ILLUSTRATION: ALBERTO MIRANDO
A FUNNY THING happened on the way to the rehabilitation of industrial policy. The World Bank, which spent three decades treating state-directed growth as an intellectual pariah, published an exhaustive report in March arguing that governments should indeed be allowed to pick favourites—provided they do so with the precision of a surgeon, the discipline of a monk and the political self-restraint of a Swiss canton. The report landed just as America, the world’s most influential economy, was demonstrating none of those qualities. The timing is either exquisite or exquisitely awkward.
The report, “Industrial Policy for Development”, has been read as an abject surrender by the high priests of the Washington Consensus. Headlines have been excitable. The Atlantic declared that “A Pillar of the Economics Establishment Admits That It Was Wrong”. The Washington Post called the document “Industrial Policy for Dummies”. Both characterisations miss the real story. The World Bank has not abandoned market principles. It has produced a user’s manual for a tool that every government is already wielding, often with a hammer where a scalpel would do. The tragedy is that the manual’s most important reader appears intent on bludgeoning its own furniture.
To grasp the shift, one must go back to 1993. That year the World Bank published “The East Asian Miracle”, a report forged in the furnace of internal dispute. Its conclusion was a qualified no: industrial policy had not caused the success of the tiger economies, and developing countries with weak institutions, poor education and high deficits should steer clear. Nancy Birdsall, the World Bank’s acting chief economist at the time, now acknowledges the report was “a product of its time”. In early-1990s Washington, free markets were taken for granted. The document was, as one insider put it, an “essay in persuasion” rather than a definitive verdict.
But the report landed with the force of a papal bull. It turned the concept of industrial policy into the economic equivalent of bloodletting, solidifying an orthodoxy around free trade, deregulation and privatisation. The World Bank conditioned loans on recipients eschewing industrial policy. For two decades the taboo held—in theory. In practice the world got on with it.
China alone now operates more than 2,500 special economic zones. Turkey has around 500; India nearly 400 each. A review of 183 national development strategies for the 2026 report found every single one targeted at least one industry. Developing countries resorted to industrial policy more frequently than advanced ones. The real question was never whether governments were doing it: it was whether they were doing it competently.
This is where the World Bank’s new position departs from both the 1993 orthodoxy and the caricature now circulating in Washington. Indermit Gill, the World Bank’s chief economist, writes that the old report’s stance “has the practical value of a floppy disk today”. But he is not peddling state-led magic. The analysis distinguishes between 15 types of intervention—from industrial parks and training programmes to import tariffs, production subsidies and competitive devaluation—and argues that their suitability depends on three factors: the size of a country’s domestic market; the technocratic capacity of its state; and the fiscal space available for experimentation and error. Not every tool suits every hand.
The report is in effect a manual for doing less rather than more. It warns against tariff-heavy approaches that invite retaliation and hurt industries reliant on imported inputs. It argues for handing implementation to agencies insulated from political pressure: a recommendation that borders on utopian. It stresses the importance of cross-party commitment so that policies survive changes of government. Above all it insists that industrial policy cannot substitute for getting the fundamentals right: a healthy, educated workforce, decent infrastructure, stable macroeconomics and a business climate that encourages private investment. Governments that use industrial policy to buy time should be using that time to fix the basics.
Where the World Bank advocates stable, predictable, multi-year frameworks, America has delivered a policy regime that changes by the tweet. Tariffs have been announced, paused, revised, struck down by courts and reinstated under different legal authorities. The result has been an endless procession of lobbyists, foreign leaders and chief executives making pilgrimages to the White House to secure carve-outs. Where the World Bank warns against picking individual corporate winners, the Trump administration has taken equity stakes in firms like Intel. Where the World Bank urges humility before the difficulty of knowing which industries to target, the White House has acted with unbounded certainty.
Mary Lovely of the Peterson Institute for International Economics puts the dilemma facing developing countries succinctly. “Lots of these middle-income countries have started to see their growth slow down a lot. And so what do you do? Well, if you look at China, what they did was industrial policy—and that seemed to work pretty well for them.” The World Bank’s answer is that China’s success owed more to market discipline than central planning. By 1993 90% of prices related to industrial output were determined by market forces. China worked hard to improve the general business climate, and its best results came from using competition, not diktat, to shift resources.
As for South Korea, the cumulative cost of its heavy-industry push in the 1970s and 1980s ran to about 2.4% of GDP over 1973-79 alone. The benefits, however, exceeded the costs: the country’s GDP today is roughly 3% higher than it would have been without the subsidies. But this industrial policy would almost certainly have failed had South Korea not simultaneously invested in education, forced firms to compete and pushed businesses to adopt foreign technologies. The intervention was necessary; it was nowhere near sufficient.
The World Bank’s re-evaluation does contain a real reversal. When researchers revisited East Asia’s experience with decades more data, they found that industrial-policy efforts had been more crucial than the 1993 report allowed. The earlier study had examined only the high-performing tigers without looking at lower-performing counterfactuals, and it had defined success so narrowly as to be almost unfalsifiable. The new work is more empirically honest.
Yet the report’s most disgusting finding is not about economics. It is about politics. The prerequisites for successful industrial policy—technocratic competence, insulation from lobbying, cross-party consensus, the willingness to kill failing programmes—are exactly the things that political systems in the grip of populist reaction find hardest to supply. The World Bank has produced a manual for a machine that most governments lack the institutional wiring to operate. To its credit, the report says so explicitly. Industrial policy belongs in the toolkit of every country, Mr Gill argues, but “the manual does not tell them to do more industrial policy; it tells them to do less while dispensing advice on how to do it better.”
The wry observer might note that this is not a U-turn so much as a highly conditional right turn, executed with the handbrake on. The world has changed, as Mr Gill says. Average GDP per head in developing countries has doubled since 1993. Schooling levels have risen. Inflation has fallen and the quality of economic management has, with few exceptions, improved. Many more countries now meet the prerequisites for industrial policy than did three decades ago. But meeting the prerequisites and using them wisely are different things. The market for policy manuals is robust. The market for political restraint, as Washington demonstrates daily, remains stubbornly thin. ■
