When tariffs bite: The short-run vs long-run welfare effects of trade wars

Non Technical Summary

Tariffs have returned to the centre of international economic policy. This paper examines the effects of tariffs differ between the short and long run. It estimates how responsive trade is to changes in trade costs across 22 manufacturing sectors, using data covering 170 countries between 1995 and 2022. These estimates are then incorporated into a model of the global economy covering 77 countries. The model is used to examine how different types of tariff increases affect countries in the short run and after economies have had time to adjust.

The results show that there is no single relationship between the length of time after a tariff increase and its economic effects. Instead, the outcome depends importantly on how widely tariffs are imposed and whether firms and consumers have access to alternative trading partners.

When one country imposes tariffs while its trading partners do not retaliate, adjustment can reduce some of the losses experienced by the countries facing the tariffs. Over time, firms can shift their purchases towards suppliers in countries that are not subject to the new tariffs. This ability to redirect trade provides an alternative source of supply and reduces the impact of the original tariff. In the paper's example of a unilateral 25-percentage-point increase in US tariffs, the welfare losses experienced by trading partners become smaller over time. The welfare gain for the US also falls as the economy adjusts, from 1.56% in the short run to 1.15% in the long run.

The picture changes when tariff increases become more widespread. If trading partners retaliate, the country imposing the initial tariffs can lose the benefits it might otherwise obtain from its improved position in international markets. In the paper's US example, the initial welfare gain turns into a welfare loss when trading partners retaliate: the US welfare effect falls from a gain of 1.56% under unilateral action to a loss of 2.21% in the short run and 2.39% in the long run.

The distinction between short- and long-run effects is particularly important in a broader trade war, where many countries raise tariffs against one another. In this situation, there are fewer alternative markets to which firms can redirect trade. As a result, the adjustment that can cushion the effects of unilateral tariffs is much less effective. In the paper's multilateral tariff-escalation scenario, average welfare losses increase as economies adjust, from 2.61% in the short run to 2.96% in the long run when intra-EU trade is exempt.

These findings have important implications for how policymakers assess the effects of tariffs. The economic impact observed shortly after a tariff is introduced may not provide a reliable indication of its eventual effect. Firms need time to change suppliers, reorganise production and establish alternative trading relationships. Whether this adjustment reduces or increases the eventual cost depends on whether alternative markets remain available.

The results therefore highlight the importance of considering the structure of a tariff shock, rather than treating all tariff increases as equivalent. A tariff imposed by one country may have different consequences from a situation in which several countries respond with their own tariffs. For policymakers assessing the potential effects of trade restrictions, both the immediate disruption and the longer-term reorganisation of trade need to be taken into account.