Judicial Procedural Involvement (JPI): A Metric for Judges’ Role in Civil Litigation, Settlement, and Access to Justice

We examine judges’ role in civil litigation by studying empirically the relationship between judicial procedural involvement (JPI) and lawsuits’ mode of disposition (MoD). Furthermore, we propose JPI as a metric for the allocation of judicial attention to litigants. Applying the framework to Israeli trial court data, we find that 60 per cent of cases included JPI (through hearings and rulings on motions) whereas 40 per cent involved only the court's institutional function. By juxtaposing JPI and MoD data, we shed light on the scope of judicial involvement in settlements, the ratio between judges’ normative public‐life function and their problem‐solving function, and other pertinent questions. Since nowadays lawsuits are rarely adjudicated, trial rates are low, and litigants in person (pro se litigants) are common, we argue that access to justice should also be construed in terms of access to judicial attention throughout the proceeding, which is readily measurable through JPI.


I. INTRODUCTION
In an era of low trial rates, what is the role of judges in civil litigation? A rich body of literature suggests that their role has gravitated considerably from final adjudication of facts and law to case management and settlement promotion. However, the scope and nature of this shift in judges' role and its that it resulted from governmental policy and legislative reforms that led to a 'decline in public use of the courts for the determination of civil disputes'. 12 In many jurisdictions, including England and Wales 13 and the United States, 14 court docket analyses are constrained by the paucity or poor quality of official data on court usage. This is also the case in Israel, where discrepancies and errors in court docket data 15 make it difficult to obtain an accurate longitudinal picture of case filings, case dispositions, and trial rates. 16 Nevertheless, it is widely held that civil case filings in Israel are on the rise 17 while the number of judges remains constant, inevitably resulting in judicial pressure to settle or otherwise terminate cases early. 18 This article begins to fill this empirical gap. We use a manually coded representative sample of civil case terminations in Israel between 2008 and 2011 to report accurate rates not only of trial and judgment, but of all case terminations. We then capture the frequency of different types of JPI in litigation and examine their relationship to MoD. * * * 12 Id. (2012), p. 6. Genn refers specifically to factors such as the cost of litigation, competition in the form of arbitration and mediation services, and cuts in civil legal aid; id., pp. 8-15. Dingwall and Cloatre similarly argue that reforms in the civil justice system have led to diverting cases from courts; Dingwall and Cloatre, op. cit., n. 10, pp. 62-66. 13 Genn, op. cit. (2012), n. 11, p. 6 ('The bare figures … leave many open questions that can't be answered because of the poverty of official data on court usage'). As an illustration, Dingwall and Cloatre report a decline in the number (rather than rate) of trials in English County Courts, aptly cautioning that '[i]t can be somewhat misleading to concentrate exclusively on absolute numbers rather than looking at rates, which correct for possible changes in the population of events that generate these data'; Dingwall and Cloatre, op. cit., n. 10, pp. 57-58. Building on the observation that settlement has become 'the modal civil case outcome', 19 scholars often associate the vanishing trial phenomenon with the emergence of a settlement culture. 20 The rarity of civil trials and the prevalence of settlements have also been linked to a shift in the focus of judicial functions: from trial and adjudication to pre-trial case management 21 and settlement promotion. 22 Some have suggested that the trial judge as a 'black-robed figure up on the bench, presiding publicly over trials … has become an endangered species' 23 that is vanishing along with trials, or at least becoming significantly transformed in nature. The meaning of this transition for judges, litigants, and the legal system is the subject of an ongoing debate. 24 The causes of this transition, on the other hand, are well established. Increased caseloads have prompted courts to shift their efforts from trial to early disposition of cases before a trial ensues. 25 It is further suggested that the popularity of alternative dispute resolution concepts and their institutional promotion within courts have contributed to a reduction in trials and an increase in settlements. 26 Finally, judges' managerial and settlement-promoting roles have become formally institutionalized. In the United States, Rule 16(C)(2)(I) of the Federal Rules of Civil Procedure (1993) states that 'the court may … take appropriate action … settling the case and using special procedures to assist in resolving the dispute'. In England and Wales, Rule 1(4)(2)(f) of the Civil Procedure Rules (1998) lists 'helping the parties to settle the whole or part of the case' as one of the court's duties. In Israel, Rule 140 of the Rules of Civil Procedure   27 authorizes the judge to conduct a pre-trial hearing 'to examine the prospects of a settlement between the litigants'. Thus, settlement promotion has become an integral part of legal proceedings; it is 'not "bargaining in the shadow of the law", but bargaining as a requirement of the law' and '[t]he dispatchers are the judges'. 28 Judges influence litigants' settlement decisions in multiple ways. First, judges are 'a ghostly but influential presence, through their rulings in adjudicated cases and their anticipated response to the case at hand'. 29 Moreover, judges 'stand ready to step from the shadows and resolve the dispute by coercion if the parties cannot agree'. 30 In addition, as we suggest in this study, judges can directly influence litigants' decisions to settle through their procedural involvement in the litigation -that is, JPI.
