Printer Friendly


INTRODUCTION                                                   1599
  I. TAKING THE MEASURE OF THE BEAST                           1604
 II. MEASURING EFFICIENCY                                      1607
     ACTIONS                                                   1615
     A. Unknowable Facts                                       1615
     B. Lack of Reporting or Misreporting                      1616
     C. Barriers to Data Access                                1617
     D. Challenges of Linking Court Data to Other Records      1620
     E. Limited Desire for More Information                    1621
        1. "How would that information help me?"               1621
        2. "That information might hurt the causes
            and institutions I care about"                     1622
CONCLUSION                                                     1622


A popular trope about long marriages is that they survive because over time the partners come to know each other's good and bad characteristics, to appreciate the good and to tolerate the bad and to decide--consciously or unconsciously--that staying together is better than the available alternatives. As a nation, the United States has been married to class actions for a quite long time: depending on one's definition of the procedure, since the early years of the Republic, (1) the 1938 adoption of the Federal Rules of Civil Procedure, (2) or the 1966 amendment of Rule 23, the adoption of which this conference celebrates. But the liaison has always been uneasy, for the notion of resolving individual rights and property claims through a representative action on behalf of absent parties has always raised due process concerns. (3) The joining of group litigation with a legal regime based on individual autonomy was long considered mainly a marriage of convenience, justified by the inefficiency of resolving large numbers of claims arising out of the same facts and law in individual proceedings. (4) And like other marriages in which the partners seem ill-suited, the amended Rule 23 evoked sharp concern among many contemporary onlookers, who foresaw dire consequences. (5)

As time passed and fracture lines in the marriage of class actions to traditional dispute resolution became apparent, there were efforts to repair Rule 23 in a dramatic fashion by eliminating (b)(3) damage class actions entirely or eliminating the rule's tri-partite structure, (6) or less ambitiously, by adopting incrementally modest changes to the existing provisions. (7) In time Congress got into the act, passing the Private Securities Litigation Reform Act of 1995, (8) intended to rein in securities class actions, and the Class Action Fairness Act of 2005 (9) ten years later, intended to subject class certification to heightened judicial scrutiny, by facilitating removal of state-law based class actions to federal court. At the time of this writing, Congress is poised to act again. H.R. 985, the Fairness in Class Action Litigation Act of 2017, (10) introduced in the House in February 2017 and passed a month later with little debate, is plainly aimed at making it increasingly difficult to certify money damage class actions on the grounds that they are abusive. (11)

As in many acrimonious marriages, over time there has been considerable airing of dirty linen, which has taken the form of charges of "abuse." (12) Like the partners in longlasting acrimonious marriages, the critics and supporters of Rule 23 have trotted out the same charges and countercharges over and over again. Recently, even the neighbors have weighed in, with advocates of adopting class actions outside the United States explaining that while they believe collective litigation could play a beneficial role in their respective jurisdictions, they are not promoting what they perceive as dysfunctional "American-style class actions." (13)

Notwithstanding the vigorous policy debate over the benefits and costs of class actions and the substantial jurisprudence that has developed in response, there has been no comprehensive sustained empirical effort to monitor the consequences of the 1966 amendments to Rule 23. Astonishingly, although critics often point rhetorically to a "flood of frivolous litigation" as a reason for curbing class actions, fifty years after the adoption of the 1966 amendments no one knows how many class action complaints are filed annually in the United States, (14) much less what sorts of substantive legal claims are most likely to give rise to class actions, what their modes of disposition are or what their outcomes are. Through the diligent efforts of researchers at the Federal Judicial Center, (15) and scholars such as Eisenberg and Miller, (16) Fitzpatrick, (17) and Baker, Perino and Silver, (18) we have learned about characteristics of federal class actions at certain points of time; short-term consequences of certain rule changes; key features of Rule 23 (b)(3) settlements approved by federal judges; and trends in filings and outcomes of specific case types. However, judicial and legislative reformers have little empirical evidence to support assertions that general rule changes are necessary and only speculation to guide consideration of what the consequences of these changes might be. A good therapist might advise that it is way past time to uncover the facts related to the stresses in our marriage to Rule 23.

In this Article, I propose a research agenda for systematically investigating the frequency and key characteristics of class action litigation, and assessing how well Rule 23 meets the goals its proponents have long held out for inefficient management of mass claims, efficient and fair compensation of losses due to defendants' negligence or statutory and regulatory violations, and optimal deterrence of illegal behavior. My intention is to begin a conversation about such an agenda, not to set the agenda myself. Parts I-IV identify the types of data we should be collecting and reporting and discuss the challenges different types of data collection present. (19) Part V considers the primary barriers to embarking on sustained systematic empirical research on class actions. (20) This Article concludes by challenging the federal judiciary to promote empirical analysis of the consequences of Rule 23.


More than three decades onwards from Professor Arthur Miller's attempt to demolish the dueling depictions of Rule 23 class actions as "Frankenstein monsters" and "shining knights," (21) we still do not have sufficient data to know what the 1966 revision of Rule 23 effected. As I discuss further, assessing the character of class actions--are they monsters or knights?--presents challenging empirical problems. But the monster appellation connotes something large and out of control, which in turn suggests that the sheer number of class action filings should disturb us, regardless of their character. It seems odd, therefore, that we have gone so long without knowing how many class action filings there are. (22)

Sound policy analysis begins with measurement of the phenomenon of interest. That measurement may take many forms, but usually the first step is to determine the magnitude and shape of the target. Numbers do not tell us everything we ought to know about class actions (or other phenomena of policy concern), but they do help us to determine what is worth arguing about. Moreover, many charges about the negative consequences of class actions, such as the assertion that the risks of class actions are so great that they force defendants to settle non-meritorious claims--so called "blackmail settlements"--rest on empirical assumptions about the pattern of disposition of class complaints, which have also gone largely untested.

Many lawsuits filed in the form of class actions are dropped, dismissed or otherwise resolved without class certification. Over time, judges, lawyers and parties have come to refer to these lawsuits as "putative class actions." These lawsuits, signaled by the filing of a class complaint, constitute the research population of interest. At a minimum, policymakers concerned about class actions ought to know:

1. The number of class complaints filed annually, by case type (e.g. securities, anti-trust, consumer fraud, product liability), party characteristics, venue and category of certification (i.e. (b)(1)(a), (b)(1)(b), (b)(2), or (b)(3)).

2. The mode of disposition of these complaints, whether dropped, dismissed, decided by summary judgment, tried to verdict or settled. (23)

3. Time to disposition. (24)

4. Whether or not these complaints were ever certified as class actions, either for all purposes or for settlement only.

5. Whether certification occurred prior to or following judicial decision on pre-trial dispositive motions.

6. Whether there were Daubert or other evidentiary hearings on certification.

7. Whether the complaint was resolved as part of a multi-district litigation (MDL). (25)

8. Whether there was appellate litigation related to the certification decision and the outcome of this litigation.

9. Whether the final outcome of the complaint at the district court level was appealed and the outcome of the appeal.

Data on filings and mode of disposition ought to be reported in the same form as data on filings and mode of disposition of all civil filings so as to permit calculations of the fraction of claims within case type categories that generate class claims and class-wide resolutions and comparison of disposition patterns between ordinary and class litigation.

Compiling and reporting annual data would permit empirically-based analyses of trends, including analysis of the consequences of doctrinal and legislative changes and rule amendments. (26) For example, reporting the number of class complaints filed annually, by case type and venue, would contribute to an objective assessment of political charges that the federal courts generally or some federal districts in particular are overrun with class actions. (27) Reporting the pattern of dispositions by mode of disposition and the timing of different events would contribute to an objective assessment of political charges that the "in terrorem" effect of Rule 23 leads to many "blackmail settlements." (28) From both normative and practical perspectives, it would be useful to know if the Supreme Court's recent decisions raising the bar for certification (29) are leading to evidentiary hearings that increasingly resemble mini-trials on the merits, as some practitioners have charged. In sum, systematically collected and objectively reported data could short-circuit certain arguments about the consequences of class actions and class action doctrine, while bringing into sharper focus those issues that empirical data indicate are truly worth fighting over.


