Trials, particularly jury trials, once played a central role in the American legal system.1 No longer.2 While trial remains a theoretical possibility in every case, the reality is quite different. Trials occur rarely, typically only in the most intractable disputes.3 The pronounced disappearance of trials seems to have largely escaped the attention of Hollywood,4 the literary community,5 and the mainstream media,6 but this development is well known to judges,7 other court personnel,8 litigators,9 and academics.10 However, even those “in the know” often do not appreciate just how rare trials have become.
This article documents and quantifies the continuing disappearance of trials. It confirms that today a trial is very much the exception, rather than the rule, regardless of jurisdiction (federal or state), type of case (criminal or civil), type of trial (bench or jury), or type of claim (contract, tort, etc.).11 The article also outlines the principal reasons why trials have diminished and some of the resulting consequences. Although trials are not likely to disappear altogether, they seem virtually certain to remain on the endangered list for years, and perhaps forever, primarily because Supreme Court decisions have made case disposition by motion more likely, and parties in both civil and criminal cases are increasingly drawn to what they perceive to be readily available, less expensive, and more attractive alternatives.
Diminishing trials are not a new phenomenon. To the contrary, there has been a “century-long decline in the portion of cases terminated by trial” and a more than “twenty-five year decline in the absolute number of civil trials.12 When the Federal Rules of Civil Procedure were first promulgated in 1938, approximately 20 percent of all civil cases were resolved by trial.13 By 1962, trials still accounted for roughly 12 percent of all civil dispositions in federal court.14 But 40 years later, the rate of disposition by trial in civil cases had fallen to less than 2 percent, even as the total number of civil dispositions grew dramatically. Today, approximately 1 percent of all civil cases filed in federal court are resolved by trial — the jury trial disposition rate is approximately 0.7 percent, and the bench trial disposition rate is even lower.15 The absolute number of civil trials in district courts is lower today than it was at any time in the last 55 years (the period for which comprehensive data are available). It is not surprising, therefore, that astute commentators believe civil trials in federal court are “approaching extinction.”16
The pattern in federal criminal cases is similar, although not quite as pronounced. By the end of World War II, plea bargains or dismissals resolved approximately 80 percent of all criminal cases in the federal courts, and trials accounted for the remaining 20 percent.17 The percentage stayed roughly the same until the early 1980s. However, following passage of the federal sentencing guidelines in the mid-1980s, the percentage of criminal cases resolved by trial significantly declined. By 2000, it was just 6 percent, and by 2010, the rate of disposition by trial in criminal cases was less than 3 percent.18 Today, trials only occur in approximately 2 percent of federal criminal cases.19 As Judge William Young of the U.S. District Court for the District of Massachusetts explains: “Today, our federal criminal justice system is all about plea bargaining. Trials — and, thus, juries — are largely extraneous.”20 Of course, many more cases are filed, and ultimately resolved, in state courts.21 But, if anything, there is even less likelihood of a case proceeding to trial in state court than in federal court.22 One study found that by 2002, civil cases were resolved by juries in state court less than 1 percent of the time.23 The comparable number for criminal cases was 1.3 percent.24 The rates are even lower today.25 Primarily due to foreclosure cases growing out the mortgage crisis, bench trials remain common in a few state courts, but rarely occur in others.26
While it is “beyond dispute” that trials occur less and less frequently,27 accurately measuring the extent of the decline is not a simple task.28 The relevant data are neither uniform nor comprehensive, especially at the state court level. Definitional issues abound, starting with the most basic: What is a trial? The Administrative Office of the United States Courts, which compiles federal judicial statistics, defines a “trial” as “a contested proceeding where evidence is introduced,” thereby including all hearings on temporary restraining orders, hearings on preliminary injunctions, hearings on contested motions, and other contested proceedings in which evidence is introduced.29
This evidence-based formulation, which certainly conflicts with the popular understanding of the term,30 expands the number of trials exponentially because a single case can actually give rise to multiple trials.31 This only makes that much more striking the paucity of reported trials measured in this manner.32 Likewise, in many state court data compilations, a trial is deemed to occur whenever a jury is impaneled, regardless of whether the jury reaches a verdict.33 This practice also tends to inflate the number of trials that are reported. In all events, data on trials need to be reviewed with care, qualified as appropriate, and interpreted carefully. We make every attempt to do so here.