We have created the JPI analytical framework to untangle the empirical relationship between the vanishing trial phenomenon, the settlement culture in courts, and the scope and nature of judges' roles in civil litigation. First, we distinguish between the judicial and institutional functions of courts. Judicial functions refer to actions of judges throughout the litigation process that rely on judicial discretion and skill. The JPI framework measures three types of such actions: (1) ruling on motions during the filing and defence-pleading phases; (2) conducting pre-trial hearings; and (3) presiding over trials. 31 The court's institutional function represents a fourth category, which is manifest in cases that terminate without JPI, such that no significant discretionary judicial activity takes place and only the court's institutional power and authority are in operation. MoD categories: judgment on the merits, judgment by way of compromise, settlement (court-approved and out-of-court), default judgment, dismissal due to lack of prosecution (abandonment), voluntary withdrawal, and an 'all other' category encompassing the remaining MoDs. We apply the analysis to a representative sample of 1,283 Israeli civil trial court cases, to evaluate the rate and type of JPI in adjudicated cases, settled cases, and cases with other MoDs, focusing particularly on lawsuits in which judges determined the outcome or had a potential impact on litigants' decisions regarding the MoD.

Research population, sample, and database
The Israeli legal system is considered a 'mixed jurisdiction' -a common-law system with several continental European influences. 32 Accordingly, while it has a code-like private law legislation that incorporates general standards typical of continental statutes, judge-made law is still an important part of the law in Israel. The legal system follows the rule of precedent, and judgments are articulated in the same way as common-law decisions. 33 In Israel, there are two general civil trial courts of original jurisdiction: Magistrate courts (Hashalom), which have jurisdiction over civil lawsuits up to 2,500,000 NIS (roughly 700,000 USD), and District courts, which have jurisdiction over lawsuits of higher value. 34 Our analysis focuses on Magistrate courts, which handle more than 95 per cent of general civil cases of original jurisdiction. 35 The representative sample in our study consists of 1,283 civil cases that terminated in six Magistrate courts 36  Judiciary, in a study headed by Weinshall-Margel and Taraboulus. 38 They used a two-stage stratified sampling method to ensure that the sample accurately and proportionally represented the population of terminated civil cases of original jurisdiction. 39 First, courts were sampled, and then, for each selected court, cases were randomly sampled to represent the distribution of cases that opened in that specific court. The sample size constitutes approximately 2 per cent of civil case terminations in the selected courts during the stated timeframe; its reported confidence level is above 95 per cent and the sampling error is below 6 per cent. The Research Division employed five law students who read all case documents and coded attributes of the case, procedural events, and the outcome. 40 The sample has several advantages for exploring the relationship between JPI and MoD. First, the dataset overcomes a common limitation of empirical studies in this area: the low accuracy and reliability of courts' registered data. 41 Specifically, the Israeli electronic court docket data include significant discrepancies and inconsistencies, 42 and certain MoD codes are particularly unreliable: some of them are only 60 per cent accurate. 43 Our analysis, by contrast, relies on data gleaned from an accurate and reliable sample of cases, which were manually coded for variables pertaining to pleadings, motions, hearings, disposition, and other such attributes. Importantly, coders identified instances of substantive settlements that were formally disposed otherwise (for example, when the plaintiff's voluntary withdrawal of the lawsuit explicitly mentioned that the parties settled).
Moreover, several attributes of Israeli civil trial courts make them a good setting for exploring the relationship between the procedural and dispositive dimensions of judges' work. First, there is no selection bias in the subset of cases that Israeli judges decide (unlike judges in jurisdictions conducting jury trials). Second, typically, the judge who handles the case in the early phases of litigation continues to preside over it if a trial ensues. Finally, Israeli trial judges rarely conduct confidential settlement conferences in chambers. In-person interactions with the parties, including settlement promotion, take place almost invariably in written decisions and public hearings. 44 Thus, in Israeli trial courts, hearings and rulings on motions are a good measure of JPI, and judicial settlement promotion has a public nature. By contrast, in both English and American courts, 'judges can locate such exchanges in chambers to which the public has no access'. 45

II. THE PROCEDURAL AND DISPOSITIVE DIMENSIONS OF CIVIL LITIGATION
Civil trials involve procedural and dispositive dimensions. The procedural component entails a public, fair, and impartial process of examining facts and legal issues arising between the parties. The dispositive component is reflected in a judicial decision that ends the legal dispute. Despite their obvious relatedness, the two dimensions do not necessarily overlap. For example, settlements can occur at any procedural phase, and trials are not commensurate with adjudication on the merits. Understanding the relationship between the procedural and dispositive dimensions is critical to any normative consideration of trial rates, MoD distribution, and the role that courts and judges play in these regards. Such analyses have bearing on a host of empirical and normative questions. As Hadfield notes: If the reduction in trial rates is a consequence of increased rates of abandonment and default, does that reflect mounting barriers to engagement in the legal process? Does it reflect increased disparities between the haves and the havenots? Are single-event trials before bench or jury being replaced by more piecemeal nontrial adjudication by judges as a consequence of increased  judicial determination constrain normative discussions about the individual and collective social functions of courts, and specifically the nature and purpose of public adjudication and the privatization of civil justice. Our analysis of the relationship between JPI in civil litigation and the dispositive outcome of the litigation taps into this discourse. Subsections 2.1 and 2.2 lay the foundation for the analysis, by briefly introducing the nature and distribution of procedural phases and MoDs in our sample. Subsection 2.3 sets the stage for Section III, in which we describe our operationalization and findings, capturing judges' engagement in litigation in terms of JPI.