Case law, commentary and the language of the rule itself have long recognized that efficient judicial management of a multiplicity of claims arising from the same factual circumstances is a goal of Rule 23. (30) Yet there is little empirical data relating to the workload class actions impose on courts. This may be because the efficiency gains of resolving hundreds or thousands of claims using an aggregative procedure--a consolidated non-class proceeding such as an MDL, a Rule 42 consolidation in a single court, or a class action--is so obviously more efficient than resolving each such claim individually with a full panoply of procedural rights. (31) Determining the relative efficiency of different forms of aggregation would be difficult because, in practice, large-scale litigation is often addressed with multiple procedural tools at the same time. For example, a single MDL may include a mix of many class actions and multiple individual lawsuits, (32) and may result in a class wide resolution under Rule 23 or a non-class "global settlement" reached with attorneys representing large numbers of individual plaintiffs. (33) Moreover, large-scale litigation is not assigned to different aggregative procedures on a random basis, meaning that the effects of different aggregative approaches on court burden would be confounded with differences in the cases themselves, including substantive law, facts, party characteristics, as well as differences in attorney strategies and judicial management preferences.

Notwithstanding these limitations on interpreting workload statistics for class actions there are times when consulting such data might help in weighing policy decisions. For example, when Congress was considering adopting the Class Action Fairness Act, the Judicial Conference worried about the effects on workload of shifting an unknown number of state court class actions to the federal court. (34) The standard approach to measuring judicial workload in state and federal courts relies on weighted case filings--i.e. the numbers of different types of cases pending multiplied by an agreed-upon measure of judicial effort per case type. According to the Federal Judicial Center, the Administrative Office of the U.S. Courts (AO) began using weighted case filings to calculate judicial workloads in 1971. (35) The AO's annual court management statistics currently include the number of weighted case filings per judgeship but do not provide data that would permit us to compare class to non-class (or other aggregated) cases, although the case weights do take into account class action-specific events, such as judicial hearings on class certification. (36) In a national judicial time study of 8320 civil cases filed from 1987-1990, the Federal Judicial Center found that putative class actions consumed "almost five times more judicial time than the typical civil case." (37) But this finding was based on only fifty-one class actions (the only class actions identified in the sample). Moreover, comparing judicial burden for class actions to the "typical" civil case--rather than to the typical complex non-class case--may not provide an accurate measure of the additional judicial burden imposed by class action lawsuits, compared to comparable non-class lawsuits. In any event, there has been no similar study conducted since.

A different measure of efficiency focuses on party and attorney effort to resolve litigation. This is difficult to measure, but legal costs and expenses, relative to the monetary and non-monetary stakes of the litigation, are a good proxy. Recent empirical research on fee awards to class counsel (38) provides clues as to the costs on the plaintiff side but because fees are frequently awarded on a percent of fund basis, that amount may be more or less than the value of the actual attorney time invested in the litigation. Moreover, at least in securities class actions, there is empirical evidence to suggest that fee awards are affected not only by the size of the settlement fund but by other factors including the characteristics of lead plaintiffs, the volume of class action litigation in the district where the lawsuit is resolved, and the managing judge's experience with class actions. Baker, Perino and Silver interpret these data as indicating that experienced class action lawyers "game" the fee award system, (39) which suggests that fee awards are a biased measure of actual costs. Although virtually all motions for fee awards in securities class actions include hour and fee data either to support requests for "lodestar" awards or to serve as a "cross-check" on percentage-of-fund awards, (40) Baker, Perino and Silver's analyses suggest that class counsel manipulate these data, (41) meaning that they also cannot be relied upon as a proxy for true expenses.

Importantly, empirical research on class counsel fees to date excludes expenses associated with the substantial fraction of putative class actions that do not reach settlement but do create expenses for plaintiff class counsel. Moreover, there are no publicly available data for defendants' costs to defend class actions. (42) In sum, when it comes to measuring the efficiency of class actions from the private perspective we are largely in the dark.

Not usually included in attorney fees and expenses, but important components of transaction costs in class action litigation, are the costs of notice (which often amount to one million dollars or more) and the costs of claims administration. These costs are reported to judges at the close of class actions, along with a report of compensation paid to class members. The ratio of these costs to compensation delivered is another measure of efficiency. Moreover, these costs should be included when calculating the sanctions imposed on defendants for the purpose of assessing the deterrence potential of class actions. The financial reports submitted to the judge should be made available to researchers for this purpose after the termination of class action lawsuits.


American scholars have argued that deterrence is the only sensible purpose for class actions when class members' claims are small (43) and perhaps the most important purpose even when claims are large. (44) However, the belief that compensation of losses large and small is an important function of class actions lives on in American discourse about class actions, (45) in the wording of Rule 23, (46) in district court decisions approving or rejecting proposed settlements, in appellate decisions reversing judicial approval of class action settlements, and in public debate over the failures of Rule 23(b)(3) class actions. (47) Notwithstanding this longstanding belief, we know little about how well Rule 23(b)(3) compensates class members--how "fair, reasonable and adequate" settlements actually are--much less how compensation varies across different types of cases or (importantly) different degrees of judicial scrutiny of proposed settlements. We also do not know how different approaches to claims administration affect the delivery of compensation to eligible claimants. The lack of objective information on Rule 23's compensation performance function fuels debate over the value of class actions and creates significant opportunities for misleading characterizations of class action outcomes.

Measuring performance of any compensation scheme is challenging. It requires data on the proportion of eligible beneficiaries who claimed; the magnitude of claimants' and non-claimants' losses; and the amount of monetary benefits paid to claimants. Often the claiming rate and the magnitude of non-claimants' losses are unknown, although perhaps possible to estimate. Claimants' reported losses may be inaccurate if they do not properly understand reporting rules, find them too difficult to comply with, or if they negligently or deliberately exaggerate their losses. In most instances, the aggregate amount of compensation actually paid out is easiest to determine. Aggregate data alone, however, are not sufficient to assess compensation performance: a fair compensation scheme will deliver the same ratio of compensation to loss to similarly situated beneficiaries. Individual-level data are necessary to examine distributional fairness. (48)

Measuring the compensation performance of class actions presents additional challenges. At the time a class complaint is filed the size of the class (eligible class members) is often not known and their losses are difficult to estimate. Moreover, as a result of applicable substantive law, putative class members' compensable losses may be different from their actual losses. (49) By the time a class settlement is negotiated attorneys possess enough information about the scale of the class and the magnitude of class members' losses to agree on an amount defendants will pay to resolve the litigation, the eligibility rules that will determine who can claim compensation, and average compensation for different types of claimants. Often, however, a good deal of uncertainty remains not just at the time of negotiations but at the time a judge certifies a Rule 23(b)(3) class and approves the settlement. As a result, approved settlements--which have provided most of our information on class action settlements to date (50)--offer an incomplete picture of the average compensation promised to class members, and virtually no information on distributional fairness. (51) Because claiming rates vary dramatically and can be vanishingly small, how much compensation was actually delivered, to what proportion of eligible class members, and how it was distributed among class members only become apparent after settlement distribution. This information is contained in financial reports by the claims administrators who typically manage the post-settlement claims process, which are usually submitted to judges when class action litigation is finally closed (and sometimes on a more frequent basis for protracted litigation). This is another reason these reports should be made available for research purposes after the termination of class action lawsuits.

