We document two potential biases in recent analyses of UI benefit extensions using boundary-based identification: from using county-level aggregates and from across-border policy spillovers. To examine the first bias, we use a regression discontinuity (RD) approach that accounts for measurement...