Case terminations by procedural phase
We classify cases into four categories that reflect the procedural phase in which they terminated, using the last procedural event that occurred prior to their termination as an indicator. Filing comprises lawsuits terminated before defence pleadings were submitted or a hearing occurred. Defence pleadings includes cases in which defence pleadings were filed but no pretrial or trial hearings took place. Pre-trial includes cases in which at least one pre-trial hearing was held and no trial hearings occurred. Judges can hold pre-trial hearings after the defence pleadings are filed 'to clarify the dispute and the procedures necessary for addressing it, in order to promote efficiency, simplification, brevity and speed in the proceeding, and to examine the prospects of a settlement'. 49 Trial includes cases in which at least one trial hearing occurred. 50 Table 1 summarizes the distribution of the procedural phase in which cases in our sample terminated. It answers the question 'Where have all the trials gone?', 51 at least with respect to filed lawsuits: of the 1,283 cases, 654 (51 per cent) did not progress beyond the filing phase, 192 (15 per cent) terminated after defence pleadings were submitted, 277 (22 per cent) terminated during 49 Israeli Rules of Civil Procedure, §140. 50 The dataset does not indicate whether a case was terminated during trial or after its conclusion. 51 Hadfield, op. cit., n. 14. See also: Genn, op. cit. (2012), n. 11, p. 6 (pointing to the importance of asking 'Where have the disputes and trials gone and what is the outcome?').
the pre-trial phase, and 160 (12 per cent) terminated during or after trial. The findings demonstrate that the trial rate in Israeli Magistrate courts is higher compared to American and English courts. 52 To summarize, 1,123 (88 per cent) of the civil lawsuits in our sample are 'vanished trials', and the adversarial nature of litigation is limited in scope; there was no formal presentation of a defence in about one half of the cases, and a public hearing before a judge took place in 437 (34 per cent) of the cases. Young and Singer suggest that the courtroom encounter between the judge, litigants, and litigants' attorneys generates a 'bench presence', and that this could be used as a metric of procedural fairness. 53 They define 'bench presence' as the time that a 'judge spends in the courtroom, conducting trials or otherwise presiding over an open proceeding … [performing a]ny task that involves the judge's presence in the courtroom in furtherance of an adjudicative purpose'. 54 Young and Singer exclude settlement conferences from their definition of 'bench presence', possibly because they define it in the context of American district courts, where settlement conferences are typically conducted in judges' chambers rather than in public. Arguably, in Israel, both trial and pre-trial hearings generate 'bench presence', even when the latter involve judicial settlement promotion. Pre-trial settlement hearings are public, and pre-trial judges may use them to exercise inquisitorial capacities, which include inviting witnesses and interrogating them, setting the agenda for the issues that need to be determined, appointing experts, and using other case management techniques. 55

Case terminations by MoD: classification and (nuanced) measurement
Court docket studies differ in their MoD definitions, especially with regard to non-contested MoDs. 56  ends the litigation, of which the court has been notified. 59 Accordingly, when a litigant filed a motion to roll a settlement into an enforceable judgment, the case was classified as a court-approved settlement. In addition, if a plaintiff's motion to voluntarily dismiss the case mentioned that the parties had reached a settlement, the case was classified as an out-of-court settlement. Thus, our empirical definitions distinguish substantive settlements not only from judgments on the merits but from all of the MoD categories that we specify below. This distinction is important because the claimed association between the vanishing trial phenomenon and a settlement culture in courts is enhanced by empirical studies that use the term 'settlement' to describe all litigated disputes that terminate without adjudication on the merits. 60 A default judgment represents cases terminated upon the plaintiff's motion to issue a judgment based solely on the lawsuit following the defendant's failure to submit defence pleadings or appear in a hearing. 61 Cases are dismissed due to lack of prosecution if the plaintiff fails to complete a required procedural action (such as filing a motion or complying with a court order) and remains inactive thereafter. Voluntary withdrawal reflects the dismissal of the lawsuit upon the plaintiff's notification of ceasing the litigation (excluding cases in which the withdrawal mentions that the litigants have settled).
A judgment on the merits is a reasoned judicial decision that determines which of the litigants' claims prevailed legally. 62 In addition, litigants may authorize the judge to issue a 'judgment by way of compromise' in accordance with §79A of the Israeli Courts Law. While the law does not define this term, 63 it is used 'in action' as a form of judicial arbitration, either to determine the outcome without providing a reasoned decision or to provide a judgment that the judge perceives to be a fair outcome, albeit not necessarily the result of a   Table 2 represents the MoDs of cases in our sample. It indicates that 132 (10 per cent) of cases terminated in a judgment on the merits, and that settlements accounted for nearly half (47 per cent) of the case terminations; 35 per cent of the cases were court-approved settlements, and an additional 12 per cent were out-of-court settlements. Another 8 per cent of the cases were voluntarily withdrawn without indicating that a settlement had been reached. Notably, in 28 per cent of the cases, the litigation ceased due to litigant inaction or disengagement with the process; default judgments terminated 20 per cent of cases and 8 per cent of cases were abandoned following the plaintiff's lack of prosecution. Finally, 4 per cent of cases terminated in a judgment by way of compromise. 65 To summarize, the 90 per cent of cases that were not adjudicated on the merits include 47 per cent substantive settlements and 43 per cent other MoDs. These findings demand that we revisit the premise of empirical and . 65 The classification of this MoD for the purposes of analysis depends on the determinant qualities that one ascribes to adjudication on the merits. A focus on discretionary judicial determination of the dispute could justify lumping §79A judgments in with adjudication on the merits, bringing the total rate of the latter to 14 per cent. By contrast, an emphasis on adjudication through public deliberation, reasoning, and application of law lends support to associating §79A judgments with settlement, bringing the overall rate of settlements to 51 per cent. However, due to the uniqueness of this MoD to the Israeli context and its relatively low frequency (56 cases), in our analysis we refer to it as neither settlement nor adjudication on the merits. economic accounts of litigation-related settlements that treat any case that is not adjudicated on the merits as settled. 66 As we elaborate in the next section, such a binary conceptualization is insufficient for a normative evaluation of trial scarcity, settlements, and judges' role in litigation.