Extensive docket analysis of putative class actions, of the type pioneered in Baker, Perino and Silver's fee studies, (52) could provide a better (although imperfect) measure of the denominator (i.e. losses) for calculating compensation rates. As I discuss later, such analysis is currently hobbled by restrictions on high-volume requests from court electronic dockets (e.g. PACER files). Determining actual compensation requires access to the final financial reports of the claims administrators. To date, courts have proved unwilling to provide access to these reports or the data contained therein. (53)


Assessing how well class actions meet their (often contested) deterrence goal is most difficult of all. (54) To properly assess Rule 23's contribution to enforcement--its "private attorney general" role--requires information on the rate and nature of legal violations that are justiciable through class actions, the proportion of these that are identified by putative class members or their representatives, and the sanctions that are imposed directly and indirectly on defendants as a result of class actions. Direct sanctions are denned as all costs imposed on defendants, including total compensation paid to class members, cy pres remedies, and the cost of injunctive relief and all transaction costs, including legal expenses and the costs of notice and claims administration. Indirect sanctions include reputational capital losses, as reflected in loss of market share and diminished share value, inter alia. Moreover, for purposes of assessing whether Rule 23 is an effective deterrence mechanism, it is important to know the accuracy of identification of legal violations--that is, the rate of false positives to false negatives. False negatives may be impossible to discern. False positives, however, may be reflected in the determination of the merits of putative class actions. Although the definition of a meritorious class action often depends on who is making the determination (55), how a case was resolved--whether by dismissal, summary judgment, or trial verdict-sheds some light on the issue. (56)

In addition, if criminal prosecution or public enforcement actions are available, data on public actions related to the same violations are necessary, as optimal deterrence should take into account all costs imposed on the wrongdoer, lest the combined outcome of private and public actions result in over-deterrence. Systematically investigating the relationships among private and public enforcement requires linking litigation data to regulatory agency data. Currently this is difficult to do on a broad scale but as we move further into a "big data" universe it should become less difficult, unless specific barriers are imposed to prevent such linkage. (57) Until then, case studies relying on small samples of lawsuits offer the most promising understanding of the relationships between private and public enforcement. (58)


Policy analysts perennially complain about the lack of sufficient objective data to support analyses of the consequences of important policies and programs. Analysts of class action policy are no exception in this regard. However, we are at a severe disadvantage by comparison to our colleagues in the fields of health, education, welfare, and national security in that we face a virtual absence of even the most basic information on how class actions operate in federal and state courts. Given the huge resources that have been poured into lobbying Congress on class action issues, the enthusiasm displayed by the Supreme Court for reinterpreting Rule 23, and the near continuous consideration of Rule 23 amendments by the Civil Rules Advisory Committee over the last twenty-five years, the lack of empirical evidence regarding class actions is startling. (59) Some of these barriers are a result of difficult technical challenges. Others, however, are a result of decisions by the Judicial Conference. In this Part I discuss five current barriers to systematic empirical analysis of the consequences of Rule 23.

A. Unknowable Facts

Some information relevant to assessing how well class actions meet their efficiency, compensation, and deterrence goals lies beyond the court system. Defendants are under no obligation to report their transaction costs. Unless individuals and entities harmed by legal violations understand that their injuries or losses were caused by illegal action, decide to claim against the wrongdoer, and have the knowledge, resources, and stamina to pursue their claims, we have no way of measuring compensable losses. (60) We may be able to determine aggregate losses and total compensation paid to victims who did come forward; however, without knowing the losses of those who did not claim, we cannot calculate total losses and therefore the fraction of losses that was compensated as a result of class litigation. Unless public officials or private individuals come forward to allege legal violations we have no way of measuring the frequency or magnitude of such violations. Without this information, we can calculate neither the ratio of violations that resulted in sanctions (as a result of public enforcement or private litigation) to undetected or unpunished violations, nor the ratio of economic penalties to the economic value of illegal behavior that went unpunished. As a consequence of all these unknowns, we can never fully measure class action litigation's consequences. However, with the right information, we could learn much more about these consequences than we know now.

B. Lack of Reporting or Misreporting

The AO has long reported statistics on the federal court caseload. These statistics aggregate data from individual case records including information provided by attorneys and information on case events from court records. (61) The digitized individual case records are contained in the federal Integrated Database. (62) For civil lawsuits, plaintiff attorneys complete a "civil cover sheet" on which they enter party names, the basis for federal jurisdiction, the nature of the suit (using a very detailed set of categories), and, in a section asking what the complaint requests, whether "THIS IS A CLASS ACTION UNDER RULE 23." (63) This class action indicator is one of the variables included in the federal Integrated Database, and has long been the basis for the AO's reports on federal class action litigation. Statistics on the frequency of cases flagged as class actions were included for a number of years in the AO's annual reports on the business of the federal courts. (64)

In the mid-1990s, in aid of its consideration of Rule 23 amendments, the Advisory Committee on Civil Rules requested that the Federal Judicial Center conduct an empirical study of class actions. (65) As described earlier in Part III, when the Federal Judicial Center researchers matched the cases they had identified from an extensive docket and document search of class action activity with the records in the Integrated Database, they found that the latter included only seventy-one percent of the cases identified through the Federal Judicial Center search. (66) This discrepancy might be a result of the fact that at the time a civil suit is filed (and the cover sheet is filled out), the suit is not, as a matter of law, a class action since it has not been certified (and may never be). Or the discrepancy could simply be a consequence of attorney failure to follow the instructions on the form. (67) Due to the concerns that the class action indicator they were relying on was in error, the AO ultimately deleted references to class actions from their annual statistical reports. However, the Integrated Database still includes the class action indicator as one of the individual case variables. Although the Federal Judicial Center study revealed that relying on this indicator to count class actions underestimates the number of civil cases that involve some class-action-related activity, (68) in the absence of more reliable information, analyzing the characteristics of the cases flagged as class actions may still have some value.

C. Barriers to Data Access (69)

For decades, scholars, educators, journalists, and others interested in the operations of the federal judiciary have been able to consult annual statistical reports published by the AO. These reports, previously published on paper only but more recently available online, offer a rich resource for analyzing civil and criminal litigation trends, including filings, time to disposition, and differences across circuits and district courts. For example, the debate over the "vanishing trial" was provoked initially by an analysis of published data on jury trial rates by Galanter, (70) whose earlier research challenging assertions about "hyper-litigiousness" was also based on AO reports and similar reports from some state courts. (71) For those without access to or resources for analyzing case-level data, these reports still offer the first and easiest way to gain an appreciation for what is going on in the federal courts. Given continuing hot controversy over Rule 23, including statistics about class action litigation in the AO reports would have great value. Rather than simply omitting data on class actions because of well-grounded concerns about the reliability of its current class action indicator, the Judicial Conference should instruct the AO or the Federal Judicial Center to conduct a study aimed at improving the reliability of the class action indicator. (72)

A second source of information on civil case filings in federal courts is the Integrated Database that was created in 1970. (73) The database includes a record for each case (civil and criminal) filed in the federal district and appellate courts. The civil case data include information from the civil cover sheet and information about case activity. Until recently the Integrated Database was not generally publicized as available for analysis by scholars or others, but it has been used routinely by Federal Judicial Center researchers as a basis for analyses requested by the judiciary, including the Advisory Committee on Civil Rules. During the 1990s, the database was sometimes shared by the Federal Judicial Center with other scholars who were conducting research on federal litigation. At some point in the 2000s, (74) a decision was made to store these data in the Inter-University Consortium on Political and Social Science Research (ICPSR) archive, which is maintained by the University of Michigan and supported by leading universities across the United States, using a membership fee system. (75) The archive was intended to expand access to research data to the academic community. Over time, it became common for government agencies to require, as a condition of funding, researchers supported by government contracts and grants to deposit their research data in the archive after completion of their studies. This requirement was intended to maximize the value of public funds invested in research.

