USING COURT RECORDS FOR RESEARCH,
TEACHING, AND POLICYMAKING: THE CIVIL
RIGHTS LITIGATION CLEARINGHOUSE
Margo Schlanger* and Denise Lieberman**
INTRODUCTION
The National Archives and Records Administration (NARA) is, wisely,
planning the future of its enormous collection of relatively recent court records.
The pertinent regulation, a “records disposition schedule” first issued in 1995 by
the Judicial Conference of the United States in consultation with NARA,
commits the Archives to keeping, permanently, all case files dated 1969 or
earlier; all case files dated 1970 or later in which a trial was held, and “any civil
case file which NARA has determined in consultation with court officials to have
historical value.” Other files may be destroyed 20 years after they enter the
federal repository.1 The question before NARA, then, is whether to destroy
nearly all of the records for cases terminated after 1969, keeping litigation files
from those years only for those few cases in which adjudication was by trial (5%
or less of the docket since 1984),2 or whether to deem some or all non-trial cases
as having “historical value” and preserve them for future research and use.
The regulation appears to embrace a view that historical court records are
usually of only low or moderate value, insufficiently important to justify the
resource allocation needed to preserve them all. In this brief article, we hope to
demonstrate that this premise is incorrect. Court records have always been vital,
and irreplaceable, sources for historical research. Made more accessible, as

*

Professor of Law, Washington University in St. Louis; J.D. 1993, Yale Law School. For helpful
comments on the presentation that formed the basis of this paper I owe thanks to the participants of
the symposium, NARA Federal Civil Court Records: New Territory for Empirical, Historical, and
Legal Research, at University of Missouri in Kansas City. Particular thanks to many of NARA’s
archivists, to Jim Woodward, Clerk of Court, U.S. District Court for the Eastern District of
Missouri, for answering many of my questions about records retention, and to (as always) Sam
Bagenstos.
**
Lecturer in Political Science and Adjunct Professor of Law, Washington University in St. Louis;
former legal director, American Civil Liberties Union of Eastern Missouri; J.D. 1995, University of
Pittsburgh School of Law.
Copyright © 2006 by Margo Schlanger. Permission is granted to copy for free or at-cost
distribution to students enrolled in a class.
1
The 1995 version was: Schedule for the Disposition of the Records of United States District
Courts, Territorial District Courts, Court of Claims, Court of International Trade, United States
Claims Court, and the Special Court, Regional Rail Reorganization Act of 1973, at 22, Information
Resources Management Manual Chapter 17, XIII GUIDE TO JUDICIARY POLICIES AND PROCEDURES
(Transmittal 3, Nov. 6, 1995). The current version, identical in pertinent respects, is
ADMINISTRATIVE OFFICE OF THE UNITED STATES COURTS, Vol. I, THE GUIDE TO JUDICIARY POLICIES
AND PROCEDURES, Records Disposition Schedule 2, Chapter 12, Part A, Section 15 (last modified
Nov. 18, 2002). Both versions are on file with the authors and are posted in this article’s
Appendix, see http://schlanger.wustl.edu, under “publications.”
2
See, e.g., Marc Galanter, The Vanishing Trial: An Examination of Trials and Related Matters in
Federal and State Courts, 1 J. EMP. LEG. STUD. 459, 533-34 (tbl. A-2) (2004).

154

UMKC LAW REVIEW

[Vol. 75:153

modern technology now allows, court records can become central sources for
much broader use for legal researchers; for historians, political scientists,
sociologists, and anthropologists; for students and teachers; and for advocates
and policymakers. In Part I of this Article, we present a model for such use—the
Civil Rights Litigation Clearinghouse, a new web-based resource built primarily
around digitized court records, sponsored by Washington University in St.
Louis.3 Accessible at http://clearinghouse.wustl.edu, the Clearinghouse collects,
indexes, and makes publicly available for research and observation a growing
universe of civil rights cases and the settlements and court orders those cases
have produced, which regulate government and private entities in myriad
important ways. We are working with NARA to include dockets and other court
documents from the National Archives that might well be destroyed, someday,
under the current document retention policy. The Civil Rights Litigation
Clearinghouse illustrates the new opportunity to make court records, including
NARA’s court records, ever more useful to—and therefore used by—a broad
array of people. It would be a great shame to pretermit such uses by illconceived archive destruction.
It is the absence of good alternatives that makes court records so important.
In Part II, we explain why other types of records cannot substitute for litigation
files, discussing the shortcomings of other sources of information about
litigation. Part III concludes by arguing, very briefly, that trial status is an
inappropriate proxy for the importance of a given piece of litigation. Trials are
simply dispositions that turn on facts, as opposed to dispositions that turn on law
(non-trial adjudication), or negotiation; trials are neither more nor less important
than dismissals, summary judgments, settlements, and all the other methods of
ending litigation. Moreover, trials are growing ever rarer, putting increasing
pressure on the poor proxy.4 We therefore welcome NARA’s invitation to
comment on the best way for it to proceed and urge that that the current policy be
changed altogether, or at least that the regulatory provision for keeping non-trial
records for cases with historical value be applied extremely liberally, to entire
case categories (e.g., civil rights of all kinds, including prison conditions; torts;
environmental law; and many others).
I. THE CIVIL RIGHTS LITIGATION CLEARINGHOUSE: NEWLY
ACCESSIBLE COURT RECORDS
A. The Importance of Civil Rights Injunctive Litigation
Civil rights injunctions have been of tremendous importance in this country
since 1955, when the Supreme Court’s second opinion in Brown v. Board of

3

The project is multidisciplinary and is supported by Washington University’s Law School, the
Law School’s Center for Empirical Research in the Law and its Center for Interdisciplinary
Studies, as well as by the Faculty of Arts and Sciences’ American Culture Studies and Legal
Studies programs.
4
See infra note 52 and accompanying text.

