When cheap imports arrive, do domestic firms innovate their way out or give up?
Foreign Competition and Domestic Innovation: Evidence from US Patents
David Autor et al. (2020) · American Economic Review: Insights 2(3)
Article ↗ · Replication data ↗ (openICPSR)
The archive requires an account we do not hold, so the files were not opened. What follows rests on the published abstract and on this platform's own observations.
What the authors found
Matching three decades of US patents to the firms that own them, Autor, Dorn, Hanson, Pisano and Shu find that rising import exposure from China reduced patenting by the exposed firms. Competitive pressure cut their sales and profitability, and innovation fell with them. The optimistic mechanism — that competition spurs firms to invent — does not show up in the firm-level data.
What our data adds
The national aggregates in this database cannot see that, and the way they fail is the instructive part. US resident patent applications rose from 149,251 in 1999 to 288,335 in 2015 before easing to 269,586 in 2020 — growth, not the decline the paper documents. Over the same period Chinese resident applications went from 15,626 to 1,344,817, an eighty-six-fold rise. Both facts are true at once. The paper's result is about which firms patented, and a national total sums the firms that retreated together with the firms that never faced the competition, so the composition change the authors identified is invisible by construction.
The aggregate series the note discusses. It cannot show the firm-level composition change the paper identifies.
| Figure | Value | Indicator |
|---|---|---|
| US resident patent applications, 1999 | 149,251 | IP.PAT.RESD |
| US resident patent applications, 2015 | 288,335 | IP.PAT.RESD |
| US resident patent applications, 2020 | 269,586 | IP.PAT.RESD |
| China resident applications, 2020 | 1,344,817 | IP.PAT.RESD |
Every figure here is computed from observations in this database and can be charted, downloaded or queried.
What this cannot show
Nothing here tests the paper's claim, and it should not be read as weighing against it. A country-level count cannot support or refute a firm-level causal result: identifying that effect required matching patents to corporate owners and to their industry's import exposure, which is exactly the structure a national aggregate discards. Patent counts are also a poor measure of innovation, and the US and Chinese figures are not comparable in quality — filing incentives, subsidies and office practice differ enough that the eighty-six-fold ratio overstates any underlying gap in invention.