Depositing the federal court data files in the ICPSR archive had the potential to expand access to federal court records for scholarly research and educational purposes. For example, a law professor who wished to incorporate empirical data on federal tort litigation or securities cases in her lectures and whose university was a member of the consortium could use the database to produce simple tabulations and statistics to share with her students. Researchers could use the data to support analyses of the consequences of changes in statutory and case law and procedures. Ironically, however, after the Integrated Database was deposited in the ICPSR archive, the AO specified that the ICPSR data managers require that anyone wishing to access the data complete a Human Subjects Review process at their home institution. (76) This process, designed to protect people who are the subjects of potentially harmful research, is usually time consuming and, on occasion, arduous. (77) Although the rationale for imposing IRB requirements was never made clear publicly, the likely explanation was a desire to protect the privacy of litigating parties. However, this concern seemed incongruous, given that the data incorporated in the Integrated Database are derived from PACER, the "electronic public access service that allows users to obtain case and docket information online from federal appellate, district, and bankruptcy courts... (which is) provided by the Federal Judiciary in keeping with its commitment to providing public access to court information via a centralized service." (78) Although any study proposing to use the civil case data would likely be deemed "exempt" by most Institutional Review Boards under current Human Subjects Protection regulations (79), and any use for instructional purposes would not come within the purview of these regulations at all, the requirement for review created an impediment to accessing case level data for research purposes. In early 2017 (after I discussed data access issues at this Symposium) the Integrated Database was posted on the Federal Judicial Center website, along with instructions for public use. (80) As useful as analysis of the Integrated Database information may be for some purposes, case docket and document data retrieved from federal court records using PACER would provide the best basis for comprehensive analysis of class action litigation in federal courts. The federal judiciary charges fees for accessing these data under 28 U.S.C. [section][section] 1913, 1914, 1926, 1930, and 1932, which can make large-scale data retrieval expensive. At the discretion of the chief judge of each district court, exemptions may be granted to academic researchers, subject to specified conditions. (81) Because current Judicial Conference policy requires separate requests to each chief judge, requesting fee exemptions can be a time-consuming and arduous process, and anecdotal information suggests that the courts' willingness to grant exemptions varies by judge and perhaps also by the institutional affiliations of the scholars requesting exemptions. (82) Moreover, the most efficient search routines for locating cases and documents of interest (e.g. complaints with class allegations, class certification motions), dubbed "web scraping," are generally blocked, increasing the difficulty of retrieving data for academic research.

D. Challenges of Linking Court Data to Other Records

Particularly for assessing the role of class action litigation in enforcing market regulations, it would be useful to link court record data to other records containing information on public civil enforcement actions and criminal investigations. Currently, researchers can identify the existence of public actions related to private class action litigation by reviewing case-specific documents and performing online searches for other information sources. (83) However, as "big data" analysis becomes more prevalent, there will likely be opportunities for linking court records with other agency records and perhaps social media data as well. This potential will raise both information technology challenges and privacy concerns that will need to be addressed. Limiting access to court data will exacerbate the technical challenges and impede efforts to assess the relative contributions of private and public regulatory enforcement.

E. Limited Desire for More Information

Not everybody shares researchers' enthusiasm for empirical data. Judges sometimes do not perceive how information that researchers are asking for would help them and therefore lack interest in providing it. Policy advocates (including those who advocate for the judiciary) sometimes fear that making more information available would hurt them (or their causes).

1. "How would that information help me?"

Several years ago, I spoke at a bench-and-bar meeting on class actions urging that judges require public disclosure of information related to the disbursement of settlements, including the number of eligible beneficiaries who came forward to claim, the amounts of cash compensation paid to them, other benefits paid pursuant to the settlement, including cy pres awards, coupons, insurance premiums for medical monitoring, how compensation payments were distributed among class members (e.g. medians, means, average payments to sub-groups for matrix schedules), and attorney fee awards. (84) Much of this information is included in reports currently submitted to the judge who presided over a class action by plaintiff class counsel at the close of the lawsuit, but is currently not publicly disclosed; other information should be reported but is unlikely so at present. As discussed in Part IV, this information is critical to assessing the compensation performance of class actions. It is also essential for evaluating the efficacy of different approaches to effectuating class settlement, including notice provisions and procedures for submitting compensation claims that judges should review prior to approving a class settlement. For example, a judge who was armed with information on how "take-up rates" vary depending on these provisions could make better decisions on what standards class counsel must meet when designing settlement plans. However, in the press of managing cases, a judge might not recognize the value to him or her of making these data available for analysis. (85)

2. "That information might hurt the causes and institutions I care about"

It is well understood that information is power. Less frequently acknowledged is that lack of information may also confer power on those who seek to shape public policy. The absence of information about the consequences of legal policies enables policy advocates to urge adoption of new laws and amendments of existing rules, expanding or restricting rights and remedies based on anecdotes or bald assertions. Moreover, some staunch supporters of the federal judiciary may believe that restricting information about judicial activities protects individual judges and the judiciary generally from public criticism. The possibility that more information might impede private policy agendas or impair the independence of the judiciary, combined with the non-empirical tradition of legal scholarship, dampen interest in promoting data collection and analysis and broad access to data that would improve understanding of the consequences of Rule 23.


Class action litigation in the federal courts is largely the creation of the judiciary, which crafted Rule 23 in the 1930s, amended it in 1966, and reshaped and reformed it in the half-century since with only occasional interventions by Congress. Private and public parties were important partners in this process, as they filed class complaints, argued for and against certification, decided to settle or not, argued for or against various types of relief, and took appeals up to the Supreme Court. Private parties also sometimes created obstacles and promoted alternatives to class action litigation that influenced the use of class actions. Judges and lawyers together created innovative class action practices that found their way into the formal rule. Ultimately, however, it is the federal judiciary that gave life to Rule 23 and has shaped its development over the past fifty years.

It is time for the federal judiciary also to take on the responsibility of promoting the analysis of the consequences of Rule 23. A first step in this direction should be to convene a task force of researchers from academia, nonprofit organizations and government agencies to identify the data necessary to support empirical research on the operations and consequences of Rule 23, propose revisions to current administrative reports relating to class actions, review barriers to data access, and recommend solutions to these problems.

Having more data on the consequences of class actions would not end all arguments about whether our marriage to Rule 23 should continue. But it could improve the odds that we are arguing about the right issues. And it might improve the chances that changes in the terms of our marriage have the consequences sought by rule reformers.


([dagger]) Judge John W. Ford Professor of Dispute Resolution, Stanford Law School.

(1) The roots of modern Rule 23 in the federal court system trace back to Equity Rule 38 adopted in 1912 and its predecessor Equity Rule 48 adopted in 1833. See generally STEPHEN C. YEAZELL, FROM MEDIEVAL GROUP LITIGATION TO THE MODERN CLASS ACTION (1987). The notion of group representation embodied in these rules of equity can in turn be traced back to the English Bill of Peace and before that to medieval group litigation in England. Id.

(2) On the history of the 1938 version of Rule 23, see Robert Bone, Walking the Class Action Maze: Toward a More Functional Rule 23, 46 U. MICH. J.L. REFORM 1097, 1100-02 (2013); see also Daniel Hopkinson, The New Federal Rules of Civil Procedure Compared with the Former Federal Equity Rules and the Wisconsin Code, 23 MARQ. L. REV. 159, 170 (1939).

(3) Because of such concerns the outcomes of representative group actions brought under federal Equity Rule 48 were not initially binding on absent parties. In Smith v. Swormstedt, the Supreme Court explained that the rule permitting representative group actions could bind absent parties as their commonality of interests would minimize the risk of unfairness:
Where the parties interested in the suit are numerous, their rights and
liabilities are so subject to change and fluctuation by death or
otherwise, that it would not be possible, without very great
inconvenience, to make all of them parties, and would oftentimes
prevent the prosecution of the suit to a hearing. For convenience,
therefore, and to prevent a failure of justice, a court of equity
permits a portion of the parties in interest to represent the entire
body, and the decree binds all of them the same as if all were before
the court. The legal and equitable rights and liabilities of all being
before the court by representation, and especially where the
subject-matter of the suit is common to all, there can be very little
danger but that the interest of all will be properly protected and

57 U.S. 288, 303 (1853).

Today, "commonality" remains a touchstone for assessing the appropriateness of class certification. See, e.g., Wal-Mart v. Dukes, 564 U.S. 338 (2011) (vacating class certification on the grounds that the class could not satisfy Rule 23(a)(2)'s commonality requirement).

(4) But see Bone, supra note 2, at 1100 (arguing that the structure of the new Rule 23 was puzzlingly formalistic given that the motivation for introducing the new rules of civil procedure in 1938 was to create a more efficient scheme for managing civil litigation).

