We illustrate the sensitivity of two-way fixed effects difference-in-differences estimates to innocuous changes in data structure. Using the staggered rollout of state-level bank branching deregulations, three outcome variables are brought to bear on the interventions: personal income growth (a replication), house prices (new to the literature), and per capita cigarette purchases (a falsification test). Estimates are sensitive to panel length, and the data structure creates the false impression of a causal effect of the interventions on all three outcome variables. We contend that any two-way fixed effects regression using this set of interventions is at risk of generating spurious results.