Appellate courts and independent experts.

• Author: Ginsburg, Douglas H.

The federal courts of appeals increasingly hear cases that have scientific or highly technical content. This is particularly true of the Federal Circuit, because of its jurisdiction over patent cases, and of the D.C. Circuit, because of its review of agency rulemakings and adjudications. (1) Many of these complex cases pose a significant challenge for generalist judges, who typically lack scientific or technical expertise.

The challenge of understanding such cases has led some prominent observers to call for the use of independent expert witnesses or, alternatively, expert staff members to assist the courts of appeals as needed. Perhaps because the use of court-appointed experts in the federal district courts has not proved controversial, the idea of appointing independent experts to assist appellate courts has resurfaced in a serious fashion. Under this proposal, neutral staff or outside independent experts would advise judges ex parte and their advice would not be reflected in the record, save perhaps in the opinion of the court.

In my view, it would be a mistake for the federal courts of appeals to retain or consult experts for five reasons. First, the practice is inconsistent with the adversary system. Second, it may cause the judiciary subtly to transfer to the independent expert its non-delegable duty and authority under Article III of the Constitution of the United States to decide cases and controversies. (2) Third, in administrative review cases, which tend to be the most challenging in terms of scientific and technical subject matter, it would undercut the courts' appropriate deference to agency expertise. Fourth, it would sacrifice the virtues of a generalist judiciary. Finally, the practice would be unworkable. Before laying out these objections in greater detail, I discuss the technical challenges faced by judges and the proposals they have elicited that the courts of appeals rely upon independent experts.

1. THE INFORMATION JUDGES NEED

   Under the hard look doctrine, which originated in the D.C. Circuit in the late 1960s and early 1970s, (3) a court, in reviewing agency decision making, performs a "searching and careful" inquiry into whether the agency's "decision was based on a consideration of the relevant factors and whether there has been a clear error of judgement" but does not "substitute its judgment for that of the agency." (4) This inquiry often forces a judge to evaluate significant technical complexities.

   Consider the antitrust case United States v. Microsoft Corp. (5) Although it did not involve review of administrative action, the case is a good example of the difficult technical questions generalist judges may face. A key issue--which, for people under thirty, may not seem challenging, but for the average judge is a bit more daunting--was the extent to which Microsoft's Web browser, Internet Explorer, was so bound to its Windows operating system that it could not be removed without crippling the operating system. (6) If Internet Explorer was in fact inextricably bound to the Windows operating system, then computer hardware manufacturers such as IBM and Dell would have little incentive to pre-install another Internet browser, namely Netscape Navigator, on their machines, nor would consumers have much incentive, after purchasing a new computer, to purchase or download and to install that browser. (7)

   Other examples abound. Recently, a panel of the D.C. Circuit reviewed regulations of the Mine Safety and Health Administration that limited concentration levels for diesel particulate matter in underground mines. (8) Although the court, sensitive to the agency's comparative expertise and to its own limitations, explained that it would "give an extreme degree of deference to the agency when ... 'evaluating scientific data within [the agency's] technical expertise,'" it still had to examine the agency's scientific explanations closely for reasonableness. (9) The court therefore considered, among other issues, whether the agency had shown diesel particulate matter poses a significant risk to miners' health; whether the proxies the agency had used to set the standard were sufficiently accurate surrogates for diesel particulate matter; and whether the standard was one feasible of achievement in the circumstances of an underground mine. (10)

   In another recent case, the D.C. Circuit reviewed a rule issued by the Federal Motor Cartier Safety Administration governing the number of hours that truckers could work and drive in seven and in eight consecutive days. (11) In that case, the court vacated the rule because the agency failed sufficiently to consider the effect the rule would have upon the health of the drivers. (12) The court also expressed concern with the agency's justifications for other aspects of the rule. (13) For example, the court questioned the agency's decision to increase the maximum daily driving time to eleven from ten hours in light of statistical evidence that crash risks increase dramatically as driving hours increase after the eighth hour. (14) Although the agency cited several studies and its own cost-benefit analysis, the court noted that the agency had failed to explain how the cited studies justified the increase and that the cost-benefit analysis failed to account for fatigue generated by time on task. (15) The court was also dubious about the agency's justification for allowing a driver using a sleeper berth--a compartment in the truck where a driver can sleep--to split an otherwise continuous ten-hour rest period; the cited studies did not seem to support the agency's conclusion. (16) These were issues as to which there was a plethora of sometimes conflicting scientific and statistical evidence, all beyond the ordinary experience of the court.

