Endogenous Border Times

We examine transaction-level Peruvian import data to show that firms are subject to significant costs of port-of-entry delays. At the transaction level, we observe the time it takes a shipment to clear each step in the entry process. Our theory shows conditions under which observed entry times are endogenous. As a result, total entry delays potentially lead to biased policy conclusions and non-informative efficiency rankings of countries' entry procedures. We make three empirical contributions that help unbundle sources for time costs in trade and border effects. First, we provide evidence that at least part of the total port-entry-time is endogenous. Second, we identify the effect of entry delays on imports based on exogenous necessary entry processing. Third, we provide evidence that trade costs due to entry delays are heterogeneous across firm types. New and large importersare more elastic with respect to entry delays. This information allows researchers and policymakers to interpret aggregate port of entry delay data and their costs across different types of firms.