In sum, the evidence suggests that unnecessary police killings decrease after departments institute increased restrictions yet the likelihood of increases in restrictions is based on demands from the most politically powerful segments of society. In order to test whether changes in police killings are determined in part through the differences in 2 political power based on the superimposition of race and class, political explanations at the city level should be considered. A review of previous studies suggests that there are two broad city level explanations that influence the likelihood that police will use deadly force against citizens: the political or conflict perspective, and the reactive or community violence perspective.