Court-System Transparency

• Author: Lynn M. LoPucki
• Position: Security Pacific Bank Professor of Law at the UCLA School of Law
• Pages: 02

Lynn M. LoPucki is the Security Pacific Bank Professor of Law at the UCLA School of Law, the annual Bruce W. Nichols Visiting Professor of Law at the Harvard Law School, and the former Senior Advisor to the National Bankruptcy Review Commission with respect to the Commission's Data Study Project. He can be contacted at lopucki@law.ucla.edu. Thanks are due to Drew LoPucki for his assistance in developing the ideas presented in this paper; to Shahab Asghar for assistance with research; and to Frances Foster, Elizabeth Gibson, Melissa Jacoby, Peter Joy, Bob Lawless, Tony LoPucki, Aaron Moshiashwili, Katie Porter, Hillary Sale, Walter Weyrauch, and workshop participants at The University of Iowa College of Law, the University of North Carolina School of Law, the University of Illinois College of Law, the Washington University School of Law, and the Rand Corporation for comments on earlier drafts.

Page 483

I Introduction

Computer algorithms capable of predicting the outcomes of legal disputes may soon be a reality. These algorithms would base their predictions not on legal analysis of statutes, regulations, and court opinions,¹ but on statistical analysis of case outcomes. Because the algorithms do not yet exist, the case characteristics from which they would make their predictions remain largely a mystery.

The characteristics would certainly include the identities of the judge²and the lawyers³ in a particular case. Those two pieces of information alone would make the algorithms highly provocative, because neither is supposed to play any role at all in case outcomes. Through the lens of such algorithms, observers would see the courts as they actually operate, instead of as they are supposed to operate.

Court transparency requires such algorithms. A court system is "transparent" for the purposes of this Article when all relevant aspects of its operation are revealed to policymakers, litigants, and the public in forms that they can readily comprehend. For reasons that this Article explains, Page 484 court systems can become transparent only when court files are maintained in relational electronic formats and the public has free, technologically unfettered access to their contents. Relational formats are the familiar formats used in data-management and spreadsheet programs. Each piece of data is tagged as the value of a characteristic of an object. Empiricists refer to relational data as "coded." Statistics programs can process such data into statistics, tables, and graphs.

The federal courts-which already maintain court files in electronic format and make them available over the Internet-are on the brink of introducing relationally formatted forms. Users of these forms "code" the data as they create it, by entering it into fields (boxes) in specified formats- essentially the way customers fill out order forms on the Internet. The Judicial Conference of the United States (the "Judicial Conference"), the regulator of the federal courts, has approved a data-enabled PDF format for use, and the United States Trustee, a division of the Justice Department, has recommended to software vendors that they begin using it for nineteen documents commonly filed in bankruptcy cases.⁴

The federal courts are not introducing relationally formatted forms merely to achieve transparency. Indeed, transparency may not even be among the courts' objectives. Data-enabled forms would improve court administration by facilitating its automation. once the technology is in place, however, the courts will be transparent unless regulators erect or maintain barriers to prevent public use. So far, the Judicial Conference has permitted a steady but cautious advance toward transparency.

This Article takes as its starting point the current state of the world's most transparent court system-the United States Courts as accessible through Public Access to Court Electronic Records ("PACER").⁵ Part II describes the physical changes necessary for that system to achieve transparency.

The remaining Parts then use systems-strategic analysis⁶ to project the consequences. Part Ill explores the benefits of transparency. Transparency would expose and reduce corruption. It would expand the power of citizens Page 485 and legislators over the courts and make the actual rules that govern society visible to the public. Transparency would promote the settlement of the most predictable cases and improve overall efficiency by focusing the system's resources on the least predictable cases. Transparency would reduce or prevent many kinds of lawyer malpractice and litigant error.

Part IV examines and refutes the arguments generally raised in opposition to court-system transparency. The costs of transparency to the public would be minimal, because academic institutions, nongovernmental organizations, individuals, and the private sector would fund and do the cheap, relatively simple data processing required. on careful examination, the variety of privacy-based arguments generally raised in opposition to transparency all fail-principally because the records to be made transparent are already public and widely available. Transparency would enhance rather than impair judicial independence because it would provide a sounder basis for the public's evaluation of judicial performance. Transparency would enable widespread copying of legal documents. But that too turns out to be a benefit rather than a detriment when analyzed in terms of system function. This Article concludes in Part V that the effects of court-system transparency would be overwhelmingly positive.

The analysis proceeds on two assumptions. First, the government chooses to release all recorded court-system data except those sealed by court order or redacted pursuant to recently adopted federal privacy standards. Second, the government provides only the present level of funding for data analysis. In accord with the systems-strategic method, the analysis then projects the consequences by taking the perspectives of various court-system participants, imagining the strategies that each would pursue in response to the data release, and speculating on the interaction among those strategies.

II The Technological Challenge

This Part briefly describes the current system for access to federal case files, the preconditions to transparency, and the mechanisms by which courts would achieve transparency. It describes some issues that would arise in removing the limitations that now prevent transparency and the effects of removing those limitations.

A The Mechanisms Of Transparency

PACER is a system operated by the federal courts. Since about 1997, PACER has made federal-court case files, including the dockets, publicly available over the Internet.⁷ A case file generally includes all documents Page 486 filed with the clerk of the court by the parties or the judge in connection with a case.⁸ The "docket" for a case is a list of all documents contained in the case file. Prior to PACER, the public could access case files only by traveling to the courthouse, ordering records by mail, or buying copies from service-company intermediaries.⁹ With only minor exceptions, PACER now makes district-court and bankruptcy-court case files-including hearing transcripts-publicly available over the Internet. A pilot project is underway to include digital audio recordings of hearings.¹⁰ The dockets are in HTML format; the documents are in PDF format.

PACER users must register online and pay eight cents per page to view or download each document, up to a maximum single-document charge of $2.40.n In addition, PACER's report function provides a limited ability to search for cases nationwide (by party or, in the district court, by nature of case) or in a specific court (by several criteria).¹²

For a court system to be transparent, up-to-date research must be available to answer millions of questions regarding the patterns of cases and case outcomes. Already hundreds, if not thousands, of researchers use data extracted from PACER. The work is labor intensive. A researcher can sometimes use PACER's report function to identify a group of cases for study, but more often researchers must identify the cases from outside sources. The researcher must select and download the documents (and incur costs) before the researcher can word-search them. Many PACER documents cannot be word-searched at all, because the documents are scanned rather than text-based PDFs, and the scans are of insufficient Page 487 quality to rescan them to text. Researchers typically print hard copies, mark the data to be extracted, and convert the data to relational form by coding and entering them into spreadsheets or databases by hand. The researchers then analyze the data...