Between 1938 and 2009, “there was a decline in the percentage of civil cases going to trial of over 90%.”34 There has also been an enormous decline in the absolute number of trials.35 In 1962, there were 50,320 total civil case dispositions — 5,802, approximately 12 percent, were by trial.36 There were also 33,110 criminal defendants whose cases were resolved that year — 5,097, approximately 15 percent, were by trial.37 The high watermark for civil trials occurred in 1985. That year, federal courts resolved 268,070 civil cases, more than five times the number disposed of in 1962, and there were 12,529 trials (6,276 bench trials and 6,253 jury trials).38 In criminal cases, 1990 was the most active year, when there were 7,874 trials (6,181 jury trials and 1,693 bench trials) in a universe consisting of 56,519 defendants.39 There has been a continuing decline in both the percentage and the absolute number of trials in more recent years.
There are a number of different metrics pertaining to trials in the federal district courts. One metric focuses on civil and criminal trials completed per year, where a trial includes any hearing or proceeding in which evidence is introduced. There were 20,581 such trials in federal district courts in 1997, 15,830 in 2005 and 11,754 in 2016.40 Thus, by this measure, trials diminished by approximately 43 percent in just under 20 years. But many of the trials were not case dispositive (i.e., they did not involve liability or guilt determinations). Jury trials constituting liability or guilt determinations show a much larger decrease over time. For example, there were 4,765 civil jury trials on the merits in 1990 and 1,758 such cases in 2016, representing a decrease of more than 60 percent.41 Similarly, there were 5,061 criminal jury trials resolving guilt or innocence in 1990 but only 1,889 such trials in 2016, a decrease of 63 percent.42 The statistics for the last 16 years are set forth in Appendix 1 (previous page). Even during this more abbreviated period, the number of civil jury trials on the issue of liability decreased by 48 percent,43 and the number of criminal jury trials on the issue of guilt decreased by 46 percent, such that the total “on the merits” jury trials in federal court decreased by 47 percent.44
Another federal court trial metric examines civil and criminal case termination based on the stage at which resolution occurs. For example, in 1990, a total of 213,429 civil cases were terminated, and 9,236 cases (4.3 percent) were ended “during or after trial” (4,783 by jury and 4,480 nonjury).45 In 2016, by contrast, 271,302 civil cases were terminated, but only 2,781 cases (1 percent) ended “during or after trial” (1,965 jury and 816 nonjury).46 Measured this way, the absolute decrease in trials was 70 percent. Similar reductions occurred in criminal cases.47 Still another measure of trial activity is the total number of juries picked to serve in civil and criminal cases in the federal district courts. That number has declined, markedly and quite steadily, as shown in Appendix 2 (at left). In 1996, a total of 10,338 juries were selected, but in 2016 the total was just 3,887. Thus, during a 20-year period, the number of juries picked in federal courts decreased by 63 percent.48
Given this decrease in trials, it is hardly surprising that the number of trials per district court judge also diminished significantly over time. In 1962, there were, on average, 21 merits trials in civil cases (10 jury/11 bench) per district court judgeship each year.49 By 1985, the number of merits trials per district court judgeship increased to 24 (12 jury/12 bench).50 But, thereafter, the number of merits trials per district court judgeship began to decline rapidly such that, by 2006, there were half as many merits trials per judgeship as in 1962.52 By 2015, the average number of merits trials per district court judgeship per year was just 4 (3 jury and 1 nonjury).51 This is not to suggest that district court judges were working less over the years. To the contrary, as discussed infra, available evidence indicates that the average hours worked by a district court judge increased substantially during this time period. The statistics plainly show the judicial time was not spent trying cases.
