How Technology Diffuses Across Borders

In 1719, the British Parliament passed legislation making it a criminal offense to induce skilled artisans to emigrate. The penalty was forfeiture of property and permanent banishment. This was not the first British attempt to keep industrial secrets at home — the Statute of Artificers had been regulating labor mobility in various forms since 1563 — but the 1719 act represents the point at which Britain explicitly recognized that its industrial advantage was embodied not in machinery alone but in the people who knew how to build, operate, and maintain that machinery. Preventing technology transfer, the legislation implicitly acknowledged, meant preventing skilled workers from leaving. The machinery could be described in patents and diagrams, but the knowledge of how to make it work existed primarily in human heads, and human heads could walk across borders.

The legislation failed. Not spectacularly or immediately — it created real deterrents and real costs for the workers who violated it and the foreign governments that recruited them — but comprehensively, over the span of the decades during which the Industrial Revolution matured. Belgium industrialized first among Continental European nations, and it did so primarily by systematically recruiting British skilled workers despite the prohibition. Liège became an early center of the Belgian coal and metal industries in part because William Cockerill, an English mechanic who had emigrated in 1799, established machine shops there that trained a generation of Belgian industrial workers in British techniques. His son John Cockerill expanded the operation into one of the largest integrated iron and machinery firms in Europe. The British government prosecuted what cases it could, stripped emigrants of their property rights when it identified them, and occasionally prevailed upon foreign governments to refuse them entry. None of this stopped the process. The economic incentives for skilled workers to emigrate were simply too large.

The economic logic of the British prohibition was sound in theory and impossible to enforce in practice. Industrial knowledge in the early nineteenth century was tacit knowledge — knowledge that resided in the judgment and muscle memory of craftsmen rather than in codified form. A cotton spinner who had spent years mastering the feel of a mule spinning machine understood things about fiber tension, spindle speed, and humidity effects that no manual captured and no patent described. This tacit dimension was precisely why skilled workers were valuable to foreign industrialists: they brought not just the abstract knowledge that a machine existed, but the practical understanding of how to make it perform. And tacit knowledge, by its nature, cannot be confiscated at the border. It travels with its possessor.

Germany’s acquisition of British industrial technology followed a slightly different pattern that reveals a second mechanism of technology diffusion: the importation of capital equipment along with the engineers who could operate it. In the 1830s and 1840s, Prussian industrialists in the Ruhr Valley imported British steam engines, textile machinery, and puddling furnaces with the workers who understood them. The Prussian government actively facilitated this, funding technical education missions to Britain, subsidizing the travel costs of skilled workers willing to emigrate, and establishing the Gewerbeinstitut in Berlin in 1821 as an institution specifically devoted to studying British industrial methods and disseminating them domestically. This was not industrial espionage in the covert sense — it was state-sponsored technology acquisition, conducted openly enough that the British Foreign Office protested it repeatedly and ineffectively.

What the German case illustrates is that the full package of industrial technology is not just machinery and not just skilled workers, but a complex system in which machinery, workers, management practices, organizational forms, and supporting institutions all interact. The Ruhr industrialists who imported British steam engines in the 1830s were buying not just power sources but templates for a way of organizing production that was alien to the guild-based craft industries that had previously dominated German manufacturing. The steam engine came bundled with implicit knowledge about how a factory should be laid out, how shifts should be organized, how quality should be monitored. Acquiring the machine without the accompanying organizational template would have been only partially useful. Acquiring the machine with British workers who embodied that template accelerated the learning process enormously.

This insight — that technology transfer requires the transfer of an entire sociotechnical system, not just its most visible components — explains why technology diffusion consistently takes longer than optimistic estimates predict. When the Japanese government launched its modernization program after the Meiji Restoration of 1868, it sent thousands of students to European and American universities, hired hundreds of foreign technical experts to work in Japan, imported complete factory systems, and established its own technical universities. This was the most systematic state-sponsored technology acquisition program in history to that point. And yet full convergence of Japanese industrial productivity with Western levels took roughly a century. The students came back with knowledge; the factories were built; the machinery worked. But building the managerial class, the engineering culture, the network of supplier relationships, the financial institutions comfortable with industrial investment — these took generations.

