded. Interlocutory appeals are appeals filed in response to a trial court’s decision on a motion and do not result in a final judgment at the trial court. Appeals of these interim decisions were also excluded. Finally, violations of court orders were excluded when such violations were not new criminal charges (e.g., probation violations and violations of civil commitment orders). An example of a violation that was retained in the sample was failure to register as a sex offender. This violation was included because it is a crime to not register as a sex offender, making the appeal of the conviction or sentence comparable to other direct appeals. Sampling ************** Each appellate court compiled a universe list of appeals using the inclusion criteria described previously. The data obtained for the 2010 SSCCA come from a stratified, single-stage probability sample of 5,045 cases derived from the universe list of appeals. Appeals were stratified by type of case or type of court (i.e., death penalty, COLR, and IAC), and appeals were sampled at a fixed overall rate within each stratum. Initial sampling rates were calculated for each stratum based on the ratios of the target sample size desired to the expected total population for each stratum. Appendix table 2 shows the target sample sizes and expected total population assumed for each stratum and the resulting initial sampling rates. A census was taken for death penalty cases. After sampling cases in about half (50%) of the courts, the number of cases actually collected (i.e., the actual yield) was evaluated. The actual yield was lower than expected, which indicated smaller population sizes. Based on this evaluation, the sampling rates for the remaining half of courts were increased. This increase in rates not only adjusted for the actual yields, it also took into account the design effect and therefore the reduction in effective sample sizes introduced by changing from the initial rates within each stratum. This was done using a mathematical program that solved for the new set of rates, taking into account the current actual yield, the expected total yield, design effects, and effective sample sizes required. Appendix table 3 shows the initial sampling rates, revised sampling rates, and total actual sample sizes. In most instances, courts were able to provide clean, study- eligible lists of cases to which the rates were applied. However, some lists provided by appellate courts were known to be overinclusive (i.e., the list included study-ineligible cases). If a list was overinclusive, a two-part approach was used. First, the rate of eligible cases was estimated by reviewing an initial sample and a partial sample. Second, a sampling rate that took into account the eligible case rate was used for the remaining sample. Data collection and preparation ********************************** All appeals were coded directly from court records, including the appellate court docket, the initial appellant or petitioner brief, and the court’s decision or opinion. Court records were accessed through various means, such as electronic court records, public information access through the court’s website, private law libraries, and hard copies (requested through the mail or acquired from in-person visits to appellate court clerks’ offices). Based on a review of these documents, the data collection team coded each appeal. Each appeal represented one record, except when the actions that occurred within an appeal prompted the completion of an additional data record. Circumstances that resulted in additional records include-- * Consolidated cases—Appeals may be consolidated at various points during case processing. If two or more cases were consolidated after briefing (i.e., after the parties had filed briefs in the individual cases), then each case was coded separately to capture all of the issues presented to the court. Cases consolidated before briefing were coded together because all issues were presented in one brief. *Cross-appeals—If either party filed a cross-appeal in a case, the cross-appeal was coded separately to capture all of the issues presented to the court. *More than one appeal disposed in 2010—If a sampled appeal accompanied an additional appeal from the same underlying trial court case that was disposed in 2010, each was coded separately. For example, an IAC may have disposed of a case in early 2010 by remanding it to the trial court for additional review. The trial court decision was made and the case was appealed to the IAC again, with a disposition on the new appeal also made in 2010. If either of those decisions was selected in the sample, both of the cases were coded. Another example is an appeal that was initially dismissed (e.g., for procedural error), then later re-filed and accepted for review. If both the dismissal and the disposition on review occurred in 2010, then both cases were coded, even if only one of the cases was on the original sample list. To determine the legal issues most commonly addressed by the appellate courts, these data required a review of all dispositive opinions, whether published or not. Data captured the content of the first six legal issues addressed by the appellate courts, and also captured a count of all legal issues presented in the opinion. Quality assurance ******************** The coding team performed a multi-level quality check. Each appeal was coded, reviewed, and then entered into a database, with each task performed by a different member of the team. The appeals were entered into a data system designed to check for logic errors during data entry. As a final stage of quality control, the project team cleaned the data by court to uncover inconsistent case processing practices within each court. During the cleaning stage, data were sorted by coder and date coded. Early coding was most likely to contain errors. Therefore, the first third of appeals coded by each team member was selected and reviewed for errors. Weighting, estimation, and variance estimation *********************************************** Because the SSCCA estimates were based on data obtained from a probability sample, each sample appeal required the assignment of a weight to provide unbiased estimates with measurable precision. A base weight was calculated for each appeal as the inverse of the overall probability of selection, which reflected stratum (i.e., death