Replication Data for: On Separation of Powers and Obfuscation in U.S. Supreme Court Opinions.

A longstanding debate in American judicial politics concerns whether the U.S. Supreme Court anticipates or responds to the possibility that Congress will override its decisions. A recent theory proposes that opinions that are relatively hard to read are more costly for Congress to review—and that as a result, the Court can decrease the likelihood of override from a hostile Congress by obfuscating its opinions: writing opinions that are less readable when congressional review is a threat. I derive a straightforward but novel empirical implication of this theory; I then show that the implication does not in fact hold. This casts serious doubt on the claim that justices strategically obfuscate opinion language to avoid congressional override. I also discuss sentence tokenization as a source of measurement error in readability statistics for judicial opinions.