Why the Same Technology Produces Different Outcomes in Different Societies

In 1485, the Ottoman Sultan Bayezid II issued a firman — an imperial decree — banning the printing press throughout his empire. The timing is notable: Johannes Gutenberg had died just seventeen years earlier, and the press had already reshaped European intellectual and religious life with startling speed. Within decades of its introduction, cheap printed texts had made Luther possible, undermined the Church’s monopoly on biblical interpretation, and flooded European cities with pamphlets, treatises, and arguments that no authority could effectively suppress. The Ottomans watched all of this happen from close range — and then, for more than two centuries, declined to adopt the technology for texts in Arabic script. The ban was finally lifted in 1727, under very different political circumstances, and even then the first Ottoman press operated under strict censorship controls. A technology that had been destabilizing and radicalizing in Europe was — temporarily, and with considerable effort — kept stabilizing and marginal in the Ottoman world.

The standard reading of this episode treats it as a story about Ottoman backwardness: evidence of a conservative, religiously constrained society failing to capitalize on a transformative technology. That reading is wrong. What the printing press episode actually illustrates is something more interesting and more generalizable: the same technology, introduced into different institutional environments, produces fundamentally different outcomes. The technology didn’t change. The societies did.

Why Institutions Determine Technological Outcomes

Technologies are not self-interpreting. A printing press doesn’t intrinsically destabilize authority — it does so only when there are literate populations hungry for unauthorized information, weak centralized censorship mechanisms, fragmented political authority that makes suppression logistically difficult, and an existing intellectual culture of disputation that printed texts can accelerate. Sixteenth-century Europe had all of these features. The Ottoman Empire in the same period had almost none of them.

The Ottoman state in 1485 was, by the standards of the time, remarkably well-organized and administratively capable. The Sultan’s court maintained sophisticated bureaucratic control over information flows. The ulema — the Islamic scholarly class — held a monopoly on authoritative textual interpretation that was structurally dependent on the scarcity and difficulty of manuscript reproduction. Printing threatened not just political authority but an entire economy of religious credentialing built around the labor of copying and authenticating texts. The ban on the printing press was not a failure of technological understanding. It was a rational response to an accurate threat assessment.

The same dynamic appears across the history of technology adoption. The Tokugawa shogunate in Japan maintained strict limits on firearms for more than two centuries after their introduction, not because Japanese smiths couldn’t make guns — they demonstrably could, and did, in large numbers during the civil wars of the sixteenth century — but because the samurai class’s political dominance depended on a monopoly of martial skill that firearms threatened to democratize. Guns were regulated out of existence because the institutional beneficiaries of existing power arrangements had the political capacity to suppress them. The technology was available. The incentive structure made adoption costly for those with the power to prevent it.

The British Industrial Revolution Was Not Inevitable

The standard account of the British Industrial Revolution presents it as a story of technological genius: Watt and the steam engine, Arkwright and the water frame, Kay and the flying shuttle. This account is not wrong, but it is radically incomplete. The inventions were real. But inventions happen everywhere. The question is not why Britain invented things, but why Britain systematically exploited its inventions in ways that transformed the production process, while comparable inventions elsewhere — and there were many — produced nothing like the same economic transformation.

The answer is institutional. Britain in the eighteenth century had a cluster of reinforcing features that made technological adoption commercially rewarding in ways that other societies didn’t. Secure property rights, including patents, meant that inventors could expect to capture at least some of the returns from their innovations. A relatively open labor market meant that displaced craft workers couldn’t suppress new machinery the way guild-protected artisans could in France and Germany. Access to colonial markets provided the scale needed to make capital-intensive production profitable. A financial system sophisticated enough to intermediate between savers and industrial investors meant that promising technologies could be backed before they were proven.

Remove any one of these institutional features and the Industrial Revolution looks different — probably slower, possibly never. France had the technology and the science, arguably better in some respects than Britain’s, and a larger domestic market. France did not industrialize on the British timeline. The difference was not in the minds of French engineers. It was in the institutional environment that determined what happened when those engineers went looking for capital, markets, and legal protection for their ideas.