As trials per judge diminished, so did the amount of time judges spent on the bench.52 In 1980, the mean total hours on the bench per active district court judgeship was 790. In 2013, the mean total hours on the bench per district court judgeship was 430, less than two hours per day, a reduction of 46 percent.53 There were some district courts averaging fewer than 200 courtroom hours per judgeship per year, less than one hour per day.54 The mean time on the bench has continued to decrease, even as the overall workload in the district courts has continued to increase.55 Once again, what has changed is the nature of the work performed by the courts, as more cases are decided by motion and as judges spend more time managing existing caseloads.56
The nature of the civil cases being tried has also changed. Professors Marc Galanter and Angela Frozena report that in 1962, more than 50 percent of all trials involved tort claims, and 20 percent involved contract claims. By 2016, less than 12 percent of the trials involved tort claims, and contract claims comprised 18 percent. Conversely, civil rights cases accounted for less than 1 percent of the trials in 1968, but they constituted more than 30 percent of the trials in 2010, and 27 percent of the trials in 2016. Prisoners’ petitions have also increased significantly over time as a percentage of trials and now represent a material share of all trials.57
With 28 approved judgeships, the Southern District of New York (SDNY) is one of the two largest districts in the federal judiciary. Historically, it has been one of the busiest courts and has been known for handling many important and high-profile civil and criminal cases.58 While not as large, the Eastern District of New York (EDNY), with 15 approved judgeships, enjoys a similar, well-earned reputation. Both courts benefit from a hard-working group of senior judges (there are 15 in the SDNY and 14 in the EDNY) who add experience, depth, and capacity to each district.59 Given their size and stature, these districts serve as a helpful case study in observing the national trend of decreasing trials at a local level.
Consistent with the national data, there has been a significant decrease in the number of trials completed in both the EDNY and the SDNY as measured using the evidence-based standard (see Appendix 3, next page).60 Once again, many of these “trials” are not liability or guilt determinations. As depicted in Appendix 4 (available online at http://judicialstudies.duke.edu/judicature), jury trials that do reflect liability or guilt outcomes show a much more pronounced decrease over time. Thus, the data capturing completed trial activity on the merits in the EDNY and SDNY are consistent with national trends.61 For example, in 1999, there were 224 jury trials completed in the EDNY (130 civil and 94 criminal) and 292 jury trials completed on the merits in the SDNY (181 civil and 111 criminal). Slightly more than 15 years later, in 2015, there were just 98 jury trials completed in the EDNY (57 civil and 41 criminal) and 131 jury trials completed in the SDNY (76 civil and 55 criminal),62 representing a decline of 56 percent in the EDNY and 55 percent in the SDNY from 1999 to 2015. In 2016, it was much the same story — 90 jury trials were completed in the EDNY, and 129 jury trials were completed in the SDNY.
The number of juries selected in the EDNY and the SDNY also confirms the diminishing trials occurring in these venues. In 1998, there were 232 juries selected in the EDNY (15.5 per judgeship) and 375 juries selected in the SDNY (13.4 per judgeship). By 2005, the figures were 184 (12.3 per judgeship) and 234 (8.4 per judgeship), respectively.63 The data for more recent years, shown in Appendix 5 below, indicate a continuing decline: In 2015, there were 127 juries selected in the EDNY (8.5 per judgeship) and 152 in the SDNY (5.4 per judgeship). In 2016, the numbers were down again: 123 juries (8.2 per judgeship) selected in the EDNY and 140 (five per judgeship) in the SDNY. Note, however, that the per-judge calculations exclude the trial contributions made by senior judges. Assuming (conservatively) that senior judges try cases at the minimum level of activity to which they must commit (i.e., one quarter of a normal schedule), the number of jury trials per active judge diminishes even more given the number of senior judges in each district.64 Judicial vacancies could also impact the calculations. While the number of trials continues to decline65, the EDNY judges and the SDNY judges averaged 706 and 645 hours on the bench each year, respectively.66 These are among the highest totals in the country.67 Indeed, the district court judges in the Second Circuit have more on-bench time than the district courts in any other circuit by a large margin.68
