Administrative law scholars have long asked whether and how systems of review or appeal can limit the discretion of adjudicators and thereby promote consistency in adjudications. From immigration courts to patent adjudications to restaurant inspections, recent scholarship has documented dramatic inconsistency in administrative decisionmaking. The resulting calls for reform have focused especially on systems of administrative review, with advocacy for quality assurance systems and peer review as supplements to (or replacements for) traditional administrative appeals. Mostly absent from these debates, however, have been empirical studies of the effectiveness of traditional systems of administrative review. This Article proposes a theory of consistency-enhancing, discretion-limiting administrative review, along with a method of evaluating the effectiveness of systems that conduct such review. The theory implies that a traditional system of review can promote consistency at three steps. First, litigants should be more likely to appeal the decisions of judges unfriendly to their claims. Second, appeals judges should disproportionately reverse decisions from such judges. Third, those judges should seek to avoid remands and therefore alter their behavior when a decision of theirs is remanded. Effectiveness at these three steps has observable implications, and I evaluate each step using a previously unreleased database, obtained by Freedom of Information Act request, of nearly 4 million Social Security Administrative Law Judge and Appeals Council decisions from 2010 to 2014. Disparities in ALJs’ allowance rates - their relative generosity in granting benefits - are notoriously large. Using the within-hearing-office variation in these allowance rates as a measure of consistency, I find that the Social Security Appeals Council promotes cross-judge consistency at the first and third stages of appeal, but not at the second. Appeals Council decisions do not, on average, distinguish between the decisions of harsher and more generous ALJs. Interviews suggest that this lack of distinction reflects the Appeals Council’s deference to the substantive findings and conclusions of administrative law judges, combined with little deference to the ALJ’s articulated rationale for a decision. Although such an approach shows respect for the factfinder’s decisions and a commitment to procedural due process - and indeed, the Appeals Council has recently won awards for its effectiveness - such review does not address cross-ALJ consistency in outcomes. I therefore propose that the Social Security Administration broaden the mission of the Appeals Council by de-emphasizing policy compliance and granting full review more often.Stepping back, I compare these results with those of previous research on immigration courts, where the largest problems occur at the first step of review (the selection process): many immigrants are unable to afford lawyers and do not appeal, regardless of the potential merit of their case. For both processes, careful empirical attention to what happens before an appeal is filed and after a decision is reached is critical to institutional design.
Read full abstract