An October working paper finds that most U.S. technologies it examined have a foreign precursor and estimates that counties exposed to more overseas ideas later produced more related inventions.
Boston University economics professor Tarek Alexander Hassan, Federal Reserve Bank of St. Louis economist Aakash Kalyani and Boston University economics doctoral student Juan Cruz Lopez Del Valle analyzed text from 13 million patents to map 19,542 technologies. The National Bureau of Economic Research (NBER) circulated their study as a working paper, a form of research released before peer review.
The researchers identified a precursor when an older technology appeared frequently in patents filed soon after a newer one emerged. They treated mentions in patents filed by local inventors as evidence that a county had begun working with a technology.
To measure earlier access to foreign ideas, they counted technologies mentioned in a county’s patents by 1980 when those patents did not mention the U.S. technologies later built on the ideas. They used historical migrant settlement patterns to estimate how access to knowledge from different countries affected subsequent invention.
The authors estimated that a 1% increase in this 1980 measure for a given foreign country corresponded to 1.4% more related U.S. technologies developed locally after 1980 through 2019.
In a separate analysis of about 350 million online job postings from 2010 through 2023, a 1% increase in the same measure was linked to 3.5% more local postings mentioning descendant U.S. technologies. That analysis included technologies that arose in 1980 or later. It counted advertised positions, not filled jobs.
The authors’ Oct. 1 paper says 56% of the U.S. technologies studied had at least one foreign precursor and places the start of its patent sample in 1894. NBER’s listing gives the figures as 54% and 1887.
The researchers assigned each technology to the first country whose inventors filed 10 U.S. patents mentioning it. They caution that this rule may misidentify origins where inventors were less likely to file U.S. patents. They also say greater innovation in one county may partly displace activity elsewhere, so the findings do not establish a net gain for the United States.