The data concerning state court trial activity are neither as comprehensive nor as current and consistent as the federal court data. The leading resource for state court data is the National Center for State Courts. The state court database — which only goes back a few years, unlike the federal data — shows that in 2015, 21 states, representing a significant portion of the country, reported data on total civil dispositions in courts of general jurisdiction, including the number of jury trials and bench trials.69
Pennsylvania reported the highest civil jury trial disposition rate at 0.53 percent. Other large states reported low civil jury trial rates, including California (0.21 percent), Texas (0.47 percent), Florida (0.18 percent), and New Jersey (0.12 percent). These outcomes were consistent with results from prior years. In 2014, Alabama reported that 1.53 percent of civil dispositions were resolved in jury trials , and all other reporting states had lower civil jury trial rates. In 2013, 20 states reported civil disposition data in courts of general jurisdiction. No state had a civil jury trial rate higher than 0.62 percent (Nebraska), and fully 18 states reported jury trial rates of equal to or less than 0.55 percent, including Pennsylvania (0.55 percent), Texas (0.47 percent), Michigan (0.47 percent), Ohio (0.38 percent), Florida (0.20 percent), and California (0.18 percent). In 2012, 20 states reported civil disposition data. Only New York, separately discussed below, had a civil jury trial rate greater than 1 percent.70 Civil bench trials are also declining, except in those states still experiencing the effects of the mortgage crisis. For example, in 2015, in states such as Pennsylvania (1.5 percent), New Jersey (0.95 percent), and Connecticut (0.36 percent ), the civil bench trial rates were low; but in states such as Florida (10.05 percent), California (11.92 percent), and Texas (14.34 percent) the rates were much higher, reflecting significant real estate foreclosure activity. In criminal cases, the jury trial rates for four large states (California, Florida, Pennsylvania, and Texas) during the last four years are set forth in Appendix 6 (at right). The jury trial disposition rate was less than 2 percent in all four states and less than 1 percent in California. Trials in felony cases typically occur more frequently than in cases involving misdemeanors, which is consistent with research suggesting that cases involving larger potential penalties are less likely to be resolved through plea bargaining.71
In 2012, the most recent year for which comprehensive civil case statistics are available, the general jurisdiction courts of New York disposed of 177,457 civil cases.72 Slightly more than 2,000 of these cases were tried to a jury, representing 1.15 percent of all civil dispositions. There were also 1,430 bench trials, representing 0.8 percent of all civil dispositions. Although tort cases accounted for 36 percent of the total civil dispositions, they represented 84 percent of the cases tried to a jury (1,724 out of 2,048). Still, the jury trial rate for tort cases as a group was slightly less than 3 percent. Medical malpractice suits had the highest jury trial rate (5.9 percent), 277 jury trials out of 4,160 total dispositions.73 The data for criminal cases in New York are both more current and complete than the data for civil cases in New York. But, once again, as depicted in Appendix 7 (next page), a picture of limited trial activity emerges with the jury trial disposition rate hovering around 3 percent or less. Consistent with other states, a slightly higher trial rate occurs in felony cases. Unlike most states, where criminal trials outnumber civil trials, there are significantly more civil trials than criminal trials in New York. In both civil and criminal cases, the trial rates in New York remain meaningfully higher than in the surrounding states of Connecticut and New Jersey, as set forth in Appendix 8 (available online at http://judicialstudies.duke.edu/judicature). This differential arises in part by differences in case mix between New York and Connecticut and New Jersey.
There are many reasons why trials have declined to the extent described here. We outline some of those reasons below, drawing heavily on prior work done by a number of leading academics and members of the judiciary.