Using JPI to evaluate civil litigation
The finding that 12 per cent of cases progressed to trial and 10 per cent were adjudicated on the merits 67 could be construed to suggest that the public role of the court as a sociopolitical institution that proclaims norms and delivers justice is fairly limited. 68 It may also raise doubts concerning the proportion and type of disputes, litigants, and legal issues that receive judicial attention, as a resource allocated by the justice system. To shed light on these issues, we use the JPI framework to capture the scope and nature of the distinct role that judges play in litigation. Specifically, we use JPI to measure the ratio between judges' public-life function (proclaiming public values) and their problem-solving function (resolving private disputes). Furthermore, we suggest that the different types of JPI provide a necessary, more intricate measure of access to justice. Our proposed JPI framework captures whether and how judges were procedurally involved in each case and the relationship between their procedural engagement and the MoD. To that end, we conceptualize the filing and defence-pleading phases, which are defined as such by litigant actions, using an alternative measure that captures JPI: ruling on motions. We use this measure in conjunction with two other types of JPI: judges' presiding over pre-trial hearings and trial hearings. Section III details the operationalization and findings of this analysis. Section IV demonstrates several analytical and policy applications of the JPI framework.

Evaluating judges' role in an era of low trial rates
In light of the rarity of adjudication by trial and judgment, these two paradigmatic judicial behaviours insufficiently capture judges' role in civil litigation. This is especially true given the ubiquity, richness, and potential impact of judicial case management and settlement promotion practices. Accordingly, the JPI framework operationalizes docket data to capture judges' procedural role throughout the litigation lifecycle (three JPI types) and the court's institutional function (cases that terminate without JPI). It then becomes possible to examine the relationship between judges' role in litigation and the dispositive outcome of litigation (MoD).

(a) The dispositive effect of JPI in litigation
There are various ways in which JPI can influence the parties' litigation behaviour and their decision to pursue a particular MoD. First, judicial written and oral interactions throughout litigation convey meaningful signals regarding the specific case. These judicial communications, which carry the weight of judicial authority, help litigants to reduce the levels of uncertainty and asymmetric information about matters in dispute as well as to develop more similar cost/risk analyses (estimates of liability, damages, and the cost of litigation). 70 The litigants may also gain insight into the court's (preliminary) views on factual and legal aspects of the case. 71 In addition to providing valuable information, judges can directly facilitate litigants' bargaining processes, 72 engage in case management techniques, 73 and create procedural mechanisms that affect litigants' decisions about the MoD. These judicial inputs can be delivered through written communication in the form of ruling on motions 74 or through judges' 'bench presence' interactions with the litigants during pre-trial 75 and trial 76 hearings. A recent observational study of pre-trial hearings in Israeli Magistrate courts documented such judicial engagement in extensive -and often intensive -behaviours that were aimed at influencing litigants' willingness to settle. 77 Obviously, judges can only influence cases in which they have an opportunity to provide input. Thus, we use the JPI framework to evaluate judges' role in cases in which they are actively involved (in contrast to cases that terminate without JPI). We also note that judges play a different role in cases that dispose for reasons such as lack of jurisdiction or litigant inaction compared to cases that dispose substantively, through adjudication on the merits or settlement.
(b) Operationalizing case terminations in terms of JPI JPI typically intensifies in conjunction with the progression of the litigation through the four procedural phases of filing, defence pleading, pre-trial, and trial. The latter two phases, which account for 34 per cent of the case terminations in our sample, are readily defined in terms of JPI. By contrast, the former two phases, which account for 66 per cent of case terminations, are defined by litigant behaviours: filing and defending a lawsuit. We reclassify these cases by JPI based on whether judges ruled on motions prior to the final disposition. We view cases in which neither ruling on motions nor hearings occurred as exposed solely to the court's institutional function.
Since our database does not include a count of judges' rulings on motions, we used as a proxy the filing of motions by the plaintiff. Case termination requires the filing of at least one motion. 78 Therefore, we classified cases in which one or fewer motions were filed (and no hearing was held) as falling under the court's institutional function (517 cases, or 40 per cent), and cases in which the plaintiff filed two or more motions (and no hearing was held) as exposed to JPI through motions (329 cases, or 26 per cent). 79 Notably, the average number of motions in the 'institutional function' and 'JPI through motions' categories was similar within both the filing and defence-pleading phases. Figure 1 summarizes this operationalization.