The historical experience of technology diffusion also reveals something counterintuitive about the role of intellectual property protection. The standard economic argument for patents is that they encourage innovation by giving inventors temporary monopoly profits, and that once the patent expires, the knowledge becomes freely available. The implicit assumption is that disclosure in the patent makes the knowledge accessible to anyone who reads it. The history of early industrialization suggests this assumption is significantly wrong for complex physical processes. British patents for spinning and weaving machinery in the late eighteenth century were freely available to anyone who visited the Patent Office in London. Continental industrialists read them. The patents did not enable successful replication, because the patents described the finished machine without describing the years of iterative refinement that had made the machine reliable and productive. The tacit knowledge gap persisted even when the formal knowledge was fully disclosed.

This gap between codified and tacit knowledge is what makes technology transfer in developing economies consistently slower and more expensive than import-substitution theorists have historically predicted. The industrialization programs of the 1950s and 1960s — in India, Egypt, Brazil, Mexico, and across the developing world — were premised on the idea that acquiring capital equipment and building factories would be sufficient to industrialize. The factories were built. The productivity convergence was disappointing. In many cases, the capital equipment imported from advanced economies required maintenance and operational expertise that took decades to develop domestically, and in the interim operated at a fraction of its designed productivity because the tacit knowledge required to run it well had not been transferred. The equipment was present. The knowledge system within which the equipment made sense was absent.

The most effective mechanism for bridging the tacit knowledge gap, across all historical episodes, is the movement of people. Not textbooks, not patents, not even formal technical training — though all of these help — but the direct movement of people who embody the knowledge from locations where it exists to locations where it doesn’t. William Cockerill’s emigration to Belgium was more economically consequential than any number of Belgian industrial espionage missions could have been, because Cockerill brought not just the knowledge of what British machines looked like but the knowledge of how to make them, run them, and train others to do the same. The difference between knowing that mule spinning machines exist and knowing how to build a mule spinning machine that actually works at scale is the difference between a diagram and a decade of experience.

The policy implications are uncomfortable for governments that care about national economic competitiveness. The most effective technology policy is also the most politically sensitive: allowing the free movement of technically skilled people across borders, in both directions. Countries that allow talented engineers and entrepreneurs to enter freely, and that are willing to send their own students and researchers to learn at the leading institutions of other countries, consistently accumulate the tacit knowledge that converts formal technological knowledge into productive industrial capacity. Countries that restrict inflows of skilled foreigners and limit the outward movement of students — whether out of security concerns, nativist sentiment, or mercantilist anxiety about brain drain — consistently find that their access to the tacit knowledge embedded in the global technical community is correspondingly limited. Britain’s attempt to wall off its industrial knowledge in 1719 failed completely. Every subsequent attempt to do the same thing, by any nation, has also failed. The knowledge moves with the people. The people move toward opportunity. The only question is which economies make themselves sufficiently attractive to the people who carry the world’s most productive ideas.

The diffusion of industrial technology from Britain to the Continent also produced a result that Britain’s protectionists had most feared: it created competitors. Belgian steel eventually competed with British steel. German chemicals and electrical equipment surpassed British equivalents by the late nineteenth century. The technology that Britain had tried to keep proprietary became the foundation for industries that outcompeted Britain’s own. This outcome is not evidence that the protectionists were right to try — the diffusion would have happened regardless, and slower diffusion would simply have meant slower global growth without preserving British industrial dominance. It is evidence that the competitive advantage of being first is always temporary when the knowledge that creates it can be learned, and that the most durable competitive advantage is not proprietary knowledge but the institutional and human capital capacity to generate new knowledge faster than competitors can absorb old knowledge. Germany ultimately surpassed Britain in chemicals not because it copied British chemistry but because it built research universities and industrial research laboratories that generated new knowledge continuously. The lesson from the history of technology diffusion is not that keeping secrets is impossible, which it is, but that generating new knowledge is a more durable advantage than holding old knowledge — and that the countries that understood this earliest built the most durable industrial leadership.