With respect to civil trials, Yale Law School Professor John Langbein contends that litigants “no longer go to trial because they no longer need to.”74 He believes, as do others,75 that modern procedural rules have played a crucial role in rendering trials “obsolete” such that “pretrial civil procedure has become nontrial civil procedure.”76
Langbein believes that the expansive discovery provisions in the Federal Rules, and the emphasis on judicial case management and settlement contained in Rule 16 and elsewhere “have had the effect of displacing trial in most cases, causing ever more cases to be resolved in the pretrial process, either by settlement or by pretrial adjudication.”77 Indeed, “precisely because discovery allows such far-reaching disclosure of the strengths and weaknesses of each side’s case, discovery often has the effect of facilitating settlement.”78 This last conclusion is consistent with research demonstrating that predictability of outcome and cost have a great deal to do with whether a civil case settles.79 Settlement is much more likely if both sides have similar expectations concerning the probability of liability and the amount of damages.80 To the extent that expansive discovery leads to greater knowledge and predictability, it will normally facilitate settlement, particularly when judges have been directed by rule to encourage such outcomes. Decisions rendered by the Supreme Court interpreting key provisions of the Federal Rules of Civil Procedure have almost universally tended to militate against the growth of trials. Many commentators identify 1986 as a pivotal year, because that is when the Supreme Court issued three decisions signaling that disposition by summary judgment should be more available than in the past.81 The message was clearly heard by the lower courts, as dispositions by summary judgment increased significantly thereafter.82
Today, a case is much more likely to be disposed of by summary judgment than by trial.83 “Because the very purpose of summary judgment is to avoid unnecessary trials, one need not be a trained logician to conclude that an increase in the availability of summary judgment will naturally have a corresponding negative impact on the number of trials.”84 The Court’s summary judgment decisions were followed, several decades later, by multiple decisions that seemed to raise the pleading standards needed to state a viable civil claim, thus signaling greater receptivity to motions to dismiss.85 That message was also heard — dispositions by motions to dismiss are now much more likely than dispositions by trial.86 Some argue that the Supreme Court’s decisions assigning courts a more active gatekeeper role regarding experts have also tended to reduce trials.87 Moreover, expansive discovery under the Federal Rules became expensive discovery, especially following the advent of email and other electronic documents, which also motivates parties to settle rather than try cases, particularly in commercial disputes.88 Expansive discovery often leads to delays that also increase the likelihood of settlement by reducing the stakes (and hence value) of a case.89
In recent years, many consumer contracts have been drafted in ways that seek to eliminate the right to trial or class treatment.90 Once again, the Supreme Court has been quite receptive to these efforts.91 Even if that were not so, there can be little doubt that the small size of many disputes, and the prohibitive costs of hiring a lawyer to handle those disputes, have contributed to the decline of trials. The number of pro se litigants has increased dramatically in recent years. As Judge Jed Rakoff of the SDNY observes, many litigants cannot afford a lawyer.92 And pro se litigants are among the least likely to proceed to trial, presumably because they lack an understanding of how to try a case.93 The availability of alternative dispute resolution, and its embrace by the courts, also has played a very significant role in the diminishing trial phenomenon.94 ADR gained traction just as the judiciary was being encouraged, by rule and otherwise, to become more involved in managing cases and controlling dockets. Inevitably, “managing cases” seems to translate into pushing settlements or other non-litigated case resolutions.95
Also relevant to the disappearance of trials is the longstanding distrust of juries, particularly among corporate defendants.96 Vanderbilt Law Professor Amanda Rose has suggested that in complex trials “the size of the potential liability, combined with the uncertainty of outcome, make trial simply too great a risk from the viewpoint of corporate defendants and those that advise them.”97 Indeed, of the more than 4,300 securities class actions filed since 1995, only 15 (less than three-tenths of one percent) have been tried to judgment.98
The prospect of punitive damages also leads defendants, especially corporations, to avoid trials.99 The outcome in Pennzoil v. Texaco made very clear to management and shareholders the dangers of “rolling the dice” in a large case with a punitive damages claim. In Pennzoil, a Texas state court jury returned a verdict that was “twenty-five times greater than [trial counsel’s] likely ‘worst case scenario’ and five times greater than [counsel’s] estimated ‘complete runaway.’”100 The need to bond the judgment forced the bankruptcy of one of the largest oil companies in the world and the loss of billions of dollars in shareholder value, even though the judgment was ultimately settled for a fraction of the original amount. But the lessons of that case, particularly for inside and outside counsel, will not be soon forgotten.
It is also important to recall that the completion of a trial does not necessarily conclude a case. Furthermore, cases that go to trial are much more likely to be appealed than those disposed of by other means.101
So, if finality is important to a party, trial is not the preferred way to proceed.
