The idea is especially trendy in the United States, where such grand government designs were traditionally frowned upon—at least in theory and except in emergencies, such as wars or the New Deal. Now, both Democrats and Republicans are fans, even if the Trump administration does not know the proper words to label its social and economic engineering intuitions, from equity participations in private companies to veiled threats against companies who don’t pursue the right “public goals.”

Is industrial policy an essential tool for prosperity, as the fast economic growth in some East Asian countries may suggest? Three recent academic papers may help us answer that question.

Rodrik’s Mercantilism

The first, by Rodrik, is currently a working paper titled “What the Mercantilists Got Right.” Mercantilism, the Encyclopedia Britannica tells us, is the “economic theory and practice common in Europe from the 16th to the 18th century that promoted governmental regulation of a nation’s economy for the purpose of augmenting state power at the expense of rival national powers.” Criticized by Adam Smith and other classical liberal economists, mercantilism was the industrial policy of the time.

In his paper, Rodrik argues that mercantilism does have some good insights. The first one is the primacy of jobs and production over consumption. The second is the benefit of a close, cooperative collaboration between business and government. Third, mercantilists see economic policy as a matter of pragmatism instead of principle.

Each of these ideas is highly questionable.

Take the idea that production and jobs are somehow more important than consumption. Some opinion surveys suggest that satisfaction with life largely comes from the “meaning” of work, but try paying workers only with meaning (and meaning-and-a-half for overtime) and they will not stay on the job very long. A recurring problem of the welfare state stems from the reluctance of people to work if they are paid when they don’t. Consider 1967, around the time President Lyndon Johnson’s “Great Society” got rolling: In the lowest quintile of household income, two-thirds of able-bodied, working-age individuals not studying full-time had a job; 50 years later, the proportion was one-third. If people can consume without working, the lure of leisure is irresistible.

Another argument of Rodrik for the primacy of work is that it gives people a feeling of contributing to the community. But why wouldn’t they get the same feeling by consuming in their community? It is true that most people enjoy some aspects of their work independent of their remuneration; in other words, one’s job has a consumption component. Academics, intellectuals, and artists who are lucky enough to have highly satisfying jobs—jobs with a larger consumption component—often have a problem understanding that not everybody is so lucky.

If work had primacy over consumption, consumers would have to beg producers to sell, instead of producers having to persuade consumers to buy. French philosopher Raymond Ruyer, in his 1969 book Éloge de la société de consommation (In Praise of the Consumer Society), described the difference between a market economy, where the consumer is sovereign, and a planned economy, where the producer runs the show (under government’s control): “In a market economy, demand is imperative and supply is supplicant.… In a planned economy, supply is imperative and demand is supplicant.”

In a society dominated by producers, factors such as “technological spillovers,” “economies of scale,” and foreign competition naturally call for government intervention. In a system where individuals pursue their happiness by consuming what they want—more generally by increasing their “utility,” as economists say—private property rights are more important and government meddling more difficult to justify.

The second proposed insight of mercantilism is that business and government must cooperate in “the public interest,” as opposed to having an arm’s-length relationship under the impersonal rule of law. The industrial policies of fast-growing East Asian countries supposedly showed the benefit of such cooperation. There, as later in China, the government “played the role of a ‘collaborative catalyst’ rather than a central planner,” Rodrik suggests, following a recent book on “developmental environmentalism.” But who determines the public interest; whose private interests will be deemed “public”? Ultimately, politicians, party members, and bureaucrats will tell businessmen what goals to pursue.

The mercantilism of Jean-Baptiste Colbert, the finance minister of Louis XIV, is often described as the first industrial policy. When he asked a merchant named Legendre what the king could do to help commerce, the answer was “Laissez faire” or “Laissez-nous faire” (“Let us do”). If Colbert had followed this advice, the country would not have lagged behind Britain, the Netherlands, or Switzerland in the Industrial Revolution. It is not obvious that Rodrik’s new industrial policy is that far from Colbert’s.

