We study how firms react to unexpected increases in import tariffs. We identify our results from a sudden removal of American preferential tariffs applied on Argentine imports under the Generalized System of Preferences, which reflected American retaliation to a dispute over intellectual property between the two countries. Critical for identification, the tariff hike affected a third of Argentine exports enjoying preferential access in the American market, but did nothing to the other two thirds. We find that the higher tariffs reduced export participation of affected Argentine firms in the US market, whereas resilient exporters dealt with the cost increase by reshuffling their export baskets away from the products whose tariffs increased. In fact, affected firms were more likely both to drop suspended products from their export basket and to start exporting new (non-suspended) products to the US. Interestingly, the extensive margin effects carry over third markets, where policy did not change: after the policy shock, affected firms selling to the US were less likely to export to other markets. This happened, however, only for firms that also exited the American market. Those findings reveal the nuanced consequences of tariff preferences on the behavior of exporting firms, highlighting that their effect tend to spill over other products and other markets.

We study the impact of exogenous news on the classic Bayesian persuasion problem. The sender supplies information over multiple periods, but is unable to commit at the onset to the information that she will supply in periods ahead. A tension then emerges between the sender and her future self. We show that by resolving this tension, more informative news can make the sender better off.

Is it better to have the first or the last word in two-round debates like common law trials? If litigants always share available witnesses then they never prefer to present first (to lead), and may prefer to follow; and they never prefer to choose the order after observing the available witnesses than to always follow. Litigants can prefer to lead if they have different available witnesses because the outcomes when some litigant leads coincide with the outcomes in an alternative game where that litigant follows, but commits to its reaction function before observing its available witnesses. Leading is therefore a commitment device.