There is a “self-perpetuating” and “self-fulfilling” feature to the disappearing trial.102 University of Wisconsin Law Professor Marc Galanter describes this phenomenon: “As lawyers who ascend into decision-making positions have less trial experience, the discomfort and risk of trials looms large in their decisions. Judges, too, accumulate less experience and, in many cases, less appetite for trials.”103 Thus, in a real sense, the more trials disappear, the less likely it is that they will reappear in the future104 absent some sort of systemic change.105
While there are certainly improvements that can be made to the trial process, thereby making it marginally more attractive,106 those changes seem unlikely to alter the prevailing landscape in a meaningful way. While changing the Rules of Civil Procedure might reverse the trend towards dispositions by motion, it would not deter those otherwise preferring to settle rather than proceed to trial.
Many of the reasons trials have diminished in civil cases apply equally to criminal cases (e.g., cost considerations, risk elimination). But one cause unique to criminal cases is the introduction of sentencing guidelines and the rise of prosecutorial discretion. In the federal courts, these guidelines explicitly offer an incentive to avoid trial in the form of an offense level reduction for “acceptance of responsibility.” Moreover, busy prosecutors are often willing, as part of the plea negotiating process, to agree to guideline levels lower than would otherwise be the case. Judge Young reports that sentences following convictions at trial are five times larger than sentences received by those who plead guilty pursuant to a cooperation agreement with the government.107 This disparity strongly motivates many criminal defendants to “cop a plea.”108
Some have suggested that the “disappearance of the American trial presents a major crisis for the legal profession.”109 But this assessment seems overstated. To be sure, the continuing disappearance of trials impacts those who actually try cases, aspire to do so, or provide trial support. However, that impact does not affect a significant percentage of the profession (and has not for quite some time, given the long-term trends).110 While the lack of trials also poses questions about courthouse resource allocations, those sorts of issues can be addressed with greater planning.111
And, it is important to remember that while trials may be diminishing, the right to trial is not being eliminated, nor could it be, under the Sixth and Seventh amendments to the Constitution. The option of trial remains available to those who wish to make use of it (and who have not otherwise bargained it away, contractually or as part of a plea deal). Increasingly, though, litigants are choosing not to pursue a trial for many of the reasons mentioned above. Some have suggested that trials, especially jury trials, perform a valuable role in a democracy, allowing for direct citizen participation and decision making.112 For example, Northwestern Law Professor Robert Burns argues that the reduction in trials reduces “the space for effective speech,” eliminates a source of public information, and abandons “an important vehicle for citizen self-governance.”113 Even accepting all that, does anyone seriously believe that parties who do not want to go to trial should be forced to do so in order to promote greater democracy? And, while there may be some citizens who will be genuinely disappointed if they do not get a chance to serve on a jury, they may well be a distinct minority.114
Another weighty concern is how the disappearance of trials impacts the development of the law itself. There are several aspects to this issue. First, there is the danger that law developed only through motions “will be arid, divorced from the full factual content that has in the past given our law life and the capacity to grow.”115 Of course, conceding that the law may develop differently does not necessarily mean it will be less worthy in a qualitative sense. And once again, can the remedy really be to compel parties to pursue a path they otherwise are not inclined to follow? Second, the diminishing number of trials will no doubt produce less law relating to the types of issues that arise at trial. This, in turn, may lead to greater uncertainty about trial outcomes and substantive law.116 To the risk-averse who walk corporate hallways, such a development might promote more, rather than fewer, settlements. Third, the lack of trials also means there are fewer actual verdicts to serve as markers or data points for valuing claims.117 But, once again, this is not a new development. Cases increasingly have been settled based on a “settlement market,” rather than on actual verdicts. Whether such an approach produces optimal resolutions is a different (and far more complicated) question.118 Ultimately, parties decide whether, and to what extent, trials are worth pursuing compared to other types of disposition. In almost all cases, disputes are now resolved without resorting to trial because that is what the parties prefer based on economic and other considerations. And that seems unlikely to change any time soon.119