2006]

CIVIL RIGHTS LITIGATION CLEARINGHOUSE

155

Education created the conditions for the first modern such orders.5 It was Brown
II that instructed district courts to “take such proceedings and enter such orders
and decrees consistent with this opinion as are necessary and proper to admit to
public schools on a racially nondiscriminatory basis with all deliberate speed the
parties to these cases.”6 The resulting long-term litigation-related oversight of
errant institutions is the essence of modern injunctive practice. The settlements
and court orders entered in civil rights cases over the past fifty years have
transformed a huge number of governmental and private institutions including
schools,7 voting districts,8 jails and prisons,9 housing authorities,10 child welfare
systems,11 employers,12 restaurants and other public accommodations,13 and so
on. This injunctive litigation model is prevalent in other types of cases as well:
environmental litigation14 and some kinds of union reform,15 for example. In
these and other arenas, injunctive civil rights cases have closed some institutions
and opened others, dominated budget politics on occasion, become models for

5

Brown v. Bd. of Educ. of Topeka, 349 U.S. 294 (1955).
Id. at 301.
7
For a list of school desegregation cases, see, e.g., American Communities Project, Brown
University, The State of Public School Integration: Desegregation Court Cases and School
Demographic Data, http://www.s4.brown.edu/schoolsegregation/desegregationdata.htm (last visited
July 16, 2006).
8
For a list of Voting Rights cases see, e.g., Voting Rights Initiative, Documenting Discrimination:
Judicial Findings under Section 2 of the Voting Rights Act, available at:
http://sitemaker.umich.edu/votingrights/final_report (last visited Sept. 9, 2006). By 2007, more
comprehensive data will be available at http://clearinghouse.wustl.edu.
9
See Margo Schlanger, Civil Rights Injunctions Over Time: A Case Study of Jail and Prison Court
Orders, forthcoming 81 N.Y.U. L. REV. 550 (2006). For information about hundreds of jail and
prison cases, see http://clearinghouse.wustl.edu.
10
See, e.g., Susan J. Popkin, George Galster, Kenneth Temkin, Carla Herbig, Diane K. Levy, &
Elise Richer, Baseline Assessment of Public Housing Desegregation Cases: Cross-site Report
(April 2000) available at http://www.huduser.org/publications/pubasst/baseline.html (listing cases)
(last visited Sept. 9, 2006). See also http://clearinghouse.wustl.edu.
11
See, e.g., National Center for Youth Law, Foster Care Reform Litigation Docket (2006),
available at http://www.youthlaw.org/fileadmin/ncyl/youthlaw/publications/fcrldocket06.pdf (last
visited Sept. 9, 2006).
12
The EEOC’s website includes descriptions and copies of many class action court orders. See, e.g,
Joint Motion for Entry of Consent Decree, EEOC v. Mitsubishi Motor Manufacturing of America,
96-1192 (C.D. Ill. June 23, 1998), available at http://www.eeoc.gov/policy/docs/mmma.html (last
visited Sept. 9, 2006). We hope that these cases will be available at the Clearinghouse in 2007, at
http://clearinghouse.wustl.edu.
13
For example, for a list of the consent decrees and other settlements reached by the U.S.
Department of Justice with various operators of public accommodations under the Americans with
Disabilities Act, see U.S. Dep’t of Justice, ADA Settlements and Consent Agreements, available at
http://www.ada.gov/settlemt.htm (last visited Sept. 9, 2006).
14
See, e.g., Peter A. Appel, Intervention In Public Law Litigation: The Environmental Paradigm,
78 WASH. U. L.Q. 215 (2000).
15
See James B. Jacobs, Eileen Cunningham & Kimberly Friday, The RICO Trusteeships After
Twenty Years: A Progress Report, 19 LAB. LAW. 419 (2004) (describing cases); JAMES B. JACOBS,
MOBSTERS, UNIONS & FEDS: THE MAFIA AND THE AMERICAN LABOR MOVEMENT (2005).
6