Less than a century later, René-Louis de Voyer, Marquis d’Argenson, an Enlightenment figure, showed some exasperation in his memoirs: “Laissez faire, for God’s sake! Laissez faire!” (Laissez faire, morbleu! Laissez faire!)

We know more about economics now than at the time of Colbert or even at the time of Adam Smith. We benefit from eight decades of public choice analysis, the economics of politics. The starting idea of public choice economics is that individuals in the public sector are no less self-interested and no more omniscient than their fellow humans in the private sector. When Rodrik suggests that “cognitive limitations” affect “consumers, communities, and corporations,” he seems to assume that government is free of those.

Rodrik’s third insight of mercantilism, that public policy interventions must be contextual and pragmatic instead of rule-based and principled, is also doubtful. He looks very moderate when he writes that a “certain dose of mercantilism can work in the right context,” and that “it is the judicious combination of the two perspectives,” mercantilism and free markets, that must be pursued. This defense of expediency in public policy ignores an important idea we learned from F.A. Hayek, a 1974 Nobel laureate in economics: Since we cannot know in advance the benefits that individual liberty will bring, pragmatism is a recipe for always ignoring the benefits of freedom in favor of the known (or forecasted) benefits of government bans or injunctions. Expediency leads to ever more dirigiste and authoritarian politics.

Hayek himself was a moderate classical liberal who believed that private actors should remain free to pursue their own separate goals. In this perspective, government imposition of industrial policy goals is difficult to defend—assuming the desirability of a free society.

Current applications / In Rodrik’s view, mercantilist insights illuminate three contemporary areas of concern (not to be confused with his three “insights” discussed above). The first is the growth strategies for underdeveloped countries. Poor countries, he says, need mercantilist industrial policy as shown by the experience of Japan and three of the four “Asian Tigers” nations: Taiwan, South Korea, and Singapore. To different degrees, these countries employed some sort of industrial policies during their rise to the status of advanced economies.

He concedes, however, that the fourth Tiger, Hong Kong, provides a counterexample, but “it was the only such case” (emphasis in original). It’s an annoying exception. If the hypothesis is that industrial policy is a necessary condition for rapid development and you find one case that refutes it, the hypothesis is false.

Hong Kong is an interesting case. In 1960, the first year for which we have data, Hong Kong’s real gross domestic product per capita (in purchasing power parity) was 31 percent of the United States’. In 1996, the year before the British territory was ceded to the Chinese state, the proportion had climbed to 96 percent. In less than two generations, the territory had nearly reached the American standard of living. That standard of living was higher than in Japan or the other three Tigers.

The case of China, which Rodrik often mentions, suggests that industrial policy is not a sufficient condition for rapid development. China’s GDP per capita stood at 5 percent of the United States’ in 1960 and had progressed to only 23 percent by 2019. It is barely higher today. Significant economic growth came only after the death of Mao Zedong, as some economic freedom was allowed. Is that a coincidence? Mao was a fan of industrial policy in the strongest sense of the term. Such policies as the Great Leap Forward and the construction of village and backyard furnaces for smelting steel were meant to change the structure of the economy. The more recent growth of economic (and political) authoritarianism in that country probably explains its current economic malaise—not to mention the rising doubts about the reliability of the Chinese government’s growth statistics.

Coming back to the four Tigers and Japan, Rodrik writes that East Asia economic growth is “the most miraculous experience of development in history.” As for China, it is “the most extraordinary case of economic growth in economic history.” Don’t these hyperboles reflect a short view of history? The most miraculous and extraordinary case of economic development happened two or three centuries ago in England and the Netherlands, rapidly followed by other Western countries. With industrial policy, developing countries are gauchely trying to imitate the “Great Enrichment” (to use Deirde McCloskey’s expression).

