Skip to main content

CentrePiece article

Research: Innovation

Return innovation: how migration shapes diffusion of new technologies

Davide M. Coluccia and Gaia Dossi


Novel products, processes and technologies are important drivers of economic growth. Davide Coluccia and Gaia Dossi show how migrants from Britain to the United States in the late 19th and early 20th century not only took new ideas with them, but also helped to bring American innovations back to the old country.

Steamships at sea with lighbulbs as the smoke from their funnels;
Steamships at sea with lighbulbs as the smoke from their funnels. Image credit: Raphael Whittle.

The diffusion of new technologies across countries is a key driver of productivity growth and economic progress. Yet surprisingly little is known about how technology moves across borders. Our research provides new evidence on the importance of one such channel: out-migration. We show that migration can help to transfer knowledge and innovation from migrants' destination countries back to their countries of origin - a phenomenon we label return innovation.

We focus on migration from England and Wales to the United States between 1850 and 1940. During this period – often called the age of mass migration – roughly 30 million Europeans crossed the Atlantic, including nearly four million from England and Wales.

We assemble two new large-scale datasets to investigate the importance of return innovation. First, we link confidential individual-level data from the American and British censuses, allowing us to follow Britons as they migrated to the United States. Second, to measure innovation in 19th-century Britain, we digitise all 300,000 original patent records issued in England and Wales between 1853 and 1899.

Identifying the causal impact of US innovation on British innovation is challenging, as migrants may choose to settle in areas that are already connected with innovation in their home districts - a pattern known as "assortative matching". For example, suppose that emigrants from Lancashire, which had a large textile sector, were comparatively more likely to settle in counties with larger textile sectors. Then the measured change in technological innovation in Lancashire could reflect pre-existing innovation similarities between the UK and US areas rather than capturing the effect of return innovation.

To address this concern, we use unexpected surges in US patenting - "innovation shocks" - as plausibly exogenous increases in exposure to innovation in the United States. We consider a district in the UK to be exposed to an innovation shock in the United States if the number of emigrants from that UK district that are exposed to an innovation shock is in the top 5% of the overall distribution of exposed emigrants, after adjusting for the effects of the county, year and type of technology. We then track the role of migrants in spreading these shocks back to Britain.

Areas of Britain with larger numbers of emigrants to the United States experienced a larger increase in patenting activity

We find that English and Welsh districts with larger numbers of people moving to the United States experienced a larger increase in patenting activity, especially in the same technological areas to which their emigrants were exposed in the United States. A back-of-the-envelope calculation suggests that following an unexpected surge in innovation activity in the United States, about 15% of this increase was transmitted to Britain via migrants' networks.

Similar starting points bring most benefit

To trace knowledge flows directly, we use methodologies of text analysis to compare the content of British and US patents. Patenting in districts with stronger migration links became significantly more similar to US patents, and this pattern holds even within specific technological fields.

But the impact of exposure to foreign innovation was not the same across all industries. We find that the largest increases in British patenting occurred in sectors where Britain and the United States had similar levels of specialisation, such as electricity, engines and scientific instruments. By contrast, the effect was smaller in fields where the United States was far ahead – such as agriculture and metallurgy – or where Britain led – such as textiles. This pattern aligns with recent theoretical work suggesting that countries benefit most from sharing knowledge with partners at similar levels of development.

In the second part of our analysis, we look at what drives the return innovation effect. The physical return of migrants played a significant role, but that accounts for only about half of the overall effect. The other half came from migration ties: even when migrants did not return home, their relationships with home communities helped to spread new ideas and knowledge.

We focus on two key ways in which these connections operated: through family ties and geographical proximity. Drawing on research in development economics that shows how technology spreads through social networks, we study whether emigrants influenced innovation activity among those they left behind. 
We find that family members of British emigrants to the United States became more likely to patent after their relatives had moved abroad. This effect takes about ten years to materialise, but its size is substantial. Return migrants had a larger impact than those who never returned, yet even emigrants who stayed abroad fostered innovation within their families.

We look at geographical proximity as another measure of local social connections. By linking British patent data with census records, we show that when people’s neighbours moved to the United States, those who stayed behind became more likely to file patents themselves. This effect is strong even after excluding neighbours of return migrants, highlighting the importance of continuing cross-country connections in spreading new ideas. Our findings shed new light on how knowledge and innovation move across borders. Migration plays a key role in this process: when people move abroad, they don’t just take skills with them – they also help bring new ideas back home. In short, migration can promote innovation by fostering the flow of knowledge into migrants' countries of origin.

Larger increases in patenting were seen in sectors that were at similar levels of development in Britain and the United States, such as electricity and engines

Bringing tech back

Henry Rowland Marsden was born in Leeds to poor parents in 1823 (Curtis, 1875). At 25, he emigrated to the United States, first to New York and then to Connecticut. There, he took on apprenticeships in engineering and metal-working firms. He obtained several engineering patents – chiefly related to steam engines and pumps, including a "stone-crusher" that is still in use today. In 1862, Marsden returned to Leeds, where he set up a flourishing business centred around his newly patented inventions. A wealthy man respected for his philanthropic endeavours, he was elected mayor of Leeds in 1873.

An 1856 article published in Scientific American described a new patent granted in the UK to Henry Bessemer (Wagner, 2008). Bessemer had discovered a new process that, for the first time, allowed the production of inexpensive steel from molten pig iron. The eponymous Bessemer process has been described as one of the most transformative technological developments of the 19th century (Rosenberg and Trajtenberg, 2004). But American inventor William Kelly complained: "I have reason to believe my discovery was known in England three or four years ago, as a number of English puddlers [a specialist iron worker] visited this place to see my new process. Several of them have since returned to England and may have spoken of my invention there." (Wagner, 2008). The veracity of Kelly’s allegations remains unverified.

They nonetheless indicate three important elements. First, American inventors knew that British immigrants posed a threat to the secrecy of their inventions. Second, technology transfer did not depend on the most highly-educated people. Skilled workers, such as puddlers, could be the agents of technology diffusion. Finally, the precise mechanism that emerges is return migration. Kelly expected British puddlers to speak of "his" invention on returning to England.


20 June 2025     Paper Number CEPCP708

Download PDF - Return innovation: how migration shapes diffusion of new technologies

Download Press Release

This CentrePiece article is published under the centre's Growth programme.

This publication comes under the following theme: Innovation drivers