Ironically, civil service exams were supposed to be the fairest way for cities to hire the best firefighters and police, while opening the doors to more minorities. Exams, it was thought, provided a color-blind way to measure performance and promote minorities into leadership roles within organizations that had clearly discriminated in the past. The problem is that, for reasons not understood, minorities have not performed as well as whites on tests.

But are multiple-choice tests to measure firefighters’ retention of information the optimal way to predict how someone would react at a four-alarm fire?