penalty, COLR, and IAC) and whether the case was sampled before or after the rate change. It also determined the effects of the two-step process for sampling from overinclusive lists. A nonresponse adjustment factor was not required because the one unavailable case was handled through substitution. Poststratification or raking was inappropriate, as no corresponding control totals were available. The sums of the resulting weights are the best available estimates of the size of the population of inference, either overall or by subgroup. The distribution of weights was reviewed overall and by the relevant factors (e.g., stratum and before or after rate change), and preliminary weighted estimates were produced as part of the review. The resulting full sample weights were used for the estimates presented. For variance estimation, a set of 100 balanced repeated replication (BRR) replicate weights was produced and used for the variance estimates, standard errors, and confidence intervals. BRR was chosen because it performs better than the jackknife for quantiles such as medians, which are often used with these data given the number of dates in the court record. In creating the BRR replicate weights, each major analytic domain (i.e., death penalty selected with certainty, COLR, and IAC) was spread across the 100 replicates, such that they received the appropriate maximum number of degrees of freedom. Replicate group assignment was carried out through sorting by court and selection order within court. The latter maximized the degrees of freedom for more detailed estimates (e.g., state and court structure). All of the estimates were calculated using WesVar along with the corresponding standard error. Numerous methods were available for estimating percentiles, and statistical software packages (e.g., SAS or SPSS) may use different methods (optional or default), which may generate different estimates. Standard error computations **************************** Probability sampling implies some imprecision in the estimates because the entire population is not enumerated. This imprecision is measurable and is reported or reflected in standard errors. These measures reflect and convey the imprecision in probability sample estimates and allow for inference to the total population, comparisons between subgroups, or comparisons across time that acknowledge this imprecision. Variances, standard errors, and confidence intervals depend on the estimate involved, the sample design, the sample size, and the subgroup sample size (if appropriate). The ratio of the standard error for an estimate to the estimate itself is a useful way to compare the relative precision of estimates across estimates. In this report, the coefficient of variation (CV) was also calculated for all estimates, representing the ratio of the standard error to the estimate. If the CV was greater than 50% or the estimate was based on 10 or fewer sample cases, the estimate is noted with a “!” symbol in the tables and figures, indicating the data should be interpreted with caution. Terms and definitions ************************** Direct appeal--occurs when the defendant appeals the final judgment of a trial court. The appeal can be taken from the final judgment of the defendant’s conviction, sentence, or both. Discretionary review/appeal by permission--an appeal that the court can choose to review. Appeal by permission cases must first be either granted or denied by the court. Only granted appeals receive further review. Dispositive opinion--resolves the appeal, resulting in a decision by the court. En banc review--all judges of the appellate court, or the entire bench, decide the appeal. Harmless errors--rulings that, while mistaken, do not require a reversal of the judgment by the appellate court or warrant a new trial. Mandatory review/appeal by right--an appeal that the court must review. Panel review--a review of an appeal before a set of judges. Panel sizes for appellate courts vary by state. Postconviction appeals--an application for habeas corpus, making a claim of ineffective counsel, or making a claim that the death penalty is unconstitutional. These appeals are filed after an initial, direct appeal has been decided by the court. Published opinions--historically available in print and held precedential authority. With advances in technology, the term published currently implies more than one meaning for appellate decisions. For the purposes of this report, all decisions and opinions were included, whether or not they held precedential authority. Reviewed on the merits--the appellate court has considered the arguments of the parties and made a decision based on the substantive legal issues raised on appeal. This occurs following oral arguments and/or is based on a review of the briefs. Reversible error--a finding by the appellate courts that the lower court incorrectly resolved the issue, and this error was substantial enough to potentially change the outcome of the trial. ******************************************************************* The Bureau of Justice Statistics of the U.S. Department of Justice is the principal federal agency responsible for measuring crime, criminal victimization, criminal offenders, victims of crime, correlates of crime, and the operation of criminal and civil justice systems at the federal, state, tribal, and local levels. BJS collects, analyzes, and disseminates reliable and valid statistics on crime and justice systems in the United States, supports improvements to state and local criminal justice information systems, and participates with national and international organizations to develop and recommend national standards for justice statistics. William J. Sabol is director. This report was written by Nicole L. Waters, Ph.D., and Anne Gallegos, M.P.A., National Center for State Courts, and James Green, M.A., and Martha Rozsi, Westat. Tracey Kyckelhahn verified the report. Irene Cooperman, Morgan Young, and Jill Thomas edited the report, and Barbara Quinn and Tina Dorsey produced the report. September 2015, NCJ 248874 ******************************************************************* ************************************************* Office of Justice Programs Innovation * Partnerships * Safer Neighborhoods www.ojp.usdoj.gov ************************************************* *************************** 8/27/2015 3:50pm JER ***************************