156

UMKC LAW REVIEW

[Vol. 75:153

statutory interventions, and generally regulated governmental and private
practices. Thousands of such cases have been filed over the past fifty years and
new cases are filed all the time; hundreds, old and new, are ongoing and remain
influential. Because relief in injunctive cases is directed at changing government
or corporate policies, the impact of these cases extends far beyond the parties to
the lawsuit.
But although these are extremely important cases, information about them
has always been exceedingly hard to come by. Only a portion of the cases have
ever been the subject of published judicial opinions. And even when judges have
written opinions, and allowed those opinions to be published in print or on-line
sources, such sources rarely tell the entire story of a litigation. For an example,
consider Brown v. Board of Education itself. Brown was, of course, the litigation
that produced perhaps the best known Supreme Court opinions of the 20th
century. In the district court for the District of Kansas, Brown lasted nearly fifty
years: It began in 1951, when lawyers for the NAACP Legal Defense Fund
featured Oliver Brown’s complaint on behalf of his daughter Linda in a class
action complaint, and litigation continued until 1999. And, somewhat unusually,
the Brown litigation was the subject of eleven reported opinions—seven in the
district court,16 two in the court of appeals,17 and two in the Supreme Court.18 Yet
those opinions merely begin to inform any one interested in the life of the
litigation. They omit, for example, the actual desegregation plan entered on
remand from the Supreme Court in 1955. Indeed, the district court’s published
discussion of that plan, that same year, states: “No useful purpose would be
accomplished by setting out the plan in detail.”19 Nor is there any published
opinion detailing the district court’s determination that the school district had by
1961 substantially complied with the 1955 plan.20 The list of features of the
Brown litigation invisible from the accessible opinions extends further, but the
point is made: Even in this uniquely high profile and important case, those
judicial opinions accessible via print or on-line publication do not capture close
to the entirety of what interested observers would want to know about the

16

Brown v. Bd. of Educ. of Topeka, 98 F. Supp. 797 (D. Kan. 1951); 139 F. Supp. 468 (D. Kan.
1955); 84 F.R.D. 383 (D. Kan. 1979); 671 F. Supp. 1290 (D. Kan. 1987); No. 51-316-RDR, 1994
W.L. 114001 (D. Kan. March 30, 1994); Brown v. Unified Sch. Dist. No. 501, 878 F. Supp. 1430
(D. Kan. Feb 21, 1995); 56 F. Supp.2d 1212 (D. Kan. 1999).
17
Brown v. Bd. of Educ. of Topeka, 892 F.2d 851 (10th Cir. 1989); 978 F.2d 585 (10th Cir. 1992),
cert. denied, 509 U.S. 903 (1993).
18
Brown v. Bd. of Educ. of Topeka, 347 U.S. 483 (1954); 349 U.S. 294 (1955). In addition, in Bd.
of Educ. of Topeka v. Brown, 503 U.S. 978 (1992), the Court vacated the 10th Circuit’s 1989
opinion cited supra, in light of Freeman v. Pitts, 503 U.S. 467 (1992), and Bd. of Educ. of
Oklahoma City Pub. Sch. v. Dowell, 498 U.S. 237 (1991), and remanded for further consideration.
19
139 F. Supp. at 469 (“There are a number of respects in which we feel that the plan does not
constitute full compliance with the mandate of the Supreme Court, but that mandate implies that
some time will be required to bring that about. The elements that we feel do not constitute full
compliance are mostly of a minor nature but since this is not a final decree no useful purpose would
be served by setting them out herein.”).
20
See 84 F.R.D. at 389 (mentioning 1961 finding of compliance).

2006]

CIVIL RIGHTS LITIGATION CLEARINGHOUSE

157

litigation.21 This dearth of knowledge is even starker in most cases, and is nearly
absolute for the many cases in which no judicial opinions have been published
(usually either because the case’s resolution occurred by settlement rather than
adjudication or because the judge decided against publication). For most
injunctive cases, then, even obtaining copies of injunctive orders that remain in
effect can be a major challenge; getting hold of less central documents or of
documentation in cases since dismissed can be nearly impossible for non-experts
or the unfunded.
This information void has interfered with the development of public policy.
As Congressman Roy Blunt recently explained in a forum discussing legislation
he is sponsoring22 that would make it easier for state and local governmental
defendants to terminate civil rights decrees, we need “some future sense of how
many decrees there are even out there. There’s no record, there’s no
clearinghouse, no way to look and see how many of these decrees are out
there.”23 Informational scarcity makes good policymaking and advocacy more
difficult, as officials, lawyers, and activists are forced to spend time finding out
by word of mouth who has encountered issues similar to the ones they confront.
And, similarly, ordinary citizens are unable to uncover the legal regime in which
the governmental institutions that affect them are situated. The same problems
undermine good scholarship, as well, when scholars are forced to study only
unusual cases because the ordinary ones are so elusive.
B. The Civil Rights Litigation Clearinghouse: Introduction
The Civil Rights Litigation Clearinghouse addresses this knowledge deficit
by posting information about many types of civil rights cases, along with
documents from each case. The first phase of the Clearinghouse includes cases
of several types: conditions of confinement cases involving jails, prisons, and
juvenile detention and corrections facilities; and cases involving mental
retardation and mental health facilities, and nursing homes. In addition, we have
posted a few chosen cases even though we do not yet have the others in the
category: Brown is there, for example,24 as are a few recent cases, like Lozano v.
City of Hazleton,25 a litigation challenging a city ordinance that penalizes those

21

The district docket in Brown from the case’s beginning until 1990 is available as document SDKS-0001-9001 at http://clearinghouse.wustl.edu; the docket from 1993 to the case’s close in 1999
is available via PACER, or as document SD-KS-0001-9000. Many other documents from the
Brown file are also available at the Clearinghouse.
22
See Federal Consent Decree Fairness Act, S. 489, 109th Cong. (2005) (introduced March 1, 2005
by Senator Lamar Alexander (R-TN)); Federal Consent Decree Fairness Act, H.R. 1229, 109th
Cong. (2005) (introduced March 1, 2005 by Representative Roy Blunt (R-MO)).
23
For a transcript of the June 2005 event, entitled “Government by Consent Decree” and sponsored
by the American Enterprise Institute, see http://www.aei.org/events/filter.all,eventID.1078/
transcript.asp (last visited July 16, 2006).
24
See SD-KS-0001 at http://clearinghouse.wustl.edu.
25
See IM-PA-0001 at http://clearinghouse.wustl.edu.

