---
title: "Shipping Cost Shocks and US Producer Prices: Normal versus Crisis Pass-through"
authors: "Simon Gilchrist; Bin Wei; Vivian Z. Yue; Egon Zakrajsek"
source: "Federal Reserve Bank of Atlanta (Working Paper 2026-13)"
source_url: "https://www.atlantafed.org/research-and-data/publications/working-papers/2026/09/21/13-shipping-cost-shocks-and-us-producer-prices"
paper_id: "Atlanta Fed WP 2026-13"
harvested: 2026-09-28
year: 2026
keywords: [supply chain disruption, shipping costs, inflation, pass-through, producer prices, state dependence, local projections, menu costs]
pdf_url: "https://www.atlantafed.org/-/media/Project/Atlanta/FRBA/Documents/research-and-data/publications/working-papers/2026/09/21/13-shipping-cost-shocks-and-us-producer-prices.pdf"
---
## Abstract

This paper examines how shipping costs pass through to US producer prices. We link confidential shipment-level import transactions from the Census Bureau's Longitudinal Firm Trade Transactions Database (LFTTD) to confidential firm-level producer prices from the US Bureau of Labor Statistics' Producer Price Index (PPI) microdata, constructing a monthly panel of more than 6,000 US manufacturing importers from 2005 to 2022. Using local projections and Bartik-style instruments based on route-level shipping-cost variation, we find that pass-through to producer prices is negligible in normal times but rises sharply during the COVID-19 pandemic, when shipping costs surged broadly across routes and source countries. We interpret this state dependence through a menu-cost model with imported intermediate inputs and endogenous substitution between imported and domestic inputs.

JEL classification: E31, F14, F41, L11. Key words: supply chain disruption, shipping costs, inflation, pass-through, producer prices. DOI: 10.29338/wp2026-13

## Why it's in the StratMac library

- **The state-dependence result is the finding SM's shock-family work needs.** Cost pass-through into producer prices is negligible in normal times and roughly ten times larger in crisis — so a supply-cost series carries almost no ordinary-times signal and a large crisis-times signal. That is a direct argument for a *conditional* reading rather than a linear coefficient on a shipping-cost or freight input.
- **Explains the asymmetry the staples/commodities rescaled cards can otherwise misread.** The mechanism is explicit: pass-through is cumulative as firms reprice under menu costs, and broad (many routes at once) rather than route-specific. Narrow, idiosyncratic freight moves and broad systemic ones are different objects.
- **Elite authorship on the inflation-transmission literature** (Gilchrist; Zakrajsek of Boston Fed and the financial-frictions literature) and a genuinely primary data construction: confidential LFTTD shipments linked to BLS PPI microdata, 6,000+ manufacturing importers, 2005-2022.
- Prepared for the Carnegie-Rochester-NYU Conference Series on Public Policy. Supports the `production`, `commodities` and `staples` sections.