The information summarized in Table 3 supports the proposition that motion activity is a useful differentiator between the court's institutional function and judicial motion activity, but not between the latter and pre-trial or trial activity. The last column in Table 3 shows that the court's institutional function was the only category in which the mean number of plaintiff motions 77 Sela et al., op. cit., n. 44. Examples of documented practices include encouraging negotiation, emphasizing the costs and risks associated with trial, opining on disputed issues, and providing implicit or explicit predictions of the likely outcome. 78 There are specific exceptional instances, such as lack of prosecution, in which the court is authorized to initiate case termination. 79 We note that 64 of the 846 cases that terminated during the filing and defencepleading phases (7.5 per cent) included more than one motion by the defendant, and that 36 of those cases (56 per cent) are captured in our analysis since they also fulfilled the criteria of 'judicial motion activity' based on motions filed by the plaintiffs.  was significantly different from other JPI categories. This finding is consistent with our stance that JPI intensifies as the litigation proceeds from written submissions to hearings. Table 3 also quantifies the intensity of each JPI category. It indicates that in the 'JPI through motions' category the number of plaintiff motions ranged between 2 and 9, averaging at 2.81 (std. = 1.17). Similarly, it shows that the average number of hearings in cases terminated in pre-trial was 1.58 (std. = 0.98) compared with an average of 2.50 hearings (std. = 1.97) in cases that ended during or after trial.
Applying the JPI framework to our sample of 1,283 case terminations, we find that 329 (26 per cent) of the cases terminated following JPI through motions, 277 (22 per cent) terminated following JPI in pre-trial, and 160 (12 per cent) terminated following JPI in trial. While the type and extent of judicial interaction with the litigants (and their attorneys) in each JPI category is different, one thing is clear: judges 'vanished' from civil litigation in 517 (40 per cent) of the cases, which ended without JPI. In these cases, the court performed an institutional function.  = 1,283 Our finding regarding the scope of courts' institutional function is critical: while it is commonly asserted that judicial case management largely drives the scarcity of trials, 80 the sizeable institutional function that we observe marks the limits of this proposition in Israeli civil courts. At the same time, by identifying that 26 per cent of the cases were exposed to JPI through motions, we reveal the scope of potential judicial influence on litigation outcomes prior to the pre-trial hearing, including litigants' decision to settle. per cent): a dismissal due to plaintiff abandonment (lack of prosecution, 15 per cent), a default judgment in the absence of defence pleadings (28 per cent), 81 and a technical termination due to reasons such as lack of jurisdiction or statistical closing (5 per cent). In addition, plaintiffs actively withdrew 8 per cent of the lawsuits that terminated without JPI. The court's institutional function also includes enabling settlements (27 per cent + 13 per cent = 40 per cent), lending support to the notion that merely filing a lawsuit (and defence pleadings) transmits information that helps litigants to negotiate a settlement. 82 In cases that terminated following JPI through motions, the most frequent MoD was settlement (38 per cent + 19 per cent = 57 per cent). This finding is consistent with the idea that information transmitted via judges' ruling on motions on issues such as discovery, appointment of experts, and extension of procedural due dates can assist litigants to reach a settlement agreement. Further studies are needed in order to describe and understand this mechanism, especially given the relatively low average number and wide range of motions per case, the variety of motions and rulings, and our reliance on a proxy to measure judicial motion activity. It is worthwhile to note that litigant inaction accounted for 30 per cent of case terminations in this category (25 per cent in the form of default judgments). 83 Cases that terminate following JPI in pre-trial hearings involve in-person judicial interaction with the litigants. Consistent with the stated goals of these hearings, settlement was the most common MoD in this JPI category (55 per cent + 5 per cent = 60 per cent). Interestingly, in 10 per cent of cases in this category, litigants opted for a judgment on the merits while relinquishing their right to a trial prior to judicial determination. 84 As expected, case terminations following JPI in trial presented the reverse distribution: adjudication on the merits was the most frequent MoD (53 per cent), followed by settlements (21 per cent). 85 81 In shortened-track lawsuits, the defendant must obtain the court's permission to file defence pleadings (Israeli Rules of Civil Procedure, §204). When the court does not grant permission to defend, the case may still be classified as a default judgment, but the scope of this unique situation appears limited: of the 147 cases that terminated in a default judgment under the court's institutional function, only four included a motion filed by the defendant. 82 Hubbard, op. cit., n. 71. 83 While there are circumstances in which default judgments in shortened-track lawsuits do not reflect defendant disengagement (when the court denies the defendants' motion to defend), the potential scope of this scenario is limited: of the 44 shortened-track cases that terminated in the 'JPI through motions' category, only three (7 per cent) included a motion on the part of the defendant. 84 An additional 9 per cent of the cases that terminated in pre-trial ended in a judgment by way of compromise, representing a different form of adjudicatory determination. 85 Judgments by way of compromise accounted for 19 per cent of the cases. Given the advanced stage of litigation, it is plausible that some of these judgments were in fact settlements.