158

UMKC LAW REVIEW

[Vol. 75:153

who do business with illegal aliens. The Clearinghouse’s cases currently
available number over a thousand.
The first challenge for this phase was identifying the relevant cases. The
difficulty of such identification is, we believe, one of the major reasons court
records are so rarely used. In any event, our goal is that this part of the archive
should include every litigation with injunctive relief or injunctive-like settlement
in the relevant case categories, current and historical. To that end, in our search
for cases to include we have supplemented traditional legal research by working
with a number of advocacy and governmental organizations that maintain lists
and brief banks. We believe that we have succeeded to the extent that the
Clearinghouse collection reflects the most comprehensive documentation ever
available of injunctive civil rights litigation. In addition, the Clearinghouse
solicits its users to propose new case categories, cases, and documents for the
collection, via the web.
The next challenge was to gather up information and documentation of the
cases. At least for federal cases, court records are available but sometimes
difficult to gather; they are usually in the relevant federal courthouse but for the
older cases, in one of NARA’s regional facilities. Copying documents from
courthouses is prohibitively expensive, so we have not generally done that.
Rather, we have obtained the documents from a wide array of alternative sources
scattered across the country. Some are from the U.S Department of Justice,
others from the National Archives, still others from various advocacy
organizations, as well as individual attorneys and litigators on cases. A great
many are from the federal court system’s web-based docketing system, known as
PACER (“Public Access to Court Electronic Records”). For each case, we have
tried to assemble key filings and orders if those documents are available—in
particular, the complaints, the settlements and court orders, and, most crucially,
the docket sheets that mark the progress of the litigation in trial court.
Importantly, the Clearinghouse recognizes that the story of a litigation does not
end with the issuance of a liability finding, or even a decree. Injunctive relief
includes continued court jurisdiction and often a great deal of post-decree action,
oversight, and/or litigation.26 Our document-gathering efforts therefore extend
through the entire course of proceedings, including any post-decretal period. In
total, the Clearinghouse’s first thousand cases will post over five thousand
documents pulled from court records. The Clearinghouse thus makes webaccessible a rich and growing documentary archive of the progression of civil
rights litigation in this country.
The archival work is supplemented with four other types of information.
First, the Clearinghouse provides citations and links to several thousand
additional opinions rendered in its cases and available on Westlaw or Lexis.
Second, case summaries, mostly written by (supervised) law students, describe as
much of the case’s procedural history as we can discover; like the linked
documents, the summaries cover post-decree as well as pre-decree proceedings.

26

See, e.g., M. KAY HARRIS & DUDLEY P. SPILLER, JR., AFTER DECISION: IMPLEMENTATION OF
JUDICIAL DECREES IN CORRECTIONAL SETTINGS (1977).

2006]

CIVIL RIGHTS LITIGATION CLEARINGHOUSE

159

Third, case pages include citations to relevant case studies previously published.
Finally, and perhaps most important, all the cases and documents are indexed,
full-text searchable, and coded with searchable metadata. The available indices
include: case name, case type, cause of action, source of relief (litigated or
consent), case and institution location (city, county, state, district, circuit),
various issues, federal or state court, name of judges and attorneys, name and
type of subject institution, and various dates (filing, order entry, closing).
The Clearinghouse is scheduled to launch in October 2006, about the same
time as this article will appear. By that time, we will be well underway in a
second phase, as well. In the next year or so, we will add EEOC pattern or
practice litigation and settlements; several thousand voting rights cases; and
perhaps some portion of the large school desegregation category. In subsequent
phases over the next months and years, we will work to add an easy-to-use
quantitative component and new case categories (fair employment,
deinstitutionalization litigation, disability public accommodations, school equity
and finance, immigration class actions, police misconduct litigation, child
welfare, public defender reform, and many more).
C. The Civil Rights Litigation Clearinghouse: New Possibilities
for Research
The Clearinghouse thus makes two major contributions to the research
infrastructure in the area of American civil rights. First, the Clearinghouse
solves the basic problem in court record research that even for a particular case
about which a researcher is aware, it can be extremely difficult simply to track
down information. Ordinarily, to obtain relevant records using either PACER or
the paper file in the federal courthouses or in NARA’s regional archive requires
knowledge either of the case name or the docket number—both often impossible
to obtain except from the parties. The Clearinghouse instead allows users to
browse (by case type or state), to search using any combination of several dozen
fields (say, jail cases in Bowling Green, or cases in which Frank Johnson was a
judge), or to search the full text of the included documents.
The second research problem is that obtaining documents requires often
impracticable trips to their repositories and payment of a per-page fee. But for
each included case, the Clearinghouse makes information more and more cheaply
accessible. Once users find cases (and documents and summaries), the
Clearinghouse allows them, without fee, to print or download any document,
print or download a list of cases; and even save a search, case, or document in a
personal briefcase. (Users can also register for updates of various sorts.)
The Clearinghouse thus not only makes it possible to find particular cases,
but allows new kinds of research by users whose interests are legal, historical or
political, whether their choice of analytic approach is qualitative or quantitative:
Policymakers interested in any of the substantive areas covered
(corrections, education, housing policy, etc.) can use the Clearinghouse to access
plans and reforms that have been implemented in other jurisdictions. Court
decrees have long been a source of regulatory innovation—and even
organizations and jurisdictions that have not been sued often want to know what