Why the Same Technology Fails Twice

One of the most reliable patterns in technology adoption history is the repeated failure of transplanted technologies — technologies that worked brilliantly in one society being imported by another and producing disappointing or destructive results. Development economists spent much of the second half of the twentieth century puzzling over this: why did the Green Revolution produce export growth and reduced hunger in some Asian countries while failing to improve food security in parts of Africa? Why did structural adjustment programs — built on the same economic logic that had guided successful market transitions elsewhere — produce such variable results across different economies?

The Green Revolution case is particularly instructive. The new high-yielding grain varieties developed at CIMMYT and IRRI in the 1960s represented genuine technological advances: more productive, more reliable, more resistant to various stresses than the traditional varieties they replaced. In the Punjab — both Indian and Pakistani — they produced exactly the transformation promised: yields doubled, then tripled, rural incomes rose, and what had been a food-insecure region became a grain surplus producer. In sub-Saharan Africa, applied to similar-seeming agronomic problems, the results were far more modest. The difference was not primarily the soils or the rainfall, though those varied. The difference was land tenure structures, input distribution systems, credit access, price support mechanisms, and the political economy of smallholder farming — all of which varied enormously between South Asia and Africa and all of which determined whether a technological package designed for one context was usable in another.

This matters because the technology-first approach to development — send engineers, train technicians, transfer knowledge — systematically underweights the institutional preconditions for technology to do what it’s supposed to do. Technology is not a solvent that dissolves institutional barriers on contact. It is more like a catalyst: it accelerates reactions that are already primed to occur and does almost nothing to reactions that aren’t.

The Digital Age Recapitulates the Pattern

The internet was supposed to be different. Unlike previous communications technologies that concentrated power in whoever controlled the infrastructure — printing presses, telegraph networks, broadcast towers — the internet was architecturally decentralized, and its decentralization was supposed to make information suppression structurally impossible. Stewart Brand’s famous formulation — “information wants to be free” — captured a genuinely new technological reality and turned it into a prediction about political outcomes that has proven considerably more contingent than its proponents expected.

China’s experience with the internet is the Ottoman printing press episode recapitulated at massive scale and with vastly greater sophistication. The Chinese Communist Party confronted a technology that had demonstrably destabilized authoritarian governments in other contexts — the Arab Spring, color revolutions in post-Soviet states — and proceeded to build what is now the most sophisticated information control system in human history. The Great Firewall is not a simple block list. It is a dynamic, AI-assisted censorship apparatus capable of selectively permitting economic use of digital technology while suppressing political use. It has worked considerably better than most Western observers predicted, which suggests that the prediction was based on a misunderstanding of the relationship between technology and institutions.

The same pattern appears in the economic effects of digital technology across different societies. Platform economies that generated broadly distributed productivity gains in some economies produced winner-take-all concentration in others. Gig economy models that raised incomes for workers in some labor markets produced precariousness in others. The technology was the same. The outcomes were determined by labor market institutions, social safety nets, competition policy, and the underlying bargaining power of workers — all of which varied by jurisdiction.

The Lesson That Keeps Being Ignored

The persistence of the technology-as-transformer fallacy is itself worth explaining. Part of the answer is that the fallacy serves commercial interests: technology companies benefit enormously from the narrative that their products are forces of nature rather than tools whose effects depend on social context. The techno-determinist view — the idea that technology has inherent and predictable social consequences — licenses a kind of political irresponsibility in which the designer of a technology bears no responsibility for how it is deployed, because the deployment is inevitable.

The deeper answer is that technology-first thinking is genuinely more comfortable than institutions-first thinking. Technology is visible, transferable, and improvable in ways that institutions are not. You can ship a printing press. You cannot ship the English common law tradition, the Reformation’s disruption of ecclesiastical authority, or the political fragmentation that made censorship logistically impossible in early modern Europe. If outcomes depend primarily on institutions, then improving outcomes requires changing institutions — a slower, more contested, more uncertain project than building better tools.

Bayezid II’s ban on the printing press held for two centuries. When the press finally arrived in the Ottoman world, it arrived into a different institutional context — one shaped by military decline, fiscal crisis, and a reforming bureaucracy that needed the press for administrative reasons and was willing to accept the risk of printed dissent. The technology was the same technology that had transformed Europe two hundred years earlier. What changed were the institutions that received it. That sequence — institutions shaping technology, rather than technology reshaping institutions — is the honest history of every major technological transformation in the record. We ignore it because it makes progress seem slower and harder than we’d like. It doesn’t make it less true.