Describing the relationship between JPI and MoD
To understand judges' role in litigation, the abovementioned MoD distributions should be contextualized by accounting for the size of each JPI category, which is exemplified by the height of the bars in Figure 2 and reported in the cross-tabulated JPI and MoD data in Table 4. Specifically, we underscore the finding that of the 766 cases that judges handled through JPI, 391 (51 per cent) settled. In addition, among the cases that were actually handled by judges, adjudication on the merits had not disappeared to as large an extent as suggested by the aggregate data: 132 of 766 cases with JPI (17 per cent), 86 compared with 132 of 1,283 cases in the aggregate data (12 per cent). Moreover, of the 437 cases that were publicly heard in pre-trial or trial, 202 (46 per cent) settled and 113 (26 per cent) were adjudicated on the merits. 87 Looking at the cross-tabulation of JPI and MoD in Table 4, the intricate relationship between judges' role and the vanishing trial phenomenon is evident. While judgments on the merits were given mostly during or after trial (85 of 132 cases, or 64 per cent), about a fifth of them (28 of 132 cases, or 21 per cent) occurred during pre-trial. 88 In those cases, a pre-trial hearing sufficed for judges to form a judgment and for litigants to bind themselves to a judicial determination while forfeiting their right to a trial. Furthermore, Table 4 shows that judges were procedurally involved in the majority of the cases that settled (391 of 599 cases, or 65 per cent). 89 However, the scope and type of JPI in out-of-court settlements and court-approved settlements were different. 90 Furthermore, we note that the majority of voluntary withdrawals (58 per cent) occurred following JPI. Thus, it is plausible that at least some of these withdrawals reflect out-of-court settlements that were not reported to court, especially when the parties were exposed to 'bench presence' interactions with the judge during pre-trial (21 per cent) and trial (5 per cent) hearings. Finally, Table 4 indicates, as reasonably expected, that certain MoDs were effectuated primarily under the auspices of the court's institutional function: 147 of 251 default judgments (59 per cent), 77 of 102 dismissals due 86 Of the 766 cases, 56 (7 per cent) were determined in a judgment by way of compromise. 87 Of the 437 cases, 54 (12 per cent) were determined in a judgment by way of compromise. 88 When looking at judgments by way of compromise, 54 per cent occurred during trial and 43 per cent during pre-trial. 89 Judges were procedurally involved in 308 of 448 court-approved settlements (69 per cent) and in 83 of 151 out-of-court settlements (55 per cent). 90 While out-of-court settlements were exposed primarily to the court's institutional function (68 of 151 cases, or 45 per cent) or judicial rulings on motions (64 of 151 cases, or 42 per cent), court-approved settlements were distributed quite similarly across the court's institutional function (140 of 448 cases, or 31 per cent), judicial motion activity (125 of 448 cases, or 28 per cent), and pre-trial hearings (153 of 448 cases, or 34 per cent).

JPI and the sociopolitical role of courts and judges
In this article, we examine the role of judges in a civil litigation landscape that is characterized by large caseloads, low incidence of trials and judgments on the merits, and a significant rate of settlement. The normative implications of these trends are subject to debate. One view is that 'the public dimensions of [judges'] work are diminishing' because 'the framework of "due process procedure," with its independent judges and open courts, is replaced by what can fairly be called "contract procedure"'. 92 This view is associated with concerns that these trends reflect a privatization of courts, eroding the 'cornerstone … of their functioning' 93 : creating, applying, pronouncing, and interpreting the law, substantiating political rights and democratic procedures, and promoting public values and exerting judicial power in a transparent process. 94 A different view is that breaking away from formal adjudication and encouraging settlements creates a space for more nuanced and inexpensive modes of dispute resolution that can improve the processing of disputes and introduce non-binary notions of justice that better serve the varying needs of disputants. 95 In this vein, Luban has articulated two conceptions of courts. 96 The 'problem-solving' conception emphasizes courts' dispute resolution function, such that 'the wisdom of the judge consists not in issuing the wisest orders, but in facilitating the quickest and most painless resolution of disputes. Rather than debating principles, the judge tries to wrestle the interests of the parties into alignment.' 97 By contrast, the 'public-life' conception of courts 91 The remaining 144 cases in these MoD categories, which altogether amount to 11 per cent of the sample, were mostly terminated following JPI through motions. Additional research is required to examine whether these cases are more akin to those that terminated under the court's institutional function or to out-of-court settlements following judicial motion activity. 92 Luban, op. cit., n. 24 emphasizes that '[a]djudication … is necessary to define and redefine the conditions of the public space' 98 as it declares and promotes public values.
Recent scholarship continues to echo Luban's framework, outside the United States as well. Writing about England and Wales, Mulcahy has lamented that while courts may still be identified as a 'vital institution in the public sphere … the civil litigation system now displays many of the features of a private dispute resolution institution with a strong preference for resolving disputes in private and a minimal interest in the public airing of important legal debates'. 99 Prioritizing one function over the other may have far-reaching practical ramifications, ranging from an insufficient number of precedents 100 to mounting barriers to access to justice in courts. 101 In fact, there seems to be little disagreement that both the public-life and problem-solving functions of courts are necessary for society to prosper; the challenge lies in determining the desirable ratio between them, given the 'forces of friction in the legal system'. 102 This is an important sociopolitical and normative question; whichever stance one takes on the matter, it should be informed by the relevant data. In this section, we demonstrate that combining the JPI framework with a nuanced MoD analysis helps to answer this question.
