The Paper | Global Mobile Inventors

The Story

When Inventors Cross Borders

The first patent ever filed in Chile for medical devices was filed in 1996 by Ismael Mena. Seventeen more patents came after that, produced by 19 other inventors. Interestingly, Mena had filed a patent in that same technology in 1994 while living in the United States.

Is Mena's experience as an inventor in the US what helped him develop the necessary knowledge to file the first-ever patent on medical devices in Chile? Are inventors like Ismael Mena, moving across borders, systematically diffusing new technologies as they move?

This paper answers that question. We study a phenomenon we call Global Mobile Inventors, or GMIs: inventors who patent in multiple countries over their career, potentially facilitating the emergence of new technologies in the countries they arrive to.

6.1M

Granted patents analyzed, belonging to 3.4 million inventors across 200+ countries (1970–2015)

Trends

The Rise of Mobile Inventors

A once-rare phenomenon has become a defining feature of global innovation.

The mobility of inventors across borders has grown dramatically over the past decades, going from a few hundred movements in the 1970s to about 20,000 per year by the mid-2010s. The vast majority of these movements take place outside the boundaries of a single firm.

Figure 1a. The number of GMI inventors (solid) and patents by GMIs (dashed) has grown exponentially since the 1970s. By 2015, nearly 30,000 inventors had moved internationally at some point in their careers, producing about 70,000 patents.

Figure 1b. While fewer than 0.5% of patents in the early 1970s involved a GMI, by 2015 roughly one in three patents involved at least one GMI (using the "always" definition). The GMI phenomenon has grown faster than Global Collaborative Patents (GCPs).

Beyond growth, the geography of inventor mobility has shifted dramatically. In the 1980s and 1990s, the largest flows were between Western countries and Japan, with the United States as the focal point. Since the 2000s, emerging markets — particularly China, South Korea, and India — have entered the top corridors of inventor mobility.

Figure 1c. Top 10 corridors of inventor mobility over time. The US–China corridor rose from outside the top 10 to become the single largest by 2011–2015. Emerging markets increasingly dominate the landscape.

Characteristics

GMIs Are Superstars

Mobile inventors are not ordinary — they are among the most productive and high-quality inventors in the world.

0

Average patents per GMI

vs 3.94 for non-GMIs

0

Years active career

vs 4.64 for non-GMIs

0

In top 100 patenting firms

vs 21% for non-GMIs

0

Countries patented in

vs 1.0 for non-GMIs

0

Firms (assignees) per GMI

vs 1.24 for non-GMIs

0

5-year forward citations

vs 6.52 for non-GMIs

GMIs patent significantly more, work at top firms, have longer careers, and produce higher-quality patents. They are almost three times as likely to participate in Global Collaborative Patents as non-GMIs. These are not ordinary movers — they are among the few inventors capable of transferring knowledge across borders and pioneering new technology life-cycles in receiving countries.

Source: Table 1 in the paper. All differences are statistically significant at the 1% level. Sample: 97,192 GMIs vs. 3,391,954 non-GMIs.

Key Finding

GMIs Pioneer New Technologies

The central result: experienced GMIs are disproportionately present in the earliest patents of a new technology in a country.

2×

Patents in the first decile (earliest 10%) of a technology's life cycle in a country are over twice as likely to involve a GMI with prior experience in that technology, compared to the most recent patents in the same country-technology pair.

Figure 2. Probability of observing an experienced GMI across deciles of the technology life-cycle. The first decile (earliest 10% of patents) shows ~0.5 percentage point higher probability compared to the 10th decile (baseline). With an unconditional probability of 0.7%, this implies a 70% higher likelihood. The effect fades after the 3rd–4th decile, suggesting knowledge is absorbed by local inventors over time.

This pattern is significantly stronger for non-OECD countries — those further from the technological frontier. In these countries, there is a 3 percentage point higher probability that a first-decile patent involves an experienced GMI, compared to just 1 percentage point for OECD nations.

Figure 2b. The same analysis split by OECD vs. non-OECD countries. The effect is much more pronounced for non-OECD countries (cyan), consistent with the idea that GMIs are particularly valuable for technology diffusion in countries further from the frontier.

Crucially, the effect is driven entirely by GMIs with prior experience in the specific technology. GMIs without prior experience in the technology class show no elevated presence in early deciles. This confirms the mechanism is about knowledge transfer, not simply that mobile inventors are more productive overall.

Mechanisms

Who Drives Diffusion?

Not all GMIs are equal. Returnees, complex technologies, and network centrality all matter.

The paper uncovers several important dimensions of heterogeneity that shed light on the mechanisms behind knowledge diffusion:

2×

Returnees are twice as effective as immigrants at introducing new technologies

Complex

The effect is driven by complex technologies that require more tacit knowledge

Post-1990

GMIs are particularly effective after the 1990s, as emerging markets join the global innovation network

Network centrality matters. The paper introduces a novel measure of "local absorptive capacity" — the speed at which a technology becomes embedded in local inventors. It finds that when GMIs are more central in the local co-patenting network (i.e., better connected to local inventors), the technology is absorbed faster. This is especially true when measured by eigenvector centrality, which weights connections by how central the collaborators themselves are.

In other words: GMIs are most effective as knowledge brokers when they are well-integrated into the local innovation ecosystem, not just when they are present. You can explore the co-patenting networks in the Data Explorer.

Implications

What This Means

Our results speak to human mobility as a central determinant of economic growth through innovation. Countries can benefit from encouraging scientists and inventors to move and bring with them their accumulated knowledge and experience, to foster the diffusion of knowledge, which is key to foster innovation and productivity.

The integration of newcomers is crucial, as team collaboration is critical to facilitate the dissemination of knowledge to local inventors. The finding that returnees — inventors returning home after time abroad — are particularly effective knowledge brokers is especially relevant for developing countries, where the main returnee corridors concern inventors returning to China, Korea, and India after a stay in the United States.

Our result that the local capacity of inventors to continue producing knowledge in technologies that originally arrived with GMIs shows that this process is sustainable: GMIs seem to ignite innovation, but local inventors are the ones that maintain the production of knowledge over time.

Explore the Data Read the Full Paper