Preemptive Trade Policy under Bilateral Severance Risk

Current trade policy discourse around friendshoring and de-risking rests on the premise that supply relationships between major economies can be severed, and that exposure to such a partner carries a cost standard trade analysis has not priced. This paper asks whether a country facing a probability of bilateral severance with one partner should preemptively tariff that partner or reduce tariffs on a safe third partner, in order to build up the stock of safe-partner relationships before severance occurs. Absent policy, private entrants do not internalize the value that safe-partner relationships will have in the severance state, so the stock carried into that state is too small. I develop a three-country, two-period dynamic general equilibrium model with CES preferences, sunk-cost relationship entry, intermediate-input amplification, and bilateral retaliation. A marginal preemptive tariff strictly raises the safe-partner stock, and sufficient conditions are given under which the marginal tariff and the marginal tariff reduction on the safe partner raise welfare. Calibrated to the United States, China, and the rest of the world with 2024 data, the tariff condition holds under a severance risk provided at least 7.6 percent of the observed 2017–2024 diversion of U.S. imports from China to the rest of the world is attributable to the tariff shock, and the condition for a reduction of the prevailing tariff on the safe partner holds at a threshold of 13.9 percent. Both thresholds fall well below any plausible attribution share, and the tariff reduction on the safe partner is the more powerful instrument for diversification.