(5) See, e.g., Eleanore Carruth, The "Legal Explosion" Has Left Business Shell-Shocked, FORTUNE, Apr. 1973, at 65 (quoting New York University law professor Joseph Weiner as stating that "Corporations in the early Thirties may have felt that they were living through the French and Russian revolutions combined, but that wasn't a patch to what is going on now."); see also id. (reporting that William May, CEO of American Can Company, declared, "We are fighting for our lives.").

(6) H.R. 5103, 96th Cong. (1979), endorsed by the Carter Administration, would have substituted a new statutory consumer class action resembling a qui tarn suit for Rule 23(b)(3). The proposal to change the structure of Rule 23, circulated in the early 1990s by Civil Rules Advisory Committee Chair Judge Sam Pointer, was intended to facilitate class actions, especially mass tort class actions. DEBORAH HENSLER ET AL., CLASS ACTION DILEMMAS-. PURSUING PUBLIC GOALS FOR PRIVATE GAIN 25-26 (2000) [hereinafter HENSLER ET AL., CLASS ACTION DILEMMAS].

(7) A multiyear discussion of amending Rule 23 that began in the mid-1990s had by the end of the decade produced only one amendment, Rule 23(f), permitting interlocutory appeals of certification decisions. Id. at 26-37. However, more significant amendments, including the addition of Rule 23(g) on the appointment of class counsel and Rule 23(h) on the award of class counsel's fees, became effective in 2003. As discussed at this Symposium, the Civil Rules Advisory Committee is again considering amendments to Rule 23.

(8) Pub. L. No. 104-67, 109 Stat. 737 (1995).

(9) Pub. L. No. 109-2, 118 Stat. 4 (2005).

(10) Fairness in Class Action Litigation and Furthering Asbestos Claim Transparency Act of 2017, H.R. 985, 115th Cong. (2017). The House adopted H.R. 985 on March 9, 2017 and sent it on to the Senate for consideration.

(11) On the purposes of the proposed statute, see Congress Taking Up Tort Reform with H.R. 9S5, FRIENDS OF THE U.S. CHAMBER OF COMMERCE (March 6, 2017), []; see also H.R.985 - Fairness in Class Action Litigation and Furthering Asbestos Claim Transparency Act of 2017, CONGRESS.GOV, [] (describing the proposed statute in full).

(12) See id.; see also Lawrence Schonbrun, The Class Action Racket, AM. THINKER (Apr. 1, 2013), []; Paul Atkins, The Supreme Court's Opportunity to End Abusive Class Action Securities Lawsuits, FORBES, (Mar. 4, 2014, 3:57 PM), []; James Burling, The End of Abusive Class-Action Lawsuits? Or the End of Industry?, PAC. LEGAL FOUND.: LIBERTY BLOG, (Dec. 30, 2013)[]; Trent Norris & James Speyer, Culling the Herd of Abusive Class Action Litigation, SELLER BEWARE (June 1, 2010), [].

(13) For a discussion of European attitudes towards class actions, see Deborah Hensler, From Sea to Shining Sea: How and Why Class Actions Are Spreading Globally, 65 U. KAN. L. REV. 965 (2017), and Deborah Hensler & Stefaan Voet, Class Actions Across the Atlantic: From Guarded Interest to European Policy (Nov. 2016) (unpublished manuscript) (on file with author).

(14) In the mid-1990s the Federal Judicial Center reviewed docket information for a random sample of 8320 civil cases filed in the federal courts from 1987-1990. The Federal Judicial Center researchers found that only seventy-one percent of cases in which class action activity was recorded in docket information were flagged as class actions in the AO's Integrated Statistical Database that provides the basis for the AO's annual statistical reports on the business of the federal courts. The Federal Judicial Center concluded "there are no reliable national data on class action activity in the federal courts." Memorandum to the Advisory Committee on Civil Rules from Thomas Willging, Laural Hooper and Robert Niemic (Feb. 9, 1995),$file/prelim_class.pdf [] [hereinafter Memorandum to Advisory Committee].

(15) See, e.g., THOMAS E. WILLGING ET AL., FEDERAL JUDICIAL CTR., EMPIRICAL STUDY OF CLASS ACTIONS IN FOUR FEDERAL DISTRICT COURTS: FINAL REPORT TO THE ADVISORY COMMITTEE ON CIVIL RULES (1996), [] [hereinafter EMPIRICAL STUDY FINAL REPORT] (assisting in the Advisory Committee on Civil Rules' then ongoing review of proposed amendments to Rule 23); EMERY G. LEE III & THOMAS E. WILLGING, FEDERAL JUDICIAL CENTER, IMPACT OF THE CLASS ACTION FAIRNESS ACT ON THE FEDERAL COURTS: PRELIMINARY FINDINGS FROM PHASE TWO'S PRE-CAFA SAMPLE OF DIVERSITY CLASS ACTIONS (2008), [] (assessing, at the request of the Judicial Conference of the United States, the impact of CAFA on federal court workload).

(16) Theodore Eisenberg and Geoffrey Miller, Attorney Fees and Expenses in Class Action Settlements: 1993-2008, 7 J. EMPIRICAL LEGAL STUD. 248, 248 (2010) (detailing settlements in class action and shareholder derivative cases in federal courts from 1993-2008).

(17) Brian Fitzpatrick, An Empirical Study of Class Action Settlements and Their Fee Awards, 7 J. EMPIRICAL LEGAL STUD. 811, 811 (2010) (describing federal class action settlements identified in a comprehensive study).

(18) See Lynn Baker et al., Setting Attorneys' Fees in Securities Class Actions: An Empirical Assessment, 66 VAND. L. REV. 1677, 1691 (2013) [hereinafter Baker et al., Attorneys' Fees] (analyzing fee awards based on a systematic investigation of federal securities class actions); see also Lynn Baker et al., Is the Price Right? An Empirical Study of Fee-Setting in Securities Class Actions, 115 COLUM. L. REV. 1371 (2015) [hereinafter Baker et al., Is the Price Right?] (discussing factors influencing attorney fee awards from the years 2007 through 2012).

(19) There is a voluminous literature on Rule 23 class actions. Although much of the literature is limited to doctrinal and normative analysis, a significant fraction of publications cite empirical data. Reviewing and synthesizing the empirical evidence presented in scholarly and public policy literature is beyond the scope of this Article.

(20) My discussion focuses on federal class actions. The problems of collecting data regarding state class actions are even more daunting than those described in Part V. A serious program of ongoing research on federal Rule 23 could provide a model for research on state class actions.

(21) Arthur Miller, Of Frankenstein Monsters and Shining Knights: Myth, Reality, and the "Class Action Problem", 92 HARV. L. REV. 664, 664 (1979).

(22) Federal court caseload statistics have been sporadically reported since 1801, when President Thomas Jefferson asked Secretary of State James Madison to send a report of federal court business to the U. S. Congress. See History of Federal Caseload Reporting, FED. JUDICIAL CTR., []. Over the next 70 years, Congress periodically requested selective statistics, apparently in an effort to inform specific policy debates. Id. The 1870 Act establishing the Department of Justice ordered the U.S. Attorney General to submit annual reports on the business of the federal courts to Congress. Id. The initial reports focused exclusively on cases involving the U.S. as a party but in 1873 Congress amended the reporting mandate to include private litigation. Id. A modern system of statistical reporting was established in the 1930s and periodically revised in subsequent years in response to recommendations from various special commissions and advisory groups. Id. The current reporting system, now relying on electronic case filing, dates back to the late 1990s. Id. For an early discussion of the usefulness of court statistics for decisionmaking, see Will Shafroth, Judicial Statistics, 13 LAW & CONTEMP. PROBS. 200, 200 (1948). For discussion of empirical legal studies in the first half of the twentieth century, see Herbert Kritzer, Empirical Legal Studies Before 1940: A Bibliographic Essay, 6 J. EMPIRICAL LEGAL STUD. 925, 925 (2009).