The first step is to overcome the empirical and definitional challenges in measuring the two court functions with available procedural and MoD data. Two reasonable premises are (1) that adjudication on the merits fulfills both public-life and problem-solving functions; and (2) that settlement fulfills a problem-solving function (typically, in a less costly and more timely fashion than adjudication) and that it is unclear whether it has a public-life function. The remaining MoDs are not as clearly indicative of either function, since they neither entail an adversarial process followed by a reasoned judicial determination nor necessarily resolve the legal dispute.
Hadfield usefully demonstrates the importance of considering the substantively different nature of certain MoD categories in evaluating the 98  vanishing trial phenomenon. 103 Specifically, she singles out lawsuits that terminate due to either the plaintiff's or the defendant's disengagement with the proceeding (lack of prosecution and default judgment, respectively) and non-final dispositions (such as transfers, stays, and dismissals without prejudice). Similarly, it is helpful to examine the ratio of the public-life and problem-solving functions by looking at judgments on the merits and settlements, respectively. Furthermore, such MoD differentiation contributes to the normative evaluation of litigation-related settlements, as 'empirical verification is essential if we are to get past the overly generalized and abstracted claims made both for and against settlement'. 104 Table 5 summarizes the results of using different lenses for considering the ratio between the problem-solving and public-life functions of courts. Adding to the nuanced MoD analysis, it indicates that the JPI framework provides a useful lens for accounting for the extent to which judges, as the primary agents of the justice system, carry out the two functions.
As a benchmark, the analysis presented in Row 1 considers the entire sample, juxtaposing cases that reached trial (12 per cent) and cases that terminated before trial (88 per cent), or cases that ended in a judgment on the merits (10 per cent) 105 and all other MoDs (90 per cent). Such an aggregation reflects a 1:9 ratio between courts' public-life function and problem-solving function (Column 5). A more nuanced analysis unbundles the bulk of uncontested MoDs by separating substantive settlements (that is, agreements between the parties that ended the legal dispute) from other MoDs. In our sample, 47 per cent of the cases were substantive settlements 106 and 43 per cent comprised other MoDs, such as default judgments, withdrawals, or dismissals due to technical reasons. 107 Through this lens, the ratio between the public-life and problem-solving functions of courts is 1:5 (Column 6), such that for every case that was adjudicated nearly five cases settled (132 cases compared to 599 cases).
Next, the analysis evaluates the role of judges, focusing on cases in which judges were procedurally involved in a way that could influence litigants' MoD decisions -that is, 60 per cent of the cases in our sample with JPI. By contrast, 40 per cent of the cases, which terminated without JPI (under the court's institutional function), did not present judges with an opportunity to influence litigants' MoD decisions. Accordingly, Row 2 of Table 5 points to a different balance between the public-life and problem-solving functions 103 Hadfield, op. cit., n. 14. 104 Menkel-Meadow, op. cit., n. 24, p. 2671. 105 We did not include in this analysis the 4 per cent of cases that terminated in a §79A judgment by way of compromise due to the ambiguity regarding its classification (see the discussion op. cit., n. 65). 106 In the sample, 35 per cent of the cases were court-approved settlements and 12 per cent were out-of-court settlements. 107 One could argue that default judgments generate a problem-solving outcome, but it seems fair to classify them in a narrow or technical institutional manner. of judges: a 1:3 ratio (Column 6) between adjudication on the merits and substantive settlement (119 cases compared to 391 cases). 108 Through this lens, for every case in which judges exercised a public-life function, there were three cases in which they exercised only a problem-solving function.
The results point to the usefulness of considering the court's institutional and judicial functions as substantively distinct, and suggest that it can help to reframe previous discussions about court functions and the claimed shift from 'independent judicial decision making' to 'technocratic administration'. 109 Finally, one may consider separately cases that terminated following 'bench presence' (or a 'courtroom encounter' between the judge and the litigants), shown in Row 3 of Table 5. This sub-sample (34 per cent) comprises cases in which pre-trial or trial hearings took place. As we reflect on judges' publiclife role in declaring norms and promoting the legitimacy of courts and trust in the administration of justice, public hearings are particularly meaningful. As Resnik notes, '[o]pen court proceedings enable people to watch, debate, develop, contest, and materialize the exercise of both public and private power' and 'to learn first-hand about processes and outcomes'. 110 Thus, even though hearings took place in less than a third of the cases, these were cases that allowed the public to perceive and address judges directly, and thus form an unmediated opinion about judges' public-life function. 111 In this subset of cases, the ratio between the public-life function (measured in 26 per cent judgments on the merits) and the problem-solving function (measured in 48 per cent substantive settlements) is significantly higher, reaching 1:2 (113 cases compared to 202 cases). Thus, the public-life role of courts and judges is quite evident in cases that the public can fully observe: for every two cases that were settled, one case was adjudicated on the merits.
