Industrial policy — government action to shape the sectoral structure of the economy — was for decades a dirty word, discredited by the failures of import substitution (the trade-and-development module) and rejected by the free-market Washington Consensus. It has now returned to the centre of economic policy worldwide, from the US green-energy push to Africa's industrialisation strategies. This module covers the revival and the market failures that justify it.
What industrial policy is, and its discredit
Industrial policy is any government policy that deliberately promotes particular industries or activities over others — through subsidies, protection, directed credit, public investment, special zones, and more — to shape what the economy produces. It fell into disrepute for two reasons: the failures of import-substitution industrialisation (inefficient protected industries that never became competitive — the trade course), and the rise of the Washington Consensus, which held that governments cannot 'pick winners', that intervention breeds rent-seeking and capture (the Political Economy course), and that markets should determine the economy's structure. For a generation, 'industrial policy' was something serious economists were supposed to oppose.
The revival
Why industrial policy is back
Industrial policy has returned because the market-failure case for it was always sound, and because the East Asian success (the developmental state, next modules) showed it can work when done well. The revival rests on identifying the specific MARKET FAILURES that justify intervention — because where markets fail, there is in principle a role for policy to correct them: • Coordination failures — many industries require complementary investments that no single firm will make alone (a steel plant needs power, ports, skilled workers, and downstream users, none of which is worth building without the others). The market may fail to coordinate the 'big push' of simultaneous investments (Rosenstein-Rodan), so the state may need to coordinate it. • Information/discovery externalities — discovering what a country can profitably produce is costly, and the discoverer's knowledge spills over to imitators (below). • Learning externalities — productivity improves with experience (learning-by-doing), but firms underinvest in learning whose benefits spill over to others or to the future. • Capital-market failures — credit constraints prevent good projects from being financed (the directed-credit rationale). • Agglomeration externalities — firms benefit from clustering, but no single firm internalises the benefit of being the first to locate somewhere. Industrial policy is justified where it corrects a genuine market failure that prevents an industry forming on its own — the same logic as any public intervention (the Public Finance course's market-failure framework, applied to the economy's structure).
Self-discovery and the information externality
Hausmann-Rodrik: economic development as self-discovery
Ricardo Hausmann and Dani Rodrik's influential argument (2003): a developing country doesn't actually know what it can produce competitively — discovering it requires costly experimentation (an entrepreneur invests to find out whether the country can profitably grow cut flowers or assemble electronics). But the knowledge generated by that discovery is a PUBLIC GOOD: if the first entrepreneur succeeds, imitators rush in and compete away the profits, so the discoverer bears the cost but doesn't capture the full benefit; if the first entrepreneur fails, they bear the loss alone. This information externality means there is too LITTLE self-discovery — entrepreneurs underinvest in finding out what the country can produce, because they can't capture the returns. So there is a market-failure case for industrial policy to SUBSIDISE the discovery process (support entrepreneurial experimentation in new activities) — not because the government knows what will succeed (it doesn't), but because the market underprovides the experimentation that finds out. This reframes industrial policy away from 'picking winners' (which governments can't do) toward supporting a process of discovery and self-correction — the central modern justification, developed in the picking-winners module.
The entrepreneurial state
Mariana Mazzucato's 'entrepreneurial state' (2013) added a provocative argument: the state has historically been not just a fixer of market failures but an active, risk-taking INNOVATOR — funding the basic research and early-stage technologies that the private sector was too risk-averse to fund, and that later became transformative. Her famous example: the technologies in the iPhone (the internet, GPS, touchscreens, Siri's voice recognition) were substantially funded by public research. Mazzucato argues the state should embrace this entrepreneurial role (mission-oriented innovation policy — directing innovation toward grand challenges like decarbonisation) and should share in the rewards (not just bear the risks while the private sector captures the gains). The argument is influential and contested (critics say it overstates the state's role and understates its failures), but it has helped legitimise an active, directional state role — and it connects to the current 'green industrial policy' moment (the US Inflation Reduction Act, the EU's response), in which climate has made industrial policy mainstream again across the rich world (the Climate Finance course). Industrial policy is no longer heterodox; it is the new mainstream.
Exercise
An entrepreneur in an African country is considering investing to find out whether the country can competitively grow and export a new high-value crop (say, avocados for the European market). No one has tried it before. (1) Explain the information externality that may lead to too little of this 'self-discovery'. (2) Explain the market-failure case for the government supporting the experiment, and how it differs from 'picking winners'. (3) Identify a coordination failure that might also block the new industry. (4) Design an industrial-policy approach to the new-crop opportunity that addresses these failures without simply subsidising a guess.