What Did the 2018 Tariffs Do to US Supply Chains? Firm-Level Evidence

TL;DR: In a working paper with Marcelo Villena, we study how the US tariffs imposed on China during 2018Q3 changed the operations of multinational US importers. Using transaction-level import data at the shipment level (2015–2019) and a difference-in-differences design, we find the tariff war caused firms to add about 1.5 more source countries, cut supply from their main source country by 0.3%, and increase lead time by 0.4%.

Why focus on 2018Q3?

During the third quarter of 2018, the Trump administration imposed three waves of import tariffs on Chinese goods — July 6, August 8, and September 18 — covering approximately $250 billion of US imports at rates between 10% and 50%, following the USTR's Section 301 investigation into technology transfer and intellectual property practices. That concentration of a sharp rate increase in a brief window is what makes the episode attractive econometrically: it behaves like a policy shock. We use data between 2015 and 2019 to keep the COVID-19 pandemic out of the sample.

In the aggregate, import values of tariffed products fell 25–30% after imposition, implying at least $136 billion of trade was redirected (Alfaro & Chor, 2023) — a substantial shock to global supply chains.

What does the paper measure?

The paper works with transaction-level import data at the shipment level for imports into the United States. That granularity lets us see, for each importer, where cargo comes from, how sourcing shifts over time, and how long shipments take — rather than inferring from country aggregates.

What did we find?

• Diversification: companies reduced the share of imports from China and increased imports from other Asian countries such as India and Vietnam, and from North American partners such as Canada and Mexico.
• Causal estimates (difference-in-differences): the tariff increase produced about 1.5 additional source countries per importer, a 0.3% reduction in supply from the main source country, and a 0.4% increase in lead time.
• Performance: alongside longer lead times, we observe a negative effect on the companies' financial results.

Why does this matter?

Diversifying away from a dominant supplier country buys resilience, but it is not free: more sources mean longer, more complex logistics, and the data show that cost showing up in lead times and performance. This connects to my related work on the environmental side of the same restructuring.

The full text, citation, and BibTeX are on the publication page.