160

UMKC LAW REVIEW

[Vol. 75:153

those decrees provide, to consider whether similar practices and procedures
might be useful for them. But the decrees have not previously been available.
The Clearinghouse will correct this situation. As Peter Martin, former Dean of
Cornell Law School and one of the founders of Cornell’s Legal Information
Institute, has observed about digital information technology more generally, a
broad searchable database can facilitate “less rigid, compartmentalized, and
hierarchical government measures that also facilitate more richly informed
discretion.”27
For both plaintiffs’ and defendants’ attorneys, as well as judges, the
Clearinghouse can be an important tool for designing remedies. A lawyer or
judge might use the Clearinghouse to pull up all the jail cases involving fire
safety, for example, gaining insight into the variety of remedial avenues others
have followed.
The Clearinghouse allows researchers or analysts interested in civil rights,
government operations, or any of the substantive areas covered to understand
better the interplay of law and public policy. Instead of searching through
reported opinions, knowing that such opinions are unrepresentative but unable to
get access to the “law in action,” researchers will be able to use the
Clearinghouse to understand the settlement and/or post-decree activity in civil
rights cases, to look at the course of the litigation involved, and to obtain records
and do searches unavailable elsewhere.
For law professors in particular, the Clearinghouse will be a tool to teach
about complex litigation with its multiple parties and phases, many decision
points, and long life. It will also be useful in substantive law school classes,
providing examples of complaints, briefs, opinions, and decrees on many topics.
For students, whether in high school, college, graduate school, or law
school, the Clearinghouse will offer an entry point into important areas of civil
rights injunctive practice, which might otherwise form an impenetrable thicket.
By posting source documents along with summaries that explain their relevance,
it creates ideal fodder for student inquiry.
In short, the Civil Rights Litigation Clearinghouse demonstrates two things:
the usefulness of court records as sources, and the potential for broadening record
access for many different types of users, by using modern digitizing, database,
and Internet technology.
II. THE IMPORTANCE OF COURT RECORDS: HOW DO WE KNOW
ABOUT LITIGATION?
We hope that part I has demonstrated that court records are extremely
valuable, even though obstacles to their accessibility have prevented full use.
But could other sources, more easily maintained and disseminated, substitute?
After all, litigation creates voluminous and easily accessible records of other
types. The answer is clear: Each of the other sources, though far more easily

27

Peter W. Martin, The Internet: “Full and Unfettered Access” to Law—Some Implications, 26
N. KY. L. REV. 181, 199 (1999).

2006]

CIVIL RIGHTS LITIGATION CLEARINGHOUSE

161

accessible than court records have been, suffer from major flaws that undermine
its ability to substitute for court records, the gold standard for litigation research.
Consider the U.S. district courts, for example. Each year, the large page,
small type Federal Supplement and Federal Rules Decisions case reporters print
thousands of pages of “reported” judicial opinions; Westlaw and Lexis publish
thousands more “unreported” opinions electronically.28 Verdict and settlement
reporters publish a description of the litigated or settled outcomes of thousands of
cases. In addition, newspapers and other media outlets report on court
proceedings. And finally, the Administrative Office of the United States Courts
publishes annual statistical summaries,29 and makes available the individual case
data underlying those summaries.30 All these are, unlike court records, sources
that researchers and lawyers can find with relative ease; they are available in
libraries and on-line, indexed, archived, often text-searchable. Yet closer
examination demonstrates that this information is far from being either
comprehensive or representative.
A. The inadequacies of reported opinions
Begin with lawyers’ favorite source of information, the judicial opinion.
Judicial opinions may in some situations give their readers a great deal of
information about a litigation, but that is not their purpose. An opinion’s purpose
is to announce the reasoning underlying the judicial resolution of some issue in
the case. Thus, even when a judge writes an opinion and makes it available,
either by designating it for publication in the print and on-line versions of the
relevant court reporters or allowing it to be included in the Westlaw and Lexis
on-line databases, or (in the past several years) posting it on a court website,
opinions do not tell the entire story of a litigation. Opinions are, rather,
snapshots of the case at one particular point in time. That point may be a
particularly crucial one, but it may not. Reported opinions can concern the
especially important issue of liability, for example, but they can also be about
more peripheral matters such as attorneys’ fees, discovery disputes, and the like.
As discussed above in the Brown v. Board of Education litigation, extremely
important matters may be totally omitted from reported decisions.

28

As discussed below, a Westlaw search of all opinions in the “dct” database issued between
October 1, 2003 and September 31, 2004, pulled up nearly 28,000 opinions, of which over 7,300
were officially reported. (In many districts, yet thousands more opinions—often labeled “orders”
but including some limited legal discussion as well as the actual result—are apparently accessible
only in case files or, more recently, via the electronic docketing system.)
29
See, e.g., ADMINISTRATIVE OFFICE OF THE UNITED STATES COURTS, JUDICIAL BUSINESS OF THE
UNITED STATES COURTS, 2004 (2005), available at http://www.uscourts.gov/caseload2004/
contents.html (last visited Sept. 9, 2006).
30
See, e.g., Federal Judicial Center, Federal Court Cases: Integrated Data Base, 2004 [Computer
file], Inter-university Consortium for Political and Social Research Study No. 4348, available at
http://www.icpsr.umich.edu.