(23) Federal court management statistics currently indicate whether cases were terminated before pretrial; during pretrial; or during or after trial. See, e.g., Federal Court Management Statistics, December 2016, ADMIN. OFFICE OF THE U.S. COURTS, [].

(24) Median time to disposition is currently reported for civil cases, by district and case type. See, e.g., id.

(25) According to John Rabiej of the Duke Law School Center for Judicial Studies:
In 2014, there [were] 117,647 non-asbestos cases in MDLs representing
36% of the total U.S. pending civil caseload. Excluding prisoner and
social security actions from the U.S. pending civil caseload, which
typically (though not always) take little time of article III judges,
the 117,647 non-asbestos cases in MDLs represent about 45% of the U.S.

E-mail from John Rabiej, Dir. of the Ctr. for Judicial Studies, Duke Law School, to author (Aug. 29, 2014) (on file with author). Some MDLs include substantial numbers of class actions. The propensity of the Judicial Panel on Multidistrict Litigation to grant MDL status has varied over time, which may complicate efforts to count class actions. See Deborah Hensler, The Role of Multi-Districting in Mass Tort Litigation: An Empirical Investigation, 31 SETON HALL L. REV. 883, 885-86 (2001) (observing that "mass torts arose in the U.S. in an era when courts generally declined to certify personal injury class actions"); Hon. John Heyburn & Francis McGovern, Evaluating and Improving the MDL Process, LITIGATION, Spring 2012, at 30 [] ("For many years, the panel regularly granted more than 75 percent of all 1407 motions. During the last two years, that percentage has dropped to about 55 percent."); Amanda Bronstad, Multidistrict Litigation Panel's Increasing Denials Reflect Heightened Scrutiny, NAT'LLJ., July 18, 2016 (showing a trend of increasing denials from 2006 through 2015).

(26) Merely measuring trends is not sufficient for analyzing the consequences of doctrinal (or other) changes. See Jonah Gelbach, Locking the Doors to Discovery? Assessing the Effects of Twombly and Iqbal on Access to Discovery, 121 YALE L.J. 2270, 2295 (2012) (analyzing the effects of changes in pleading standards); David Engstrom, The Twiqbal Puzzle and Empirical Study of Civil Procedure, 65 STAN. L. REV. 1203, 1213 (2013) (highlighting the challenges of assessing the consequences of procedural change). However, trend data provide a first indication of what is changing.

(27) See, e.g., Federal Courts in Two Top Hellholes Issue Refreshingly Reasonable Class-Action Decisions, JUDICIALHELLHOLES.ORG (July 19, 2016), [] (reporting two court decisions in which federal judges reduced plaintiffs' attorneys' fees, rejected several class claims, and "decided that a class action may be decertified after a jury returns a verdict but before the court enters a final judgment."). The "hellhole" designation has been awarded both to districts that are claimed to certify an unusual number of class actions and districts whose judges are claimed to certify non-meritorious class actions and approve unreasonable settlements. See CONGRESS WATCH, PUB. CITIZEN, CLASS ACTION "JUDICIAL HELLHOLES": EMPIRICAL EVIDENCE IS LACKING, (2005), []. The "hellhole" project, begun in the 2000s and sponsored by the American Tort Reform Association, focuses primarily on state court litigation, including class actions. Id. Its reports were used successfully in lobbying Congress to pass the Class Action Fairness Act in 2005. Id.

(28) The Supreme Court has referred to this allegation in several recent decisions, apparently accepting its veracity. See, e.g., Amgen Inc. v. Conn. Ret. Plans & Trust Funds, 133 S. Ct. 1184, 1200 (2013) (citing AT&T Mobility LLC v. Concepcion, 179 L. Ed. 2d 742, 757 (U.S. 2011)). If--as fragmentary data suggest--a large fraction of class actions is certified after the strength of claims has been tested through pre-trial motion and discovery practice and plaintiffs have prevailed on these preliminary matters, it seems more likely that defendants have concluded that class-wide settlement is preferable to proceeding to trial on an individual or class basis, than that they have been terrified into submission. See generally Deborah Hensler, Can Private Class Actions Enforce Regulations? Do They? Should They? (highlighting class action certification timing), in COMPARATIVE LAW AND REGULATION: UNDERSTANDING THE GLOBAL REGULATORY PROCESS (Francesca Bignami & David Zaring eds., 2016).

(29) Decisions interpreting Rule 23 in a more restrictive fashion than previous holdings or otherwise restricting plaintiffs' ability to proceed in a class form include Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338 (2011) (vacating certification of a class of female employees, applying a heightened commonality standard) and Comcast Corp. v. Behrend, 133 S. Ct. 1426 (U.S. 2013) (vacating certification of an anti-trust class, on the grounds that the damage model did not fit the class definition).

(30) See Am. Pipe & Constr. Co. v. Utah, 414 U.S. 538, 553 (1973) ("A contrary rule allowing participation only by those potential members of the class who had earlier filed motions to intervene in the suit would deprive Rule 23 class actions of the efficiency and economy of litigation which is a principal purpose of the procedure."); FED. R. CIV. P. 23(b)(3) ("A class action may be maintained if .... the court finds that the questions of law or fact common to class members predominate over any questions affecting only individual members, and that a class action is superior to other available methods for fairly and efficiently adjudicating the controversy."); Linda S. Mullenix, Ending Class Actions as We Know Them: Rethinking the American Class Action, 64 EMORY L.J. 399, 421 (2014) ("A third rationale in support of the class action rule posits that class action procedures enhance judicial efficiency and economy, which is largely a utilitarian justification for the rule."). Mullenix sought to debunk this and other "romantic" notions of the value of Rule 23 class actions. Id.

(31) As Mullenix and others have argued, aggregation's most significant consequence for judicial workload may be in facilitating litigation that would otherwise not occur. Mullenix, supra note 30, at 422. Whether that should be regarded as a positive or negative consequence is a question that requires both empirical and normative analysis.

(32) The recent Volkswagen MDL, for example, is reported to have consolidated 180 putative class actions and about 300 individual lawsuits. In re: Volkswagen "Clean Diesel" MDL, U.S. DIST. CT. FOR THE N. DIST. OF CAL., [] [hereinafter Volkswagen MDL].

(33) The 9/11 recovery workers' personal injury litigation managed by Judge Alvin Hellerstein is an example of this. See Allen Hellerstein, James Henderson & Aaron Twerski, Managerial Judging: The 9/11 Responders' Tort Litigation, 98 CORNELL L. REV. 127 (2012). Judge Hellerstein chose not to rule on the Rule 23 motion filed in that litigation apparently in the belief that this would facilitate a global settlement. See id. at 168 ("The court had empowered the parties to reach a global settlement and had positioned itself to act as an impartial guarantor of the fairness of any settlement that the parties might reach.").

(34) EMERY LEE III & THOMAS WILLGING, FED. JUDICIAL CTR., IMPACT OF THE CLASS ACTION FAIRNESS ACT ON THE FEDERAL COURTS: PRELIMINARY FINDINGS FROM PHASE TWO'S PRE-CAFA SAMPLE OF DIVERSITY CLASS ACTIONS 1 (2008) (reporting preliminary findings measuring CAFA's impact on litigation activity and judicial rulings as well as case characteristics of class actions in the two years preceding CAFA's effective date).

