As part of recent efforts to assess the economic consequences of a second Trump administration (Gensler et al. 2025), Olarreaga and Santander (2025) ask: who pays for the US tariffs? A key question in the context of a trade war is whether the burden of tariffs falls on foreign exporters or domestic importers.
This question brings us back to the broader concept of the incidence of cost shocks in international trade. Such shocks can affect exporters’ production costs, trade costs, or exchange rates. A natural concern for policymakers is how much of these foreign shocks are passed through to import prices – in other words, what their incidence is on domestic importers.
In a recent paper (Fontaine et al. 2023), we examine the incidence of cost shocks in international trade using a firm-to-firm trade model with search frictions.
International trade is supported by a network of firm-to-firm relationships. Leveraging new data on these networks, trade economists have developed models that incorporate such relationships and have proven useful in understanding the vulnerability and resilience of trade flows. However, in many of these models, firms’ pricing strategies are so stylised that they cannot be used to study the incidence of cost shocks. A recent exception is the work by Alviarez et al. (2023), which examines price negotiations between a seller and a buyer with differing market power. In their framework, trade relationships are taken as given, and the pass-through of shocks depends on the relative bargaining power of the two parties. We take a different approach, assuming search frictions. Buyers meet new suppliers infrequently which provides a reduced-form way to capture the various obstacles that complicate the formation of trade relationships and has implications for prices and trade flows.
In our model, each buyer knows only a limited number of trading partners. The limited strength of competition within these restricted pools has consequences for the price they are offered. Over time, however, buyers encounter new potential suppliers. Meeting a new supplier creates an opportunity to switch if the newcomer is more competitive than the current supplier, or to push prices downward if the new supplier forces the buyer’s existing supplier to reduce its markup.
The model allows us to replicate several trade patterns observed in international trade networks. First, this approach implies significant heterogeneity in prices and markups across trade relationships. The price charged by a supplier depends on the buyer’s outside options, which varies with the buyer’s experience in its search process. A natural corollary is that markups vary within a supplier, across its downstream partners. Such variation in prices across firm-to-firm relationships is consistent with evidence presented in our earlier work (Fontaine et al. 2020).
Second, prices tend to decline within trade relationships, over time. The reason is straightforward: as the buyer continues to meet new potential suppliers, its outside options expand, putting downward pressure on the current supplier’s price. This pattern is consistent with evidence recovered from our data, as well as evidence in other studies (Heise 2024, Monarch and Schmidt-Eisenlohr 2023).
Armed with this rich framework, we can explore how search frictions shape the incidence of cost shocks. As a thought experiment, we consider a scenario in which costs rise in a particular country. All foreign suppliers from that country are affected by the shock, and we examine how this impacts buyers who were initially matched with these suppliers.
The model shows that the shock can impose very different burdens on different buyers. Some buyers are not affected – their supplier absorbs the shock as they can’t raise price due to competition pressures exerted by the buyer’s alternative potential suppliers. In other cases, the supplier can’t absorb the shock, and the buyer needs to switch to its second-best supplier and pay a higher price. Last, the supplier’s direct competitor within the buyer’s network is also affected by the shock, enabling the supplier to raise its price. Moreover, because buyers continue to meet potential suppliers after the shock, the prevalence of these situations evolves over time.
Ultimately, the incidence on buyers is an empirical question and depends on the size of the shock, the level of search frictions, and the relative competitiveness of suppliers. We use our model, along with detailed data on the exports of French firms to their EU buyers, to estimate search frictions across sectors and markets. We then simulate the incidence on EU buyers of a 10% cost shock affecting French suppliers.
One month after the shock, we find an average 3.6% increase in the import price paid by EU buyers who were matched with a French supplier at the time of the shock, corresponding to an incidence of 36% borne by the buyer. However, this average conceals significant heterogeneity across different dimensions.
In the average market, 25% of buyers experience an incidence of 100%, while 40% are left unaffected by the shock. More than 25% of buyers switch to another supplier, incurring an average 30% price increase. We further show that the incidence varies across markets with different levels of friction: in markets in which French suppliers benefit from relatively higher matching rates, buyers suffer a larger price increase. In our estimates, the average incidence is twice as large in markets with low versus high frictions. Finally, the total incidence on buyers decreases over time, as does the dispersion of incidence across markets. This convergence is driven by the gradual reconfiguration of buyers’ sourcing networks: as search frictions diminish, firms initially reliant on French suppliers broaden their options and increasingly switch to non-French alternatives.
In response to shocks to foreign suppliers, governments are often willing to support domestic firms relying on imported inputs.
The findings of our study suggest that targeting programmes to firms simply based on their interactions with affected suppliers might not be efficient. Some buyers will not suffer any cost increase, while others will bear the full burden. One option would be to target segments (products x destinations) where a higher incidence on buyers is expected – such as those facing greater search frictions. Another option is to provide buyers with assistance in diversifying away from their supplier ex post. Buyers who face a higher burden would be more likely to participate in these programmes, while those who do not would opt out.
References
Alviarez, V I, M Fioretti, A K Kikkawa and M Morlacco (2023), “Two sided market power in firm to firm trade”, NBER Working Paper No. 31253.
Fontaine, F, J Martin and I Mejean (2023), “Frictions and Adjustments in Firm‑to‑Firm Trade”, CEPR Discussion Paper No. 18110.
Gensler, G, S Johnson, U Panizza and B Weder di Mauro (2025), “The economic consequences of the second Trump administration: A preliminary assessment”, VoxEU.org, 18 June.
Heise, S (2024), “Firm to Firm relationships and the pass through of shocks: Theory and evidence”, The Review of Economics and Statistics, Advance online publication.
Olarreaga, M and S Santander (2025), “Who pays for US tariffs?”, in G Gensler, S Johnson, U Panizza and B Weder di Mauro (eds), The economic consequences of the second Trump administration: A preliminary assessment, CEPR press.
Monarch, R and T Schmidt‑Eisenlohr (2023), “Longevity and the value of trade relationships”, Journal of International Economics 145: 103842.