Figure 1 shows estimated historical statistics by the Maddison Project for real GDP per capita in a few Western countries and China (or the territories they now occupy) since the mid-18th century. The complete series from which these data are extracted goes back to the year 1 AD. Except for a small increase starting around 1500, there was no growth in the standard of living over more than a millennium and a half. According to these estimates, the world GDP per capita varied between $445 and $615 (in 1990 dollars) from year 1 to 1700; in Europe at the end of that period, it was $998. With the Industrial Revolution, the standard of living in the West started growing exponentially in the 18th and 19th centuries. The chart also shows how Chinese growth only started at the end of the 20th century.

Of course, long historical series for GDP per capita, a 20th century statistical concept, rely on a host of assumptions to construct what are still only ambitious estimates. They do, however, often confirm our less quantified knowledge of history, notably about the Industrial Revolution (Griffin 2013).

Granted that, compared to this original model, the Tigers (and Japan, and now China) have shown higher annual growth rates. But economic take-offs are understandably more rapid now that there are more rich people to trade with, lower transaction costs, and easily available technology.

Globalization backlash / The second area of concern that Rodrik sees illuminated by mercantilism is the backlash against globalization. The so-called “China shock” in the United States (and other advanced countries) from the increase of imports from China between 1999 and 2011 led to “negative externalities” for “society at large”—whoever “society at large” is—and eventually to the rise of the “right-wing populist movement.” I agree that this last development is worrisome, as would be the rise of a left-wing populist movement, but any industrial policy that could have prevented this economic adjustment would have entailed a very high cost in terms of general prosperity.

Consider the job losses attributed to the China shock. Change and creative destruction are a permanent feature of a free and dynamic economy responding to consumer demand. It is estimated that of the 5.8 million manufacturing jobs lost in the United States during that period, between 1 million and (at most) 2.4 million were due to the China shock. During the same period, so many new jobs were created in other industries that the net number of jobs increased by more than 6 million. Moreover, it is estimated that most of the decrease in manufacturing jobs was due to technological progress, automation, and higher productivity; it simply took many fewer workers to manufacture the same amount of goods as in generations past. An industrial policy to stop that would have hit other workers in industries not favored by the government.

Climate transition / Rodrik’s third area of concern is that the “green transition”—the shift to less environmentally harmful production—requires a mercantilist industrial policy. To justify his policy preferences, he invokes the concept of “social optimality.” This old utilitarian concept is fraught with danger because it assumes the possibility of interpersonal utility comparisons or the existence of an arbitrary social welfare function that ranks different levels of “social welfare.” According to Rodrik, we need mercantilist insights to guide the green transition.

Ironically, Donald Trump’s own mercantilist industrial policy has interfered mightily with “the development of renewables and green industries” that Rodrik favors. Granted that Trump’s industrial policy is especially chaotic, but politics is always chaotic (Riker 1982); it’s just a matter of degree. If Americans had not been impeded by their own government—from Barack Obama, to Trump’s first term, to Joe Biden, and now Trump’s second term—in their imports of green products from China (notably solar panels and electric cars), a voluntary “green transition” would have progressed in America. “China’s green industrial policies bear the hallmarks of mercantilist statecraft,” Rodrik writes approvingly. And so do Trump’s. (To be fair to Rodrik, the Juhász et al. paper seems less tempted by protectionism.)

In reality, mercantilism is necessary for any seriously coercive “statecraft.” But a state powerful enough to impose a good industrial policy is also powerful enough to impose a bad one.

Primacy of collective choices / Rodrik writes that “Smithians believe in market solutions, even when they understand markets do not work very well on their own.” There is a reason for this belief: It is easier to justify—economically and morally—than the mercantilists’ belief in political solutions, even when they understand that politics does not work very well. Mercantilists and advocates of industrial policy believe in the primacy of collective and political choices; heirs of Adam Smith and other classical liberals believe in the primacy of individual and private choices.