(35) FED. JUDICIAL CTR., History of Federal Caseload Reporting, supra note 22, at 4. Until the early 2000s, weights were calculated based on judges filling out time sheets (sometimes termed "diaries"). PATRICIA LOMBARD & CAROL KRAFKA, FED. JUDICIAL CTR., 2003-2004 DISTRICT COURT CASE-WEIGHTING STUDY: FINAL REPORT TO THE SUBCOMMITTEE ON JUDICIAL STATISTICS OF THE COMMITTEE ON JUDICIAL RESOURCES OF THE JUDICIAL CONFERENCE OF THE UNITED STATES 23 (2005) [hereinafter CASE-WEIGHTING FINAL REPORT]. In 2004, after deliberation with representatives of the judiciary and extensive research and analysis by Federal Judicial Center staff, the Judicial Conference adopted a new approach to calculating weights, based on event analysis. Id. at 1-2. In brief, the method involved collecting docket data on the frequency distribution of different types of case events--e.g. hearings--by case type and estimating models of average judicial effort per case using these data. Id. This approach relieved judges of the burden of filling out time sheets and arguably took greater account of the differences in the ways judges handle different types of cases. Id. An attractive feature of this approach is said to be that it is easy to adjust weights over time to reflect differences in judicial case management strategies. Id. (comparing the different strategies). The case weights were updated in 2016, based on new research by the Federal Judicial Center. E-mail from Carol Krafka, Fed. Jud. Ctr., to Judith Resnik, Professor, Yale Law Sch. (Oct. 5, 2016) (on file with author). However, the most recent research did not ask judges to complete time sheets; rather, several hundred judges were surveyed and asked to estimate their time associated with different case types. Id.

(36) CASE-WEIGHTING FINAL REPORT, supra note 35, in appendices.

(37) EMPIRICAL STUDY FINAL REPORT, supra note 15, at 7. The judicial time study was based on judges completing "diaries" recording time spent on individual cases. Id. at 7-8.

(38) Eisenberg, Miller, and Fitzpatrick analyze settlement and fee award data for the universe of federal class action settlements. Eisenberg & Miller, supra note 16; Fitzpatrick, supra note 17. Baker, Perino & Silver analyze data for securities class actions. Baker et al., Attorneys' Fees, supra note 18. Eisenberg, Miller, and Fitzpatrick relied on judicial decisions approving class action settlements and deciding fee awards. Eisenberg & Miller, supra note 16; Fitzpatrick, supra note 17. Baker, Perino and Silver's analysis is based instead on an extensive docket search for all securities class actions that were certified and settled from 2007-2012 that were included in the Stanford Securities Class Action Clearinghouse database, which is widely regarded as the most comprehensive database on securities class actions available. Baker et al., Attorneys' Fees, supra note 18. In addition to providing important new information on the fee award process in securities class actions their research demonstrates the value of analyses based on federal court electronic dockets.

(39) Baker et al., Attorneys' Fees, supra note 18, at 1396 (reporting that in securities class action cases attorneys ask for a higher percentage of funds from judges in districts with lower volumes of securities class actions and from individual judges with less experience managing class actions).

(40) Id. at 1417.

(41) Id. at 1418.

(42) Hensler et al. discuss the results of their attempt to obtain data on defendants' legal expenses for a small number of consumer and mass tort class actions. In three class actions where defendants were willing to provide these data, their expenses ranged from twenty percent of the amount awarded to plaintiff class counsel to an amount equal to the plaintiff class counsel award. HENSLER ET AL., CLASS ACTION DILEMMAS, supra note 6, at 441-42.

(43) See, e.g., Brian Fitzpatrick, Do Class Action Attorneys Make Too Little?, 158 U. PA. L. REV. 2043, 2047 (2011) (arguing that small-stakes class actions serve no compensation function but rather only deterrence); Myriam Gilles & Gary Friedman, Exploding the Class Action Agency Costs Myth: The Social Utility of Entrepreneurial Lawyers, 155 U. PA. L. REV. 103, 105 (2006) (indicating that the main purpose of these claims is deterrence).

(44) See generally David Rosenberg, Decoupling Deterrence and Compensation Functions in Mass Tort Class Actions for Future Loss, 88 VA. L. REV. 1871 (2002) (discussing deterrence claims generally and concluding that mandatory collectivization is necessary to optimize deterrence); David Rosenberg, Individual Justice and Collectivizing Risk-Based Claims in Mass-Exposure Cases, 71 N.Y.U. L. REV. 210 (1996) (concluding that deterrence is the strongest, if not only justification for these large tort law claims).

(45) See Mullenix, supra note 30, at 409 (identifying the compensatory function of class actions as a component of the "romantic narrative" about Rule 23).

(46) The Rule states:
If the (settlement) proposal would bind class members, the court may
approve it only after a hearing and on finding that it is "fair,
reasonable, and adequate." Although the "fair, reasonable and adequate"
standard arguably could refer to deterrence, the Advisory Committee's
Note to the 2003 Amendments to Rule 23(e) makes clear that the
reference is to the procedure's compensatory function.

FED. R. Civ. P. 23(e)(2).

(47) For example, criticism of "coupon settlements" is grounded in part on the perception that they offer no monetary benefit to class members. See generally J. Brendan Day, My Lawyer Went to Court and All I Got Was This Lousy Coupon! The Class Action Fairness Act's Inadequate Provision of Judicial Scrutiny Over Proposed Coupon Settlements, 38 SETON HALL L. REV. 1085 (2008). Critics view low "take-up rates" (the fraction of class members who come forward to claim the benefits for which they are eligible) as demonstrating that class actions do not accomplish their alleged compensation function. See, e.g., id.; Robert Klonoff & Mark Hermann, The Class Action Fairness Act: An Ill-Conceived Approach to Class Settlements, 80 TUL. L. REV. 1695 (2005-2006). Some critics view substituting cy pres remedies for monetary compensation to class members as an admission that the class action procedure is an ineffective compensation mechanism. See generally Martin Redish et al., Cy Pres Relief and the Pathologies of the Modern Class Action: An Empirical and Normative Analysis, 62 FLA. L. REV. 618 (2010).

In civil law jurisdictions, where private enforcement litigation is seen as illegitimate in most instances, "collective redress" is seen as the only appropriate purpose of class actions. Recommendation 2013/396, 2013 O.J. (L 206) 60 (EU), [] [hereinafter European Commission Principles on Collective Redress]; see also Deborah Hensler & Stefaan Voet, Class Actions Across the Atlantic: From Guarded Interest to European Policy (Nov. 2016) (unpublished manuscript) (on file with author).

(48) Procedural fairness ought also to be a concern in assessing compensation schemes. Here I focus only on performance with regard to outcomes. For a discussion of the challenges of designing compensation systems, see KENNETH FEINBERG, WHO GETS WHAT: FAIR COMPENSATION AFTER TRAGEDY AND FINANCIAL UPHEAVAL (2012) (describing the challenges of designing the 9/11 and other mass compensation programs); Deborah Hensler, Alternative Courts? Litigation-Induced Claims Resolution Facilities, 57 STAN. L. REV. 1429, 1432 (2005) (noting twelve key decisions that the designers of any compensation program face); Byron Stier, The Gulf Coast Claims Facility as Quasi-Public Fund: Transparency and Independence in Claim Administrator Compensation, 30 MISS. C. L. REV. 255, 258 (2011) (advising on how to best utilize and enhance the functioning of the quasi-public claims fund).

(49) See Baker et al., Is the Price Right?, supra note 18, at 1438 (observing that "at the onset of a litigation lawyers cannot predict settlement size with any degree of precision.").

(50) See, e.g., Memorandum to Advisory Committee, supra note 14; Eisenberg & Miller, supra note 16; Fitzpatrick, supra note 17. In the late 1990s study of class actions that I led, which included an intensive analysis of documents and interviews with attorneys in ten certified class actions, we found substantial differences between the approved settlement amounts and actual compensation paid to class members. HENSLER ET AL., CLASS ACTION DILEMMAS, supra note 6, at 458-59.

(51) Cy pres remedies and injunctive relief are not strictly speaking components of compensation but, as I discuss later, need to be taken into account when assessing the deterrent potential of class actions.

(52) See supra note 18 and accompanying text.