The JPI framework enables exploring the potential scope of another hypothesis about judicial influence -that judges inject public-life functions into (some of the) settlements that are reached subsequent to JPI. The gist of this idea is that such settlements are more likely to be shaped, scrutinized, or otherwise influenced by judges than are settlements reached under the court's institutional function. The idea that judges inject certain attributes of their public-life function into settlements through JPI is supported by a recent observational study of pre-trial hearings in Israeli trial courts. 112 Reporting on judicial settlement promotion practices, the study documents examples of significant judicial scrutiny, input, and influence on settlement terms, and suggests that judges may develop or enforce desirable social values, at least in the context of the private dispute, by discussing them or promoting their injection into the settlement agreement as a condition for its formalization in a judgment. 113 Since in Israel such settlement-related judicial interventions take place in public, their public-life effects may extend beyond the private dispute.
A future study is required to explore the merit of this claim, but the JPI framework points to its potential scope. In the most expansive sense, judges were procedurally involved and had the potential to influence the majority of settlements (65 per cent), by ruling on motions or presiding over hearings. In a narrower sense, 34 per cent of settlements were reached following pre-trial or trial hearings, where judges exercised 'bench presence' and could be more actively involved in shaping settlement terms.

JPI as a criterion for access to justice and procedural reform
While litigants' demand for adjudication-backed remedies shows no signs of subsiding, it rarely leads to a full trial process and judicial determination of the merit of litigants' claims. Thus, these components of the legal process are insufficient criteria for evaluating access to justice and procedural justice, and it becomes both useful and normatively desirable to develop complementary measures. One measure can be access to judicial involvement, while acknowledging that different types of cases may require different types of JPI. In this vein, Young and Singer propose using 'bench presence' as a measure of due process or access to justice. They define 'bench presence' as the 'time that a district judge spends on the bench, presiding over the adjudication of issues in a public forum' and suggest that it could be used as a 'rough but meaningful proxy for many components of procedural fairness, by quantitatively capturing the degree to which parties and the public are directly exposed to the judge's practices and procedural safeguards'. 114 The concept of 'bench presence' could be extended by referring to varying degrees of judicial involvement -the three types of JPI. Such an analysis provides a more granular metric for access to justice, capturing different levels of exposure to judicial involvement. From a public policy perspective and as a court system design consideration, JPI provides an expansive and detailed measure for the distribution of judicial involvement as a resource. As Levy suggests, 'judicial attention has come to be a scarce resource', and it is important to evaluate whether it is currently reasonably allocated, given the broad objectives of courts. 115 Arguably, in examining the allocation of judicial attention (or judicial involvement), we should consider also what proportion of this resource is distributed to different members of society, types of litigants, legal issues, and procedures. A great deal of social and symbolic capital resides in courts. 116 Conceptualizing access to judicial involvement as a form of capital, it is important to question how social actors (potential and actual litigants) struggle over the possession and distribution of this capital, given that courts engage with situated practices and repertoires of power and inequality. Thus, the JPI framework can enrich our understanding of power discrepancies in access to justice in courts. Specifically, it may provide a complementary measure for 'how the "haves" come out ahead'. 117 Pursuing such a nuanced measurement of judicial engagement seems particularly relevant in the context of self-represented litigants. As Sorabji explains, 'in the absence of lawyers … courts have had no choice but to carry out administrative functions ordinarily the province [of] lawyers'. 118 In such cases, judges' active engagement, measured by JPI, appears directly linked to access to justice.
In order to further such an understanding of access to justice, it is useful to measure the level of access to JPI according to the characterization of various groups of litigants (for example, represented versus self-represented, or individuals versus corporations), case types (for example, fast-track procedure versus standard civil procedure), or legal matters (for example, copyright versus personal injury). Thus, it is possible to identify the types of cases or litigants that are systematically exposed to reduced or heightened levels of judicial involvement. In this sense, the JPI framework offers a new measure for examining whether and how courts maintain, mitigate, or advance power relations and socioeconomic stratification.
The JPI framework can also inform procedural reform. One immediate implication relates to the institution of online courts, which is underway in several legal systems. 119 Online court procedures typically include a mix of automated triage, case preparation, case management, and settlement procedures, followed by online facilitation of settlements by court staff and, if necessary, adjudication based on the available information or following additional phone, video, or in-person hearings. 120 For example, based on the recommendations of the Civil Courts Structure Review Report, 121 Her Majesty's Court and Tribunal Service in England and Wales is in the process of transferring civil money claims of up to £25,000 to an online court process, which will typically be conducted without an in-person hearing before a judge.
Online courts have sparked significant concerns, including the fear that they limit access to justice, infringe upon the public nature of court proceedings, and take away the ability of litigants to experience procedural justice and bring their case before a judge. 122 The JPI framework enables identification of case categories in which certain procedural safeguards, such as a hearing before a judge, are usually not exercised in the current state of affairs. Therefore, in these cases, moving the litigation online will not deprive litigants of these specific safeguards. For example, in our sample, we found that 40 per cent of the cases went through civil trial courts without any substantive judicial treatment or consideration. An additional 26 per cent were terminated after a document-based process that could likely migrate online in a fairly straightforward fashion. From the point of view of both litigants and courts, there might be better procedural mechanisms for handling these disputes, since in some cases litigation can be both costly and qualitatively inferior compared to other alternatives.
Another important application of the JPI framework in this context is the detection of types of litigants and disputes that exhibit heightened risk of litigant disengagement (default judgments or dismissals due to lack of prosecution). Subsequently, courts may design procedures to increase JPI in