Of course, no social–economic–political system is perfect, but some are more imperfect and dangerous than others. Some collective choices may be desirable or unavoidable, but there are good arguments for recognizing a general presumption for individual choices—that is, for individual liberty. This presumption lies at the basis of economic theory and provides a more solid underpinning for beneficial cooperation in society.

Another Look at East Asian Growth

The second recent article that bears on the issue of mercantilism and industrial policy, titled “Industrial Policy, Asian Miracle Style,” appeared in the Fall 2025 issue of the Journal of Economic Perspectives. The authors, International Monetary Fund economists Reda Cherif and Fuad Hasanov, observe the rapid growth of the Asian Tigers and, before them, Japan. Something in their industrial policies, the authors conjecture, must have been more effective than the belief of the so-called “Washington Consensus” (in the 1980s and 1990s) that good governance, credit markets, foreign investment, and trade make for strong economic development. “Starting from low levels of per capita income, [these countries] joined the group of affluent industrialized economies” (emphasis in original) within a couple of generations, Cherif and Hasanov note. They try to find the causes of this “miracle.”

Figure 2 shows the growth of real GDP per capita (in 2017 dollars at power purchasing parity) for the four Tigers, plus Japan and China for comparison purposes. (The data, from the Penn World Tables 10.01, are those used by Cherif and Hasanov. The data start at 1950 for Japan and the first year available for the other countries.) The four Tigers have grown at (real) annual rates of 4–6 percent since 1960. Note how Hong Kong’s growth was overtaken by Singapore not long after Hong Kong was brought under the Chinese state. China’s growth started much later and reached similarly high rates.

Cherif and Hasanov argue that the Tigers’ growth required industrial policies to solve “coordination failures, learning externalities, and market imperfections.” Clusters of domestic firms had to be created in “sophisticated industries.” What was needed was a “developmental mindset” and “a host of policies beyond subsidies or protectionism.” “Internationally competitive export-oriented and innovative domestic industries and firms” were developed. Taking issue with the literature on state capacity, they claim that “good institutions” such as the rule of law were not necessary. Indeed, East Asia had its share of authoritarian rulers.

The authors see the Tigers’ industrial policies as characterized by “three key principles”:

• “Proactive and continuous state action to steer resources into the development of sophisticated [domestic] industries.”
• Exports and international competition, without the protectionism of traditional industrial policy.
• Accountability for those who receive state support, based on long-term performance criteria.

Strangely, they claim that these policies are similar to those of the British and American governments in the 19th and early 20th century.

Cherif and Hasanov dismiss protectionism, which “creates complacency and reduces competition.” They argue that export orientation distinguishes the Tigers from other economies—say, Malaysia—with lower economic growth. Because of the Tigers’ participation in world trade, domestic competition is fierce and inefficient firms are eliminated.

Besides protectionism, however, their arguments are largely similar to Rodrik’s. The two IMF economists confirm that their “approach is in stark contrast with a more laissez-faire philosophy.” In their model, the developmental state acts as a venture capitalist instead of a planner. The state creates winners instead of picking them. It relies on “market signals” to set its goals.

This industrial policy with a human face seems too good to be true. The taxpayers and other forced savers still must provide the government’s resources. The interventions certainly falsify and distort domestic price signals by orienting production away from what consumers want, even if the distortions are less obvious than under a more traditional, protectionist, and muscular industrial policy. Perhaps the authors were unduly inspired by the more interventionist models of Taiwan and South Korea?

The ideal state / A basic problem with industrial policy, economic planning, and social engineering is that they presuppose an ideal state. The state is the cheerleader who (of course) carries a big stick in case the carrots don’t work. The politicians articulate national goals, “usually in the form of a long-term ‘vision,’ ” Cherif and Hasanov write. Pardon me? Politicians have a discount rate lower than markets? It is true the authors are not necessarily speaking of democratic governments, but autocrats are not angels either.

