A bedrock principle of administrative law is that when a court finds an agency has erred, the court generally remands the action for the agency to consider anew (as opposed to the court deciding the matter itself). The conventional understanding is that this ordinary remand rule is part of the suite of judicial deference doctrines in administrative law. In our contribution to The George Washington Law Review’s Annual Review of Administrative Law, we argue that this understanding is incomplete—at least when it comes to high-volume agency adjudication. In that context, the vast majority of agency adjudication decisions never make it to federal court. Judicial remands in cases that reach the courts allow the courts to engage in a dialogue with the agency, in turn improving agency decision-making in similar cases that never make it to federal court. Indeed, courts have developed and utilized a variety of tools to engage in a richer dialogue with the agency on remand. Remand, thus, can be a tool for judicial engagement and dialogue, not just one for judicial deference.
This argument, however, assumes that a dialogue between courts and agencies actually takes place—that remand is not just a judicial monologue. This Article explores the empirical realities of that assumption by presenting the findings of two separate studies: a cross-agency study for the Administrative Conference of the United States on agency appellate systems and a FOIA-based study of agency immigration decisions on remand. Although much more empirical work needs to be done, the findings from these studies provide an empirical window into how agencies engage with and respond to courts on remand. In light of these preliminary yet promising findings, we argue that courts (and agencies) should consider how to better engage in a dialogue on remand in order to produce a more systemic effect on high-volume agency adjudication systems.