Review of Administrative Agency Policy Decisions
The goal of preventing administrative agencies from making ideological decisions is not a new one. Under the Administrative Procedure Act, a federal court in a proper challenge can strike down a regulation if it is “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.”7 Interpreting that provision, the U.S. Court of Appeals for the District of Columbia Circuit, which hears appeals of many administrative law challenges, has found that the role of the courts is to ensure that the agencies have “genuinely engaged in reasoned decisionmaking.”8 A court is required to intervene if it “becomes aware, especially from a combination of danger signals, that the agency has not really taken a ‘hard look’ at the salient problems.”9 This doctrine has come to be known as the “hard-look” doctrine, and although it is technically the agency that must look hard, in practice the courts also carefully consider the administrative record to ensure that the agency has done that. Although the doctrine is not specifically designed to prevent ideological decision making, as opposed to decision making that is arbitrary because of shortness of time or other factors, it often serves that purpose. The Supreme Court’s most famous application of hard-look review occurred in a case involving a politically controversial application of the National Traffic and Motor Vehicle Safety Act of 1966.10 In 1972 the National Highway Traffic Safety Administration (NHTSA) decided to require, in addition to seat belts, some form of passive protection (air bags or automatic seat belts) for front-seat passengers. Before that decision went into effect, the Secretary of Transportation decided to suspend the requirement in 1976, but his successor disagreed, requiring that either air bags or passive seat belts be installed beginning in 1982. In 1981, however, a new Secretary of Transportation reopened the rulemaking and canceled the previous standard. It was obvious to all that different philosophies of government played a role in the decision, with the Reagan administration more skeptical of such intervention than the Carter administration. In a review of a lawsuit challenging the Reagan administration’s decision,11 the Supreme Court showed a willingness to scrutinize the particular arguments offered in favor of the change. The NHTSA had observed that manufacturers were complying by introducing passive seat belts rather than air bags, but the passive seat belts could easily be detached and thus were ineffective. The agency, however, never explained why it did not respond to this problem by mandating air bags instead. The effect of the Court’s ruling, intended as a guide to how lower courts should make similar assessments, was to vacate the agency’s decision. The agency would still have an opportunity to make the same decision, but it would have to provide a better explanation, for example, by showing that an air bag–only rule would be too expensive. Of course, it is not always easy for an agency to convince a court the second time around to approve a decision if the court believes that the arguments in favor of that decision are inherently weak. The hard-look doctrine thus provides some means for the courts to police ideological decision making. There are, however, at least two problems with this approach. First, in some cases the courts’ own ideology will simply replace the administrative agency officials’ in determining whether regulations are acceptable. Perhaps having an additional check against aggressive regulatory decisions is useful, but there is no inherent reason to prefer the status quo, and whether the court will serve this role in any event will depend on which judges happen to be selected to hear a particular case. An unfortunate consequence is that judicial review may be unpredictable, and administrative agencies sometimes waste a great deal of time on rulemakings that they could not be sure would fail. Second, taking the courts’ explanations of the doctrine at face value, it leaves a great deal of discretion for administrative agencies. At least in theory, in any case in which reasonable arguments can be developed in support of a position, the courts are supposed to defer to the administrative agencies. The system thus tolerates a relatively high degree of flexibility in decision making. Perhaps this is essential because of the variance in ideological leanings of the judiciary, but it is not inherently desirable. Normative prediction markets in issues being considered by administrative agencies could become useful data points when the courts consider lawsuits claiming that regulations represent bad policy and should be struck down. If such a market were to forecast that a broadly representative pool of decision makers would generally approve of a particular set of regulations, judges who do not like the policy result might nonetheless approve it. Indeed, agencies sometimes might have incentives to sponsor normative prediction markets if they believe that the markets will tend to support their projects. In other cases, a normative prediction market might show that most decision makers would be expected to reject particular regulations, and this might help persuade judges to strike down the regulations. Of course, an agency might not have an incentive to sponsor such markets, but private parties might do so as a means of justifying the claim that particular regulations are ideologically motivated. If such markets became commonplace, the courts might be suspicious in cases in which agencies do not create them. Naturally, prediction markets cannot be expected to be considered by judges until they become better accepted as relatively accurate and objective predictors. Normative prediction markets could help respond to both concerns about the hard look doctrine detailed above. First, such markets might not merely be genuinely useful for judges seeking unbiased assessments of the reasonableness of decisions but also might be effective in unmasking ideological decision making by judges. For example, if two very conservative judges happen to be chosen to review a relatively liberal agency decision, they might hesitate to write an opinion striking down the regulations as poorly reasoned if the prediction market indicated that virtually all agency decision makers would have come to the same conclusion. Second, normative prediction markets might facilitate more skeptical judicial review of decisions that most agency officials would be expected to reject. Aside from promoting processes that might improve agency decisions, the hard-look doctrine is useful primarily because it can be expected to stop the most egregious agency decisions. It might, however, be useful to block less egregious decisions that most decision makers would nonetheless reject. Normative prediction markets provide a relatively objective assessment of how a wide range of decision makers would view the agency’s decision. They could thus facilitate a shift in the hard-look doctrine away from an analysis of the reasoning underlying a decision and toward an analysis of the decision itself. In the absence of such markets, a shift such as this would compromise the relative objectivity of the doctrine. Normative prediction markets could allow people to focus on what ultimately matters: the decision an agency makes rather than the reasoning of an agency in getting there. Of course, an agency still might explain its reasoning process regarding some questions, because market participants would assess the degree to which these arguments are likely to convince a future decision maker. But the purpose of such explanations would be to persuade, not merely to fill out paperwork that will satisfy a court. Thomas McGarity has complained that the notice-and-comment rulemaking process has produced an “ossification” of the administrative state, preventing agencies from getting much done.12 To the extent that procedural requirements stop agencies from advancing narrow ideological agendas, they should be embraced. By focusing directly on whether decisions are ideological, normative prediction markets may free agencies from procedural obligations that do not promote substantive goals. No legislative reform would be needed (as it would be for the vast majority of the proposals advanced in this book) before courts could consider normative prediction markets in conducting hard-look reviews. The review process is already relatively open-ended, and the in the air bags case the Supreme Court scrutinized technical studies. All that would be necessary is for a policy analysis organization or other private party to create a normative prediction market, with appropriate specification of how the ultimate decision maker would be chosen. The prediction market might focus on a question related to a particular aspect of the agency’s reasoning or on the overall justifiability of the agency’s decision. A litigant might then present evidence about the normative prediction market. Only a bold judge would be willing to consider such unconventional data today, but if enough normative prediction markets for assessing agency decisions were created, judges might come to appreciate their virtues.
Leave a Reply