Powerful, autonomous, and competent bureaucratic agencies (such as the Ministry of International Trade and Industry in Japan or the Economic Planning Bureau in South Korea) “accumulate a deep knowledge of industries”—“context- and sector-specific knowledge”—and coordinate “firm actions” (again quoting Cherif and Hasanov). As Hayek suggested, it is far from clear how they can obtain a deeper knowledge than what market prices provide. Getting the “feedback from all the stakeholders involved” is impossible, if only because government intervention multiplies “stakeholders” and free riders. The authors do admit that government bureaucrats need “sufficient power to coordinate various stakeholders.” In other words, it is impossible to do economic planning without somebody planning somebody else’s decision in some important way.

The Tiger countries were certainly mixed economic systems with some economic freedom and some government intervention. But relative to other countries in the world, they had much economic freedom. Consider the Economic Freedom of the World (EFW) Index, which scores most countries in the world by measuring certain components of economic freedom grouped in five areas:

• size of government
• legal system and property rights
• sound money
• freedom to trade internationally
• regulation

An index is, by necessity, more indicative than definitive because it depends on its choice of components and the weight assigned to each. That said, the EFW index, which covers the period 1970–2023, ranks Hong Kong and Singapore among the very freest countries and generally ahead of the United States. Taiwan is in the top 10, but it got there only recently. South Korea is lower in the index, which suggests it has more industrial policy even if happily disciplined by world markets. (Of course, economic freedom is a necessary but, at least in the short term, not sufficient condition for individual liberty in general.)

Back to Hong Kong / For proponents of industrial policy as a tool of economic development, we have already seen that Hong Kong is a thorn. At first sight and for a long time, it did not have any industrial policy in the sense of the government trying to change the industrial structure. Granted that in 1960 (the earliest year data are available for Hong Kong) it was less poor than the other Tigers, but that may reflect its more classical liberal British administrators after its Japanese occupation during World War II. The country is devoid of any natural resources except for its harbor.

Some analysts do argue that the colonial administration imposed some industrial policies. Hong Kong’s economy was certainly not a pure laissez-faire regime. Cherif and Hasanov claim that “Hong Kong under British rule deployed an effective industrial policy through a network of councils relying on indirect tools such as support for research and development.” Other, usually mild, government economic interventions have been reported.

Yet it is pretty clear that, since at least the end of World War II until its handover to the Chinese state in 1997 (or the 1980s, when the handover was decided), the Hong Kong economy was much closer to 19th-century laissez-faire than to an industrial policy regime. An important figure in the colonial administration, John Cowperthwaite, arrived there in 1946, became deputy financial secretary in 1951, and financial secretary in 1961, before retiring in 1971. Cowperthwaite was a classical liberal influenced by Adam Smith, and he constantly focused on economic freedom.

Milton Friedman famously celebrated Hong Kong as a natural experiment in laissez-faire. Its residents faced no domestic constraints on imports, little regulation, no central bank (and thus no monetary policy for the Hong Kong dollar), no central development or planning agency, a flat income tax, and low government expenditures. The government supplied infrastructure, public education programs, and assistance to immigrants (mainly Chinese fleeing communism) through the subsidization of housing. It was not exactly anarchy, but many contemporaneous observers and some Hong Kong businessmen actually complained about the prevailing laissez-faire.

Cowperthwaite argued that collecting macroeconomic statistics was only necessary in countries with “high taxation and more or less detailed Government intervention in the economy,” and thus was not needed in Hong Kong. We saw above that, by 1996, the GDP per capita in Hong Kong had nearly reached the US level. The Guardian’s obituary of Cowperthwaite noted that, in 1961, the average Hong Kong resident earned about a quarter of the average British resident’s income, but more than 100 percent by the early 1990s.

The EFW index confirms the high level of economic freedom in Hong Kong. In 1970 it ranked as the most economically free country in the world, while the United States was third, Japan was 14th, and the other Tigers were quite far behind. In 1980, these rankings had not changed much. In 1995 Hong Kong was still first, followed by the United States and Singapore.

