Question

A researcher is interested in studying the effects of city law enforcement resources on local crime rates. He measures law enforcement resources with the number of police officers per capita. His dependent variable is the number of violent crimes per capita. If he conducts a traditional multiple regression analysis, would his results be subject to reverse causation? If so, why?

Answer 1

Yes, his result might be subjected to reverse causation. Because, in direct sense, number of violent crimes per capita should be dependent on law enforcemengt resources, because, the more is resouces, the less should be the violent crime. But in other way, in response to an increase in number of violent crimes per capita, the city government might increase law enforcement resources to mitigate the rate of violent crime. In that sense, violent crime per capita causes the voulme of law enforcement resources. So, it might cause a direct causation