In this article, I critique Jonathan Floyd’s method of normative behaviourism (NB): that we should measure political preference for a political system from levels of crime and insurrection. First, I distinguish between problems with the data and problems with the theory. I proceed to examine 6 groups who present a difficulty for NB and identify the common thread: NB abstracts the capacity of groups to commit crime and insurrection, and therefore, misreads them in the data as normative approval of a (...) political system. Next, I argue that this a problem, especially as that capacity is often caused by the conditions of the political structure such groups live within. This lack of capacity often means they are amongst the most vulnerable. Consequently, NB needs to be careful of overlooking corrective justice. Subsequently, I offer some simple amendments to NB, followed by two complementary approaches: ethnography and fictional narratives. (shrink)