162

UMKC LAW REVIEW

[Vol. 75:153

Moreover, only a small percentage of cases have any library-accessible
opinions at all;31 the precise number varies with case type, geography, and other
factors.32 Indeed, even though Westlaw and Lexis have increased the accessibility
of some “unreported” opinions, observers estimate that only five to ten percent of
cases have any accessible opinion at all.33 I conducted a study of 2004
dispositions in federal district courts that confirms this estimate more
dispositively. I constructed a database of each opinion available from Westlaw
for fiscal year 2004 (October 1, 2003 to October 1, 2004)34; there were 27,890
such opinions.35 I compared those opinions, by district, to the Administrative
Office of the U.S. Courts’ reported number of civil and criminal cases
“terminated”36 in each district.37 Because most (though by no means all) of the

31

For many years, “reported” and “published” meant simply that opinions appeared in the printed
court reporter volumes—for district courts, the Federal Supplement and Federal Rules Decisions
series. There were, however, numerous other sources of opinions. See Allan D. Vestal, Reported
Federal District Court Opinions: Fiscal 1962, 4 HOUS. L. REV. 185, 186-87 (1966). As they have
grown, the electronic research services have begun to publish, electronically, more and more
opinions not designated for official “reporting.” Following ordinary usage, I call these
“unreported” opinions. Different district judges have different practices with respect to both
designating opinions for full, print, publication and allowing their inclusion in the Westlaw/Lexis
databases. Telephone interview with Jim Woodward, Clerk of Court, U.S. District Court for the
Eastern District of Missouri (Feb. 14, 2006).
32
See Susan M. Olson, Studying Federal District Courts Through Published Cases: A Research
Note, 15 JUST. SYS. J. 782, 790 (1992) (noting publication of just 37 of 697—6%—district court
cases in chosen case categories in the District of Minnesota from July 1, 1982 through June 30,
1984); Peter Siegelman & John J. Donohue III, Studying the Iceberg from Its Tip: A Comparison of
Published and Unpublished Employment Discrimination Cases, 24 LAW & SOC’Y REV. 1133, 1140
(finding that only 6% of federal employment discrimination cases filed in 1974 had district court
opinions accessible via Lexis; by 1986, the percentage had risen to 27%).
33
Evan J. Ringquist & Craig E. Emmert, Judicial Policymaking in Published and Unpublished
Decisions, the Case of Environmental Civil Litigation, 52 POL. RES. Q. 7, 9 (1999).
34
The search was in the “dct” database, and included only date restrictions. For example, I ran the
search “da(aft 12/31/2003) & da(bef 2/1/2004).” I searched month by month because Westlaw
pulls up only the first 2000 results of each search.
35
The database I constructed is available as part of this article’s online Appendix. See
http://schlanger.wustl.edu, under “publications.” To derive the number in text, I took out cases
from non-district courts that are (for whatever reason) included in the Westlaw’s dct database—for
example, the Court of International Trade and the Court of Appeals for Veterans Claims.
36
“Terminated” is a word used by the Administrative Office of the U.S. Courts to encompass all the
various dispositions that take a case off of the district court’s active docket—not only
adjudications, settlements, and defaults, but also remands, removals, transfers, and statistical
closings. For details on each of these dispositions, see Tech. Training & Support Div., Admin. Off.
of the U.S. Cts., Civil Statistical Reporting Guide (July 1999) (on file with authors and posted in
this article’s Appendix, http://schlanger.wustl.edu, under “publications”).
37
The Administrative Office of the U.S. Courts publishes terminations data each year, by fiscal
year. For this project, I used the on-line publication of 2004 data, see ADMINISTRATIVE OFFICE OF
THE UNITED STATES COURTS, supra note 29, tbls. C-4a & D-1. The data are also available in raw
form as study no. 4348 at the Inter-university Consortium on Social and Policy Research, see supra
note 30.

2006]

CIVIL RIGHTS LITIGATION CLEARINGHOUSE

163

Westlaw-retrievable decisions are for the kinds of dispositions that the
Administrative Office includes as “terminations,” this comparison produces a
rough national estimate of the proportion of case terminations that leave a record
in an accessible opinion. That proportion turns out to be a mere 8.7%, for 2004.
The proportion of “published” opinions (published, that is, in F. Supp. or F.R.D.)
to terminations is, as one might expect, far smaller—currently just 2.3%.
It would matter less that so few cases are observable from the judicial
opinions if those cases fairly represented the entire docket. But they do not.
Most importantly, opinions are limited to those cases with some judicial
decisionmaking. But fewer than half of district court cases end by adjudication,
even broadly defined; the others are either settled or withdrawn without
definitive court action.38 Second, even of those cases that are adjudicated, only a
portion have a written opinion. Judgment as a matter of law, bench trials, and the
like are the grist of the opinion mill. But many cases are instead resolved by jury
trial, summarily, or with an orally announced rather than written judicial
decision. Thus reported opinions offer little insight into these types of
outcomes—each of which is very common. The two factors above mean that the
world of the reported opinion exhibits a large skew towards judicial rather than
jury or party decisionmaking. Third, of those cases with opinions, in only a
portion does the judge in question submit that opinion to the court reporter
publisher for publication. Finally, in some very small subset of the publicationdesignated cases, the publisher may decide against publication.39
Moreover, even when judges are the decisionmakers, there is now
voluminous evidence that they choose to devote the time to fully developed
opinion writing in non-representative ways.
Comparing published to
unpublished outcomes, one study of Northern District of Illinois employment
discrimination cases found many more plaintiffs’ victories among the published
group than the unpublished.40 Such outcome non-representativeness seems to be
common. As Evan Ringquist and Craig Emmert summarized in 1999:
The most thorough comparison of political influences on published and
unpublished district court decisions, that of Rowland and Carp (1996),
reaches three general conclusions regarding this comparison. First, in almost
all instances, court decisions in published cases were more “liberal” than in
unpublished cases. . . . Second, more often than not, appointment effects were
greater in published case decisions than in unpublished case decisions. . . .
Finally, while Ducat and Dudley (1989) found that the published decisions of
federal judges were much more likely to support the policy and legal

