Chad Oldfather has a new article, Universal De Novo Review, just out in the George Washington Law Review, asking why appellate courts always review legal issues de novo without deference to the trial court. Chad runs through some of the common explanations and find them wanting, and suggests that perhaps there should be exceptions to the general rule where appellate courts defer to trial court legal rulings.

  I found it in an interesting and thought-provoking article, and I wanted to blog a few thoughts about it. Some of my take matches Chad's, and some is different: I mean this more as a general take on these important issues than a specific agreement or disagreement with Chad's views.

  Let's start with the general function of standards of review. In my view, the key to understanding standards of review is that they distribute power among institutions. A standard of review defines the power distribution between the reviewing institution and the institution under review. A deferential standard of review gives power to the institution reviewed; a de novo standard of review retains power for the reviewer.

  Standards of review vary because different institutions serve different functions. Different institutions — trial courts, agencies, appellate courts — serve different roles, and the standards of review distribute power to enable them to serve those roles effectively.

  When understood in that way, I think most standards of review begin to make sense. Consider the kinds of trial court rulings that receive deference by appellate courts. Trial courts are given deferential review of case-specific types of determinations — most obviously factual determinations, and equally importantly, questions of trial procedure such as evidentiary rulings that have no broader significance beyond that particular case. Why give deference for these issues? The reason is that structurally speaking, trial courts are designed to make judgements that are primarily case-specific. As Judge Sotomayor put it in her Duke comments, the focus is on the details of that particular case, and decisions typically only apply to that particular case.

  The deferential standard of review enables that function. It gives trial courts the power to do their job in individual cases. Granted, appellate courts provide some check on that. But they act as a check only when the trial court rulings are way off, such as when a factual finding is clearly erroneous or an evidence ruling is an abuse of discretion. Most of the power to decide case-by-case issues is given to the district court, and the standard of review reflects that.

  Now turn to the focus of Oldfather's article, the general rule that appellate courts review trial court legal interpretations de novo. Why is that? In my view, the reason is that structurally speaking, appellate courts are primarily designed to settle questions that cut across cases. From a structural standpoint, what matters is that there is some body that settles the law — some institution that announces the rule or standard that others must follow.

  De novo review of legal questions enables that function. It lets the institutions that have the power to announce the binding law do so unencumbered. If appellate courts deferred to trial court determinations of law, you would either have a situation in which no binding law is ever created (if stare decisis does not apply to the upheld trial court legal reading) or else the first trial court to get to an issue binds all others (if it is). Neither outcome would allow appellate courts to fulfill their designed role as institution that settles questions across cases.

  The idea that standards of review enable institutional roles largely explains the main exceptions to the usual rule of de novo review. For example, Chevron deference in the case of judicial review of agency interpretations of law is needed because modern administrative agencies are designed to have substantive powers in a zone of delegated discretion. They can't exercise that power effectively if courts can second-guess every legal interpretation with de novo review; more deferential review is needed for agencies to fulfill the function they were designed to serve. Similarly, deferential federal appellate review of district court interpretations of state law is allowed because it isn't the role of a federal court to settle state law. That's primarily a question for state courts, and deference to the federal trial court interpretation limits federal court intervention into the proper powers of state courts.

  One issue on which Chad and I agree is that the "expertise" rationale for standards of review isn't very illuminating. We can certainly hope that the institution designed to wield a certain power has or will develop expertise in executing it. Sometimes that is true, sometimes it isn't. But the real issue is the distribution of power, not whether we can in the abstract say a particular institution is more "expert" than another.