ABSTRACT
David Greenberg has done a masterful job of considering the methodological and data requirements for a study of the New York crime drop. Similarly, if a series trends now, it should also have trended in the past; if it did not, trend explanations become questionable. Closer attention to the processes that underlie crime rates might be helpful in resolving these causal attribution questions. No organized group or collection of individuals has much to gain from making crime rates go higher, yielding no incentives to sustain a positive trend. Although a random walk can regularly increase or decrease for many periods, it will eventually alter its course when conditions change. The few studies that investigate crime rate processes also generally support the random walk hypothesis. One recent analysis does claim to find multiple structural breaks in US crime rates, but it assumes that the rates vary around deterministic trends.