Hong Kong even retained the top EFW rank right after the handover to the Chinese state as a “special administrative region” under the so-called “one country, two systems” policy. The main reason for Hong Kong’s post-handover performance in the EFW index is that it retained its free-trade status even though the other major components (except for sound money) decreased under Chinese dirigisme. In 2023, the latest year EFW rankings are available, Hong Kong was still first, Singapore second, the United States fifth, and Taiwan had climbed to seventh place. (New Zealand and Switzerland were third and fourth.) Neither Japan nor Korea were in the top 10, while China was 108th.

A recent econometric analysis concludes that Cowperthwaite’s liberalism “did not actually boost the territory’s economic growth” (Geloso et al. forthcoming). The authors conjecture that the size of its government has been underestimated because of the neglected factor of land ownership and allocation: Virtually all land in the colony was owned by the government and leased to chosen occupants. This could have been the equivalent of an industrial policy that partly cancelled the potential of laissez-faire. In other words, Hong Kong would have grown fast despite government intervention, as the other Tigers did. Note also that Cowperthwaite was at the helm of the colonial administration for only 10 years, although his colleagues were themselves more laissez-faire than the mother country’s government.

In the early 1990s, two University of Hong Kong professors, A. Yeh and M.K. Ng, happily reported that “the Hong Kong government had finally abandoned its laissez-faire industrial policy” (my emphasis). But they still believed that “the Hong Kong government [was] not doing enough and [was] lagging far behind.” They foresaw “a general lack of long-term economic development” compared to Singapore, Taiwan, and South Korea. Figure 2 suggests they were wrong about the past, as Hong Kong had thrived under “laissez-faire.” And its growth has now been stifled by Chinese planning.

Dangers / It seems we cannot find, at least in Hong Kong before the handover, the ambitious view of industrial policy proposed by Cherif and Hazanov. Moreover, their vision of “policies and firm actions … coordinated and adapted to the changing circumstances” requires much faith in political power and the state.

At a more general level, it is tempting for the economist to assume the role of a philosopher-king in advising policymakers. A large part of economic inquiry during the past century or so—from ordinal utility and welfare economics to social choice theory, public choice analysis, and constitutional political economy—has pointed to methodological problems in the philosopher-king approach. (Nobel economics laureate James Buchanan and his collaborators have much to teach us on these matters.) There is a risk that brilliant policy-oriented economists forget that real wealth is predicated on consumer sovereignty (or individual sovereignty) rather than (or at least more than) on the interests or ideas pushed by politicians, bureaucrats, and the political process in general.

There is a danger for technical economists to forget that free markets are generally more efficient than politicians and bureaucrats at coordinating individual actions and fostering prosperity. If, by getting closer to laissez-faire, some countries grew much faster than most other underdeveloped countries, it might be wise to move in that direction instead of toward the millennia-long empowerment of political authorities. As shown by EFW analysis, there is a strong correlation between economic freedom and economic growth (and other indicators of human welfare, including life satisfaction). Is this a coincidence?

Was the World Bank Right?

In the same JEP issue as Cherif and Hazanov, former World Bank senior economist Nancy Birdsall offers a moderately skeptical view of industrial policy. Her article, “The World Bank’s East Asian Miracle: Too Much a Product of Its Time?” defends a 1993 Bank report, The East Asian Miracle: Economic Growth and Public Policy, that was skeptical of industrial policy for developing countries. The report focused on the four Tigers plus Japan and China.

The report attributed those countries’ success to free markets, shared economic growth (as opposed to elite exploitation), and “export-push” policies. Its free-market approach, Birdsall claims, reflected the “neoliberal” Washington Consensus and, to show she’s not similarly afflicted, she jabs at the “free-market religiosity” of the time—as if the re-ascendant state idolatry were not a more dangerous religion.

