The Effects of Spatial and Temporal Influences on Anti-refugee Violence

Using German district-level data on attacks on refugees from the three years 2014 to 2016 we test a number of determinants that may explain why an attack occurred in a given district on a given day. Besides traditional approaches mainly from sociology, such as economic deprivation and the contact hypothesis, we also test the strength of extreme right-wing and populist parties. Furthermore, we contribute to existing research by taking into account variables with a strong temporal variation. Specifically, we not only test (1) a possible contagion effect stemming from previous attacks but also (2) draw on a systematic collection of critical events to test their effect on the probability of attacks. These events comprise terrorist attacks in Germany and neighbouring countries as well as reported cases of crimes committed or allegedly committed by refugees among others. The combined information which event has occurred, when, and in which district, can then be used to test a radiating effect of these events on other districts. This allows for determining which kind of event matters, how much and for how long. Finally, we extend previous studies on the role of media for violence against foreigners (2) employing a variable that measures the extent of negative statements made in the press about refugees over time.