Taking and testing jurisprudential regimes seriously: A response to Lax and Rader

Lax and Rader critique our use of the Chow test in our series of articles on jurisprudential regimes on the grounds that individual justices votes are not statistically independent, which constitutes a violation of assumptions underlying the Chow test. In this response we point out that the Chow tests constituted only one part of our analysis; we also conducted a sensitivity analysis to look at the strength of the Chow tests compared to other sequential splits. Most importantly, we required that the observed statistical patterns of change be theoretically consistent with the legal changes made by the regime changing decisions; we note two areas where we did preliminary analyses that produced statistically significant results, but where those results did not make sense in light of the jurisprudence. We repeat both our Chow tests and individual interaction tests taking into account the clustering of observations. Our reanalysis provides support for some, but not all, of our original results.