Adam Smith at 250
Next year will mark the 250th anniversary of the ratification of the Declaration of Independence, the founding document of the United States. But another foundational document, essential to our understanding of economics, will reach the same milestone in 2026: The Wealth of Nations by Adam Smith. At a time of rapid economic and structural transformation, his ideas deserve to be revisited.
Two ideas stand out. One is that the “invisible hand” of markets allocates resources efficiently, provided that certain conditions are met —including a stable currency, a degree of trust and moral rectitude among economic actors, and credible property rights. Externalities (the unpriced impact of one entity’s activities on others) and information gaps and asymmetries reduce the efficiency and performance of the invisible hand.
The second idea, possibly more important, is that the efficiency and productivity of an economy are enhanced by the “division of labor,” known today as “specialization.” A specialized economy is driven by diverse clusters of knowledge and expertise, which leverage economies of scale, learning, and stronger incentives for innovation. Since specialization cannot function without a reasonably efficient method of exchange, it depends on Smith’s invisible hand. As specialization advances, so does the complexity of the economy.
However, as Smith noted, specialization is limited by the “extent of the market”: a small market cannot generate enough demand to sustain a wide variety of specialized businesses. That is why improvements in transportation and communications, which reduce the cost of accessing an expanding market, have enabled greater specialization.
Another important potential limit to specialization is the risk it inevitably creates. Because an economy’s patterns of specialization are structural, they take time to change. Thus, if the trading system is disrupted, or if certain skills or industries become obsolete (due to technological innovations or shifts in demand patterns), individuals, firms, and even entire economies must undergo a transition that can be difficult and prolonged.
In the 19th and 20th centuries, as economies became more specialized, policies, institutions, and conditions gradually emerged —from antitrust legislation to social safety nets to the maintenance of macroeconomic and monetary stability— to mitigate the associated risks. But these were largely national-level solutions, and after World War II, specialization became globalized.
What began as a means of supporting postwar economic recovery soon became a comprehensive transformation. Colonial empires were abandoned, along with their asymmetric economic structures, and mercantilism gave way to free trade. Added to this were advances in transportation and communication technology, accelerated by the digital revolution, and the first limit on specialization —the “extent of the market”— was radically relaxed.
For developing economies, this was a game changer. Given their low per capita GDP, they could not generate enough domestic demand to benefit from the efficiency and productivity gains of specialization. But once they gained access to foreign markets and technologies, they leveraged their comparative advantages and achieved rapid GDP growth. Thus, increasing specialization was accompanied by a geographical shift in economic activity.
The resulting structural disruptions outpaced the evolution of governance structures capable of mitigating growing risks. For a time, this did not seem to matter much: advanced economies, especially the United States, continued to finance international economic governance, setting the rules and sponsoring the institutions that kept the system running. But eventually, the shift in global economic power reached a tipping point: the demand constraint on specialization loosened to the extent that the risk constraint began to bind. As structural disruptions became more pronounced, popular frustration mounted in advanced economies, fueling a social and political backlash. Then, a growing series of shocks —the impacts of climate change, the COVID-19 pandemic, the wars in Ukraine and Gaza, and rising geopolitical tensions— reinforced this shift. Donald Trump’s return to the White House, with his “America First” foreign policy and preference for bilateral deals, cemented it.
As a result, many countries now see economic security as inseparably tied to national security: although specialization remains intact within economies, it is being partially reversed at the international level. While it is impossible to know exactly where this process will lead, adverse consequences for productivity and growth can be expected —in effect, the price of greater resilience and lower risk. Countries with less capacity to generate domestic demand —either because of low per capita GDP or small population size— will suffer more, and the magnitude of their losses will depend on how much access to global markets they retain.
But Smith’s model of specialization may soon face an even more fundamental shift. Recall that it is based on the creation of specific niches of knowledge and expertise that are not easily acquired or transferred. Yet generative artificial intelligence models, among their many effects, now seem on track to provide expertise in almost any area, to anyone who wants it, at very low cost.
The potential consequences are far-reaching. If expertise becomes less scarce, the price it commands will fall. Only knowledge and skills that remain difficult to transfer —for example, because they cannot be easily described or documented— will rise in value. In other words, a significant portion of human capital may not be worth as much in the future as it has been over the past 250 years, while another portion could be worth much more. A question that now needs to be explored is how large each of these portions will be.
Nearly 250 years after Smith introduced the concept of specialization, it remains a key feature of our economies. But it has also changed profoundly. It is partially receding in the global economy, as perceived risks of interdependence rise. Artificial intelligence is unlikely to reduce specialization, but by altering the equation of knowledge transfer, it could shift the relative value of the human capital associated with different types of specialized knowledge.
*Michael Spence, Nobel laureate in economics, is Professor Emeritus of Economics and former dean of the Stanford University Graduate School of Business, and co-author (with Mohamed A. El-Erian, Gordon Brown, and Reid Lidow) of Permacrisis: A Plan to Fix a Fractured World (Simon & Schuster, 2023).