38

See, e.g., Herbert Kritzer, Adjudication to Settlement: Shading in the Gray, 70 JUDICATURE 161
(1986); Gillian K. Hadfield, Where Have All the Trials Gone? Settlements, Nontrial Adjudications,
and Statistical Artifacts in the Changing Disposition of Federal Civil Cases, 1 J. EMP. LEG. STUD.
705, 720-22, figs. 5-7 (2004).
39
See Olson, supra note 32, at 784.
40
Siegelman & Donohue, supra note 32.

164

UMKC LAW REVIEW

[Vol. 75:153

positions of the presidents who appointed them, Rowland and Carp (1996)
find no such cohort effect for unpublished decisions. . . .41

Besides outcome skews, it is plausible that opinions are about different
kinds of cases than are typical on the docket—I would hypothesize that they are
more likely to concern complex than simple cases, more likely to be issued in
large stakes than small stakes cases, and so on.42 On the other hand, it does not
appear to be true that only unimportant cases are decided without opinion. In one
recent study, 81% of reported appellate opinions were from district court
decisions without reported opinions.43 Perhaps more telling yet, half the Supreme
Court’s docket of cases that originated in federal district courts during the study
period consisted of cases whose district court decisions were originally
unpublished.44
And finally, reported opinions exhibit a huge geographic skew of an
unstudied nature. The figure below demonstrates, presenting a histogram of the
94 district courts, grouped by the relationship between the number of opinions
that appear in Westlaw and the number of terminations reported by the
Administrative Office. It shows that while over half the districts (58 of the 94)
produced opinions at a rate of 5% or less of terminations, the others were more
diverse, ranging from 6-10% to 40-45% and even (for two outliers) 60-80%.
Because some of the more opinion-prone districts are quite large, the resulting
skew is severe: The twelve districts with the highest opinion-writing rate
produced over half of the opinions available on Westlaw, although they account
for only 12% of total terminations.45 Most notably, the Northern District of
Illinois, the Southern District of New York, and the Northern District of Texas
between them terminated 9.2% of all federal district cases in 2004, but produced
33.4% of all opinions.

41

Ringquist & Emmert, supra note 33, at 15-16 (discussing C.K. ROWLAND & ROBERT A. CARP,
POLITICS AND JUDGMENT IN FEDERAL DISTRICT COURTS (1996), and Craig R. Ducat & Robert L.
Dudley, Federal District Judges and Presidential Power During the Postwar Era, 51 J. POL. 98
(1989)).
42
Donald Songer & Reginald Sheehan, Who Wins on Appeals? Upperdogs and Underdogs in the
U.S. Courts of Appeals, 36 AM. J. POL. SCI. 235 (1992); Donald Songer, Danna Smith & Reginald
Sheehan, Nonpublication in the Eleventh Circuit: An Empirical Analysis, 16 FLA. ST. U. L. REV.
963 (1989). Even of cases reversed on appeal—a very small proportion—only a small fraction
were appeals from published opinions in the district court.
43
Donald R. Songer, Nonpublication in the United States District Courts: Official Criteria Versus
Inferences from Appellate Review, 50 J. POL. 206, 213 (1988).
44
Id. at 208.
45
These twelve districts are (in ascending order by publication rate): the Northern District of
Illinois, the District of Oregon, the District of Connecticut, the District of the Virgin Islands, the
Eastern District of Louisiana, the Western District of New York, the District of Kansas, the
Southern District of New York, the District of Delaware, the Northern District of Texas, the
District of Maine, and the Western District of Wisconsin.

2006]

CIVIL RIGHTS LITIGATION CLEARINGHOUSE

165

FIGURE: OPINIONS AS PROPORTION OF TERMINATIONS
FEDERAL DISTRICT COURTS, FISCAL YEAR 2004 (N = 94)

40
20

Number of districts

60

58

12
5

6

5

2

1

1

1

1

0

2
0

.2

.4

.6

.8

Opinions as proportion of all terminations
Source: Administrative Office of the U.S. Courts and all opinions in
Westlaw’s dct database; see supra text accompanying notes 34-37.

