The Cost of a Global Tariff War: A Sufficient Statistics Approach

In an age marked by growing trade tensions and the unraveling of multilateralism, the threat of a global tariff war looms large over the international economic system. But what would such a conflict actually cost the global economy? A recent study provides a powerful and policy-relevant answer to this question. The paper introduces a novel, tractable method to estimate the cost of a hypothetical tariff war across a broad set of countries and industries, revealing that the risks associated with protectionism are not only substantial but have escalated sharply in recent years.

Traditional methods for evaluating the cost of tariff wars compute the global Nash equilibrium of a tariff game, that is, the situation where each country sets optimal tariffs given all other countries’ tariffs. This approach requires complex numerical simulations and is often limited to small samples of countries or sectors. The author’s innovation lies in the development of a “sufficient statistics” approach which radically simplifies the analysis. This approach is based on a numerical approximation which, in the author’s own words, “is strictly weaker than the small open economy assumption.” The result is a set of formulas that allow us to estimate the cost of a tariff war directly from a set of observable data: trade shares, trade elasticities by industry, and sector-level markup distortions. This approach is computationally light and allows for the inclusion of dozens of countries and sectors, covering the real complexity of today’s interconnected global economy.

Applying this methodology to data from the World Input-Output Database for the years 2000 to 2014, spanning 44 countries and 56 industries, the author finds that the average loss in real GDP from a global tariff war would be 2.8%. In absolute terms, this amounts to a global income loss of $1.7 trillion in 2014 alone, a figure equivalent to erasing the entire economy of South Korea from the world map. The prospective cost of a tariff war has more than doubled between 2000 and 2014. This sharp rise is attributed to two structural transformations: the increasing reliance on imported intermediate goods and the global rise in market power, reflected in higher markups.

As shown in the figure above, the cost associated with a potential tariff war increased steadily across all model variations, with the most pronounced effects seen when markup distortions and input trade are jointly considered. This suggests that the world economy has not only grown more interconnected, but also more exposed to inefficiencies that are magnified under protectionist policies.

The study offers important insights not just on the magnitude of the loss, but also on who loses the most. Countries that are small and heavily reliant on imported intermediate inputs, so-called “downstream economies”, face disproportionately large losses. Estonia, for instance, which imports inputs amounting to 30% of its total output, would experience a 10% drop in real GDP under a global tariff war. Similar levels of vulnerability are seen in countries like Bulgaria, Latvia, and Luxembourg. By contrast, economies like Norway, which are net exporters of upstream products such as oil and metals, can even gain from the imposition of global tariffs. These results complicate the traditional narrative that small countries are always losers in trade wars. Instead, the nature of a country’s integration into global value chains, whether upstream or downstream, becomes a more decisive factor.

The graph clearly illustrates how losses are not uniformly distributed: as dependence on imported inputs increases, so does the economic damage of a tariff war, rising exponentially for the most import-reliant countries. Luxembourg and Malta, for instance, appear particularly vulnerable on this front.

The methodology’s speed and simplicity are also noteworthy. While conventional methods can take hours or days to simulate the Nash equilibrium tariffs across countries, the author’s analytical framework produces results in seconds. This opens the door for policymakers and international institutions to conduct real-time assessments of trade policy risks using only basic data inputs. The method also makes it possible to explore alternative scenarios, such as the gains from cooperation.

Strikingly, the paper finds that just as the costs of conflict have risen, so too have the potential benefits of cooperation. Coordinated tariff policies that correct inefficient global markups could have boosted global GDP by $347 billion in 2014, nearly doubling the estimated gains from the year 2000. In a world where trade talks have stalled and protectionist pressures are rising, this finding underscores the vast economic potential of renewed multilateral engagement.

The study also contributes conceptually to the literature by showing how tariffs interact with existing market distortions. In particular, it demonstrates that in a world of rising markups, Nash tariffs, those set unilaterally by countries without cooperation, tend to disproportionately target high-markup industries. This leads to a contraction in output precisely in sectors that are already inefficiently small from a global welfare perspective, thus compounding pre-existing misallocations.

Moreover, standard modeling choices that aggregate the “rest of the world” into a single player significantly overstate tariff war losses by assigning too much bargaining power to large blocs. Thanks to the new methodology’s ability to work with disaggregated data, such distortions can now be avoided, further refining how we understand the distributional impacts of trade wars.

This approach reframes how we quantify the stakes of global trade policy. Rather than comparing a world with trade to one of autarky, the author focuses on a more realistic margin: the difference between multilateral cooperation and non-cooperative retaliation. This distinction is particularly relevant in the current geopolitical context, where tariff threats are often used as bargaining tools rather than measures of last resort.

In sum, this research offers policy makers a fast, flexible, and data-light toolkit to anticipate the economic consequences of trade conflict. It also reinforces the value of international cooperation—not merely as an abstract ideal, but as a measurable and rising economic opportunity.

Ahmad Lashkaripour

The Cost of a Global Tariff War: A Sufficient Statistics Approach

Journal of International Economics 131 (2021) 103419