   The decision of the Supreme Court in Whitman v. American Trucking Ass'ns, (17) which reversed the decision of the D.C. Circuit in American Trucking Ass'ns v. EPA, (18) is best known for confirming the rumored death of the nondelegation doctrine. (19) But on remand from the Supreme Court, the D.C. Circuit still needed to consider whether, for example, it was arbitrary and capricious for the EPA to select a relatively stringent annual standard and a relatively lax daily standard for particulate matter (20) and for the EPA to choose 0.08 parts per million as the appropriate National Ambient Air Quality Standard for ozone, as opposed to 0.09 or 0.07, the alternative levels considered by the agency. (21) Again, included in the record were various scientific studies and analyses, (22) and the panel needed to determine whether they supported the agency's position. (23)

   In these cases and in others like them, it is not uncommon for the joint appendix filed with the court--the appendix that brings to the court the documents and the part of the rulemaking record that the parties think are most relevant to the issues on review--to be four, five, six or eight volumes and to stack up to two feet. (24) Of course, the joint appendix is only a fraction of the full record, more of which the court may need to examine in certain cases. Former Chief Judge Patricia Wald of the D.C. Circuit described just how bad it can get:

   In Sierra Club [v. Costle], the D.C. Circuit was faced with a record of several thousand pages, twelve volumes of appendices, eight hundred pages of briefs and a forty-five-page, three-column, single-spaced statement of the agency's rationale. The case raised highly complex questions, such as the "technological feasibility" of meeting a ninety-percent sulfur dioxide reduction standard in coal burned in utility plants. Neither I nor the law clerk who worked on the case had any technical background in antipollution equipment or coal mining. As a result, we suffered through endless hours of deciphering innumerable designs of electrostatic precipitators, baghouses, etc. We agonized over statistical projections and even discovered a number of mathematical miscalculations in the agency's own rationale. We sent out a supplemental interrogatory requesting additional data on a technical point, the variability of scrubber performance at a particular plant. Although we asked for a "brief memorandum," we received in reply approximately 150 pages of explanation. At times we were overwhelmed and insecure, isolated in two rooms, surrounded by mountains of documents. (25) From time to time, many associates in law firms surely feel the same way Judge Wald did. With seniority, they may grow out of it--but the judges will not.

2. INDEPENDENT EXPERTS IN THE FEDERAL COURTS

   The idea of government by experts stems from the Progressive movement of roughly the 1880s to the 1920s. Progressives viewed the judiciary, and government generally, as encrusted with inefficient and vestigial practices. (26) As the Progressives sought to make government more efficient, there was a parallel movement toward "scientific management" in business; managers of factories conducted time and motion studies (27) and experimented with altering working conditions (28) in order to determine the most efficient methods of production. In general it was thought the "scientific"--in this context meaning systematic--study of any social problem would enable experts to discover the one best solution. (29) This is also the principle that brought us eugenics, public housing, and many other such disasters. (30)

   In the Progressive Era, experts in various fields, particularly in the sciences (both physical and social), were drawn into government at all levels to an unprecedented degree. To give a prominent example, the Congress established the United States Forest Service in 1905 and President Theodore Roosevelt appointed Gifford Pinchot, the most prominent advocate of scientific forestry and the person generally regarded as the father of American conservation, (31) to be the first chief of the Service. (32)

   Against that background, it is not surprising that many judges and...