In short, federal district court litigation as it appears in accessible opinions
is likely unrepresentative of the entire docket with respect to type of disposition,
subject matter, outcome, and location. As Alan Vestal noted in 1966:
There seems to be an overemphasis in certain areas, while other areas are
under reported. Some judges do not participate; others apparently are not at
all selective in their contributions to the reported law. Reporting is left to the
author of the writing or to the publisher of the series of reports. Neither acts
unbiasedly in the process. 46

B. The Inadequacies of Other Sources
There are two other readily available sources of information about federal
litigation—settlement and verdict reporters, and the Administrative Office of the
U.S. Courts terminations databases. Like judicial opinions, each is inadequate
because it is either unrepresentative of the full district court docket or
insufficiently informative or both.

46

Vestal, supra note 31, at 220.

166

UMKC LAW REVIEW

[Vol. 75:153

Settlement and verdict reporters. The settlement and verdict reporters47 are
even worse as a source of unbiased information about the mass of cases.
Because they are mostly passive recipients of information given them by lawyers,
their sample is highly skewed towards the kinds of cases lawyers wish to brag
about.48
The Administrative Office of the U.S. Courts databases. When a federal
district court case ends, personnel in the court clerk’s office fill out a
computerized form that assembles information about it—its name, filing date,
district, docket number, the nature of the claim (using about a hundred “nature of
suit” codes), how the case was disposed of (by default, settlement, or various
kinds of adjudication), which party won, at what stage of the litigation, and so on.
The Administrative Office of the U.S. Courts assembles these observations into a
database and uses it to produce its published statistics. It also releases the raw
data to the Inter-university Consortium on Social and Political Research, which
in turn makes the database available on the web. The data include information
about every case filed in federal district court, since 1970.49 There are two
problems with the AO terminations data, which make it far less useful than court
records, at least for detailed inquiry about the content of litigation. First, the data
are extremely summary in nature.50 Second, they have significant error rates in
some fields.51
* * *
In short, for anyone who hopes to understand litigation—one specific
litigation or an entire field of litigation—there is no substitute for court records.
Historians, sociologists, anthropologists, political scientists, legal researchers,
and policymakers all need court records if they are going to understand either a
type of case or a particular litigation, whether it is individually important or
studied as an exemplar. There is simply no other source of information about the
substantive or legal issues, the conduct of the disputes, or their resolutions. We
are left, then, with an unmet need for access to court records—a need that the
National Archives can and should meet, for those records within its purview.

47

The Federal Jury Verdict Reporter premiered in October 2005, but federal verdicts and
settlements were previously available in state-by-state reporters.
48
See, e.g., Herbert M. Kritzer, Advocacy And Rhetoric vs. Scholarship And Evidence In The
Debate Over Contingency Fees: A Reply To Professor Brickman, 82 WASH. U. L.Q. 477, 499-502
(2004).
49
The database is maintained non-longitudinally, as ICPSR Study Nos. 8429 (1970-2000); 3415
(2001); 4059 (2002); 4026 (2003); and 4348 (2004), available at http://www.icpsr.umich.edu/.
50
For information on the available information, see the codebooks for the ICPSR studies cited
supra, and Admin. Off. of the U.S. Cts., Civil Statistical Reporting Guide, supra note 36.
51
See Theodore Eisenberg & Margo Schlanger, The Reliability of the Administrative Office of the
U.S. Courts Database: An Initial Empirical Analysis, 78 NOTRE DAME L. REV. 1455 (2003);
Hadfield, supra note 38.

2006]

CIVIL RIGHTS LITIGATION CLEARINGHOUSE

167

III. CONCLUSION
As described in the introduction, NARA is committed to retaining all trial
records for cases after 1969. It is only the non-trial cases that are at risk of
elimination from the National Archives. A great deal of recent work reminds us
just how rare trials are: In 1970, only ten percent of civil cases in federal district
court were terminated during or after trial, and the proportion has shrunk in
recent years to two percent.52 More important is that it is simply not true that all
(or even most) important cases see trials. Indeed, purely legal issues—which
surely can rank high in importance—are typically adjudicated without trial.
Settlements, too, can be extraordinarily important. We therefore welcome
NARA’s invitation to comment on the best way for it to proceed and urge that
the regulatory provision for keeping non-trial records for cases with historical
value be applied extremely liberally, to entire case categories. Only if the records
for entire case categories are preserved can future research understand any given
litigation context.
There are numerous case categories whose historical and political
importance is obvious—not only civil rights of all types (including prisoner
cases) but torts, environmental law, antitrust, securities law, and many others.
For any area of law in which the law is of public concern—which is to say,
nearly any area of law—it would be terrible to deprive future generations of the
opportunity to understand the way litigation works, from filing to disposition,
whether that disposition occurs by trial, on consent, or by non-trial adjudication.
In addition, we urge the Archives to continue collaborating with other
entities to increase access to, and use of, court records. For the vast body of
information currently hidden in district court records to be socially useful, it
cannot merely be warehoused. We have offered the Civil Rights Litigation
Clearinghouse as an anti-warehouse model. Because information in the
Clearinghouse is easily and quickly accessible, fully indexed and searchable, and
because the documents can be downloaded or copied, we are hopeful that it will
transform the use of the included records. Digitized information, that is, is more
usable information. This kind of access is not only a technocratic success, but a
democratic one as well.

52

See, e.g., Galanter, supra note 2, at 533-34 (tbl. A-2); Marc Galanter, The Hundred-Year Decline
of Trials and the Thirty Years of War, 57 STAN. L. REV. 1255 (2005). For a literature review of this
topic, see Margo Schlanger, What We Know and What We Should Know about American Trial
Trends, 2006 J. DISP. RESOL. 35.