(53) See generally William Rubenstein & Nicholas Pace, Shedding Light on Outcomes of Class Actions, in CONFIDENTIALITY, TRANSPARENCY AND THE U.S. CIVIL JUSTICE SYSTEM (Joseph Doherty et al. eds., 2012) (highlighting the futility of the task). Analysis of post-settlement claims administrator data would also facilitate understanding of how claims administration design affects claiming rates. See Deborah Hensler, Reflections on Settlement Class Actions: Paper Prepared for the George Washington University Roundtable (Apr. 8, 2015) (unpublished paper) (on file with author). Kevin Clermont, Brian Fitzpatrick, Alexandra Lehav, Geoffrey Miller, Nicholas Pace, William Rubenstein, Charles Silver, Bryan Wolfman and I have submitted a formal proposal for a new Rule 23(1) on "Class Action Disbursement Disclosure," to the Advisory Committee on the Civil Rules that is currently considering Rule 23 amendments. To the best of my knowledge, the Committee has taken no action on our proposal. However, [section]1719 of the newly proposed Fairness in Class Action Litigation Act of 2017 calls for the collection and transmission of information on class action disbursements to the Federal Judicial Center. Fairness in Class Action Litigation and Furthering Asbestos Claim Transparency Act of 2017, H.R. 985, 115th Cong. [section]1719 (2017).

(54) Reviewing the vast theoretical literature on the difficulty of measuring litigation's effectiveness as a deterrent mechanism is beyond the scope of this Article. Instead I propose a commonsense agenda that would shed light on Rule 23's contribution to regulatory enforcement via private actions, although it would not yield a conclusive answer to the question of Rule 23's value as a deterrence mechanism.

(55) HENSLER ET AL., CLASS ACTION DILEMMAS, supra note 6, at 416-24.

(56) For a discussion of inferring merit from disposition patterns, see Hensler, supra note 25.

(57) It has become increasingly common for public policymakers to prevent access to the sort of information necessary to link data on the grounds that doing so would violate privacy protections. However, the names of litigating parties are a matter of public record. While regulatory agencies frequently decline to provide information on putative violations that they ultimately decide not to prosecute, basic descriptive information on cases that are criminally prosecuted or result in civil enforcement actions should also be made available.

(58) Hensler, supra note 25, discusses empirical studies attempting to link private and public enforcement data related to class actions.

(59) I am regularly asked by reporters for empirical data relevant to high-visibility class action lawsuits, Supreme Court decisions on class actions and regulatory agency and congressional decisionmaking related to class actions. I have to respond--to the astonishment of the reporters--that neither I nor anyone else has readily available data. One Bloomberg reporter who talked with me frequently in the period prior to the Court issuing its decision in Wal-Mart v. Dukes asked me how could I "keep from tearing (my) hair out."

(60) The theoretical and empirical literature on patterns and determinants of claiming is voluminous and now includes research in many different countries. See William Felstiner et al., The Emergence and Transformation of Disputes: Naming, Blaming Claiming ..., 15 LAW & SOC'Y REV. 631 (1980-81) (providing a framework for studying the processes by which injurious experiences are or are not perceived, become grievances, and ultimately disputes); Herbert Kritzer, The Antecedents of Disputes: Complaining and Claiming 1 (Onati Socio-Legal Series, Nov. 6, 2011), [] (examining what is known regarding the emergence of grievances and their communication as complaints and claims in several countries); DEBORAH HENSLER ET AL., COMPENSATION FOR ACCIDENTAL INJURIES IN THE UNITED STATES (1991) (broadly identifying some of these patterns for the United States).

(61) History of Federal Caseload Reporting, FED. JUDICIAL CTR., supra note 22. At one time the case record data were recorded painstakingly by court clerks on individual case data cards that were sent to the AO. Id. Now they are extracted from the electronic files. Id.

(62) Federal Court Cases: Integrated Database Series, INTERUNIVERSITY CONSORTIUM POLITICAL & SOC. RES., [].

(63) U.S. Courts, Civil Cover Sheet, JS-44, revised 8/16, Section VI, [] (emphasis in original).

(64) I am still researching when the AO first reported class action statistics in its annual reports and when it ceased doing so.

(65) This was also the period when Congress was debating the Private Securities Litigation Reform Act. The intensive study that followed the Federal Judicial Center's preliminary analysis highlighted securities class actions. See EMPIRICAL STUDY FINAL REPORT, supra note 15.

(66) Memorandum to Advisory Committee, supra note 14.

(67) Notwithstanding concern among Federal Judicial Center and AO staff about the unreliability of the class action indicator derived from the civil cover sheet, there has been no systematic attempt to analyze the reasons why the indicator is error-prone. Such an analysis might yield information on how to revise the cover sheet to increase reliability.


(69) This discussion draws in part on discussions at the University of Pennsylvania Law School's workshop, Increasing Access to Federal Court Data, October 9, 2015, organized by Jonah Gelbach, and discussions at the Federal Court Civil Data Project Roundtable co-sponsored by the ABA Standing Committee on the American Judicial System and the Duke Law Center for Judicial Studies, held in Washington, D.C., on October 23, 2015. I am grateful to Professors Gelbach, David Engstrom, and Herbert Kritzer and to other participants at these meetings for sharing their experiences accessing federal court data.

(70) Marc Galanter, The Vanishing Trial: An Examination of Trials and Related Matter in the Federal and State Courts, 1 J. EMPIRICAL LEGAL STUD. 459 (2004).

(71) Marc Galanter, The Day After the Litigation Explosion, 46 MD. L. REV. 3, 15-28 (1986).

(72) The federal judiciary is now completing the rollout of its new case management and electronic case filing system, dubbed "NEXTGEN." Incorporating an improved indicator of class action litigation does not appear to have been on the agenda of the NEXTGEN system designers.

(73) Federal Court Cases: Integrated Database, 1970-2000, NATIONAL ARCHIVE OF CRIMINAL JUSTICE DATA, [].

(74) I have yet to be able to determine when the database was deposited in the ICPSR archive. However, at the time of archiving, the entire database back to 1971 was deposited.

(75) See supra footnotes 62-63 and accompanying text.

(76) Every institution that receives federal funding is required to establish an Institutional Review Board and a process to review proposed research involving human subjects, typically termed "IRB Review." See 45 C.F.R. [section] 46.101 (2012) (setting forth the Institutional Review Board process and requirements).

(77) E-mail from Margaret Levenstein, Dir., Inter-University Consortium for Political and Soc. Sci. Research, to Judith Resnik, Professor, Yale Law Sch. (Dec. 29, 2016) (on file with author).

(78) PACER: PUBLIC ACCESS TO COURT ELECTRONIC RECORDS,[]. Although there is a fee schedule for retrieving information from PACER, any member of the public can view case information or documents from a courthouse using public access terminals. See PACER, ELECTRONIC PUBLIC ACCESS FEE SCHEDULE (2013), [] [hereinafter PACER FEE SCHEDULE].

(79) E-mail from Adam Bailey, Manager, Stanford University Institutional Review Bd. (nonmedical research), to author (Sept. 8, 2016) (on file with author). As a member of the Stanford University Institutional Review Board, I am required to be familiar with the human subjects' protection regulations and their application.

(80) Integrated Database (IDB), FED. JUDICIAL CTR., []. Discussions about making the database available through the Federal Judicial Center were long ongoing and the decision likely resulted from the joint efforts of the Federal Judicial Center researchers and academics interested in analyzing the data. See supra note 69 and accompanying text.

(81) PACER FEE SCHEDULE, supra note 78.

(82) Jonah Gelbach, Professor, Univ. of Pa. Law Sch., Proposal to the National Science Foundation for a Workshop on Increasing Access to Federal Court Data (unpublished manuscript) (on file with author).

(83) For example, the ongoing litigation arising out of the Volkswagen "clean diesel" misrepresentations includes consumer and shareholder class actions, state attorney general enforcement actions, and criminal investigations by the U.S. Department of Justice. Volkswagen MDL, supra note 32.

(84) In 2015, a group of law faculty submitted a proposed new provision for Rule 23 calling for public disclosure of such information. See supra note 53.

(85) Aside from its usefulness for research, making such data available publicly would also promote more informed reporting of class action outcomes by journalists.

Please Note: Illustration(s) are not available due to copyright restrictions.
COPYRIGHT 2017 University of Pennsylvania, Law School
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2017 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Special Issue on Class Actions
Author:Hensler, Deborah R.
Publication:University of Pennsylvania Law Review
Date:Jun 1, 2017

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters