Starting with the Text — On Sequencing Effects in Statutory Interpretation and Beyond

Abstract

What difference do starting points make? The question is important for decision making in general and for law in particular, including the interpretation of statutes. Judges must begin the interpretive process somewhere. Today, Supreme Court opinions sometimes promote the idea of starting with the text of the statute at issue. But what does this mean, in practice, and does it matter to decisions? “Start with the text” could be a signal of allegiance to an interpretive school and an indication that some interpretive tools are more important than or even lexically superior to others. At present, however, the statement delivers neither of those messages well. Instead, we might think about the statement as a rule for sequencing sources. Although the mere sequence in which information is considered does not have any clear logical significance for case results, the idea of starting with statutory text can become almost unnervingly significant — and without adding lexical priority.Decades of studies show that the order in which information is presented can influence decisions, apart from what formal logic dictates. But the direction of order effects can be counterintuitive and sensitive to the decision environment, which suggests complications for a “start with the text” sequencing rule. Depending on several factors, the first item of information will matter most, the last item will matter most, or there will be no order effect. Furthermore, even if order effects are predicted accurately, some psychological mechanisms that produce order effects are normatively problematic for judicial use. Finally, an effective sequencing rule requires an implementation strategy with a foundation far away from standard theorizing about interpretive method. Foregrounding implementation issues and the real world of interpretive architecture suggests that, if judges want to harness order effects, they probably should turn to the most important sources last, not first.