We illustrate the sensitivity of two-way fixed effects difference-indifferences 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...