Birdsall agrees with the report that government support for exports through means such as credit subsidies was “a very effective way of enhancing absorption of international best-practice technology.” Such policies were the only “industrial policy” worth recommending. That was enough to solve externalities and market imperfections (even if, as we are supposed to know, they are as numerous as the sand on the seashore).

Industrial policy for all / The 1993 report defined industrial policy as “government efforts to alter industrial structure to promote productivity-based growth.” This looks like a precursor to the Juhász et al. definition, but “productivity-based growth” is less open-ended than the ethereal “some public goal,” which can be anything and everything. Birdsall also notes that, with the progress of service industries, “industrial” should not be understood as only manufacturing. She suggests that industrial policy can be simply “defined as interventions by the state.” This interpretation of industrial policy would seem to be a reversion to the ethereal conception. So, what exactly is industrial policy?

The more one obscures or extends the meaning of industrial policy, the more likely the conclusion follows that it was the cause of fast economic growth in East Asia. Or the cause of everything good.

For example, in Hong Kong, was investment in public housing an industrial policy because it reduced the cost of labor, as Birdsall seems to suggest? In theory, whether the subsidization of housing reduced or increased the cost of labor depends on the size of the income and substitution effects in individual choices between leisure and the subsidized good. But the point here is that industrial policy can apparently be anything. Could we consider free immigration in Hong Kong an industrial policy? Is any government intervention or abstention an industrial policy? If so, everything is industrial policy and the concept is useless. Birdsall’s reflections, perhaps unwittingly, lead to such questions.

Some good points / The article is useful in other respects. Birdsall defends the World Bank’s Miracle report and its reservations about industrial policy. She emphasizes the benefits of participating in world trade for underdeveloped countries. Like Cherif and Hasanov but contrary to Rodrik, she does not fall for the call of the protectionist sirens.

She also reminds us of the danger that rent-seeking represents in underdeveloped countries with dirigiste governments—dangers that were happily attenuated by institutional and cultural factors in East Asia. The report, she writes with an academic litote, “was not completely off the mark in its caution about industrial policy for other regions of the developing world.”

There is still no consensus on the usefulness of industrial policy for underdeveloped countries, she admits. We must explore “whether and how ‘industrial policy’ [scare quotes in the text], however defined, can advance the cause of development.” What about the Hippocratic injunction “Primum non nocere” instead? Certainly, if one armed with social engineering tools is not certain whether coercion will harm more than help—harm some fellow humans more than help other fellow humans—abstention would seem in order.

The Old Demon

Did we learn anything new about what is industrial policy from these three papers? A critical reading of them suggests three possible answers:

It’s somewhere in-between: Somewhere, in some interval between total laissez-faire and absolute control of all the subjects’ lives by political authority, stands industrial policy. But where? Perhaps some mechanism of collective choice or a philosopher-king decides, or is it just a terminological convention? Industrial policy often seems to simply mean some government intervention.

Apparently, it also means non-intervention. For example, if preventing restrictions on economic freedom by, say, prohibiting minimum wages or trade union privileges is part of industrial policy, as Birdsall suggests, then everything—government intervention and non-intervention—is industrial policy. If political authority does not interfere with my freedom to work for less than the minimum wage it decrees, is that an industrial policy? Remember Yeh and Ng’s “laissez-faire industrial policy.” No wonder scholars are still debating whether there was industrial policy in Hong Kong! At best, the expression is a descriptive label for the results of political accommodations, the exchange of state privileges against political support, and the rent-seeking of special interests.

It yields structural economic change: Another conception of industrial policy—that it is an official intention to change “the structure of economic activity”—is more precise and more analytically useful. Or is it? How much structural change is required for an intervention to become industrial policy? And what counts as economic activity? What is structural change anyway? For example, would a natalist policy or a Mao-style anti-natalist policy deserve the good name of “industrial policy”?

As we saw, there are deeper problems. How do politicians and bureaucrats know what should be the structure of the economy? Is any “public goal” to be pursued? Would it be an acceptable public goal to discriminate against Jews or blacks or the poor, or against the main industries they work in or buy from? What if a ruler who thinks that he embodies “the people” chooses—on the latter’s behalf—the industries to be favored and thus those to be handicapped, or the countries whose goods will be taxed when Americans buy them, or who believes in his gut that the elimination of trade deficits is a priority public goal?

One objection to this line of reasoning is that the authors of the three papers are doing a positive, not normative, analysis of industrial policy, and they just take the “public goal” as given. In this perspective, a public goal is simply what a government wants—perhaps, or perhaps not, supported by some majority of the population. As a positive analyst, the economist would merely recommend the best means to achieve the government’s goals. After all, we often find economists advising the worst governments. This last reflection points to the problem with the objection. Certainly, the economist should reflect on whether his recommendations will further the construction or maintenance of a free society in the general sense of the individuals’ equal freedom of choice. Otherwise, the positive economist will put himself at the service of despots instead of being primarily an adviser to the people (“people” in the plural sense).

It’s just the good government interventions: The third and simplest conception of industrial policy, without scientific varnish, is that it is simply a set of good government interventions. This characterization may remind us of a line that future California governor Jerry Brown used as a high school debater. Whatever the debate topic, he would end his speeches with, “What we need is a flexible plan for an ever-changing world.” Yes, we want that!

All that being said, we can concede that some industrial policies, however defined, are worse than others. Government interventions that interfere directly with trade and prices cause especially serious damage to prosperity. So does the old demon of mercantilism. Industrial policies that pretentiously aim at changing the structure of social cooperation in pursuit of “some public goal” are potentially more detrimental than humble government interventions focused on simple considerations of economic growth based on market prices.

Readings

• Birdsall, Nancy, 2025, “The World Bank’s East Asian Miracle: Too Much a Product of Its Time?” Journal of Economic Perspectives 39(4): 127–148.
• Bolt, Jutta, and Jan Luiten van Zanden, 2020, “Maddison Style Estimates of the Evolution of the World Economy. A New 2020 Update,” Maddison Project Database, v. 2020.
• Cherif, Reda, and Fuad Hasanov, 2025, “Industrial Policy, Asian Miracle Style,” Journal of Economic Perspectives 39(4): 101–126.
• Friedman, Milton, 1998, “The Hong Kong Experiment,” Hoover Digest, July 30.
• Geloso, Vincent, Jamie Bologna Pavlik, and Yang Zhou, forthcoming, “An Almost Laboratory Experiment: John Cowperthwaite and Hong Kong’s Economic Prosperity,” Public Choice.
• Griffin, Emma, 2013, Liberty’s Dawn: A People’s History of the Industrial Revolution, Yale University Press.
• Gwartney, James, et al., 2025, Economic Freedom of the World: 2025 Annual Report, Fraser Institute.
• Juhász, Réka, Nathan Lane, and Dani Rodrik, 2024, “The New Economics of Industrial Policy,” Annual Review of Economics 16: 213–242.
• Lemieux, Pierre, 2023, “Why the Great Enrichment Started in the West,” Regulation 46(2): 8–10.
• Lemieux, Pierre, 2025, “Trumpian Chaos as an Attempt at Central Planning,” Regulation 48(4): 4–5.
• Lemieux, Pierre, 2026, “Today, through Hayek’s Eyes,” Regulation 49(1): 57–61.
• Maddison, Angus, 2003, The World Economy: Historical Statistics, Organisation for Economic Co-operation and Development.
• Riker, William, 1982, Liberalism Against Populism: A Confrontation between the Theory of Democracy and the Theory of Social Choice, W.H. Freeman.
• Rodrik, Dani, 2025, “What the Mercantilists Got Right,” National Bureau of Economic Research Working Paper No. 34353, October.
• Yeh, A., and M.K. Ng, 1994, “The Changing Role of the State in High-Tech Industrial Development: The Experience of Hong Kong,” Environment and Planning C: Government and Policy 12: 449–472.