Do Markets Always Know Best?
The free-market orthodoxy holds that competitive markets allocate resources more efficiently than governments ever could. Yet history is littered with cases where state intervention — dismissed by economists at the time — produced transformative economic outcomes. This is the central puzzle of industrial policy.
"The question is not whether governments should intervene in the economy, but how they should do so and to what end. All governments, without exception, intervene in their economies."
— Ha-Joon Chang, 23 Things They Don't Tell You About Capitalism (2010)Understanding what governments are challenging when they intervene
Comparative Advantage
David Ricardo showed that countries benefit from specialising in goods they produce relatively more efficiently, even if another country can produce everything more cheaply. Trade then allows consumption beyond domestic production possibilities.
The implication: a labour-abundant, capital-scarce country like 1960s Korea should specialise in labour-intensive goods — textiles, footwear — not capital-intensive steel or shipbuilding.
Static vs. factor endowments and productive capabilities change over time, allowing countries to shift into industries they could not initially compete in."">Dynamic Comparative Advantage
Ricardo's theory describes comparative advantage at a single point in time. But comparative advantage can be created. Countries do not have fixed endowments — capital accumulates, skills develop, technologies diffuse.
Chang's core argument: if Korea had followed comparative advantage in 1968, it would still be producing cheap textiles. Instead, it built steel. And steel built everything else.
Why Markets Underprovide Strategic Industries
Information asymmetry: Private investors cannot fully evaluate the long-run spillovers from an industry entering a new sector. The social return exceeds the private return.
Coordination failure: Steel needs ports; ports need ships; ships need steel. Each investment is only profitable if others happen simultaneously — the market cannot coordinate this.
Capital market failures: Banks won't fund 20-year industrial projects without proven returns. The market horizon is too short.
Hamilton, List & Protection
Alexander Hamilton (1791) argued that new industries need temporary protection to reach the scale and learning required to compete internationally. Friedrich List formalised this: developing nations should prioritise productive power over short-run gains from free trade.
The irony noted by Chang: the United States and Germany industrialised behind protectionist walls, then pulled them down and told everyone else to trade freely.
The intellectual foundations of industrial policy
Friedrich List
German economist who argued for protecting infant industries. Influenced German and American industrialisation strategies.
Alexander Hamilton
US Treasury Secretary who made the first systematic case for industrial policy in his Report on Manufactures (1791).
Dani Rodrik
Harvard economist who formalised industrial policy theory. Argues governments should discover which activities are profitable to promote.
Ha-Joon Chang
Cambridge economist. Argues rich countries used active industrial policy to develop, then removed the "ladder" after climbing it themselves.
The World Bank president's critique of developing-country ambition — 1949–63
The Highway
Grand infrastructure projects that consumed vast capital with limited productive spillovers. Prestigious to build, but often bypassed the rural poor who most needed connectivity.
The Integrated Steel Mill
Capital-intensive heavy industry that defied comparative advantage. Black argued these were economically irrational vanity projects that would drain government budgets — POSCO proved him wrong.
The Monument
Symbols of national prestige — airports, government buildings, stadiums. Expressions of political aspiration with poor economic returns. The term "white elephant" entered economic discourse here.
How states intervene in the 21st century
Mission-Oriented R&D
The US Defence Advanced Research Projects Agency funded early-stage research that produced the internet, GPS, and touchscreens. Private markets would never have funded these — the timescales and uncertainties were too great. Mariana Mazzucato's "entrepreneurial state" thesis argues the state takes the risks; private firms reap the profits.
Strategic Sector Targeting
China's explicit policy to dominate ten strategic sectors including EVs, semiconductors, and AI by 2025. State banks provided directed credit; SOEs absorbed losses in early years. The EV sector in particular achieved dramatic cost reductions — BYD had, by some measures, overtaken Tesla in global EV sales by 2023 — though comparisons depend on methodology (BloombergNEF, 2024).
The New Consensus?
The Inflation Reduction Act (2022) was projected to mobilise up to $369bn in climate and clean energy investment over ten years (Congressional Budget Office estimate, 2022) — though actual spend depends on take-up of tax credits. The EU's Green Deal Industrial Plan followed. The IMF's fiscal monitor (2023) acknowledged that targeted industrial policy may be warranted to address climate externalities — a significant shift from the institution's historically sceptical position. Some economists argue the debate has shifted from whether to how — though others maintain the evidence for effective industrial policy remains contested (see Criscuolo et al., 2022).
The Government Failure Counter-argument
Friedrich Hayek argued governments lack the dispersed knowledge that price signals aggregate. Public Choice theory (Buchanan) warns that governments respond to political pressures, not economic signals — creating rent-seeking, cronyism, and industries that survive on subsidy rather than innovation. The question is not whether governments can pick winners, but whether they do.
📚 Sources, Evidence & Academic Notes
▼ EXPANDMinister of Economic Development
You are the chief economic minister of an unnamed country. Across three critical decades, you must allocate a real budget across competing sectors. Advisors will brief you — but they disagree. The country's identity is hidden until the final reveal.
Each scenario is based on a real historical nation. How close will your instincts get to what actually happened — and what the outcomes were?
POSCO — Pohang Iron and Steel Company
Click each event to expand the detail
🎯 Factors of Success
- Strong state with clear, long-term industrial vision and political will
- Meritocratic but disciplined management — Park Tae Joon's military approach
- Technology transfer agreement with Japanese companies aligned incentives
- Export orientation forced POSCO to be internationally competitive from the start
- Downstream linkages: steel demand from shipbuilding and construction was built simultaneously
- Government used performance targets — support was conditional on results
⚠️ Risks That Were Managed
- Lack of raw materials solved through long-term import contracts — raw materials became cheaper at scale
- Political risk managed by keeping POSCO outside factional politics
- Capital scarcity addressed through Japanese reparations and later World Bank lending (which came once success was visible)
- Workforce skills gap addressed through intensive training programmes in Japan
- Overcapacity risk managed through careful phased expansion tied to demand forecasts
📖 The Lesson for Economists
POSCO is the canonical case against the static comparative advantage argument. Korea did not have a comparative advantage in steel — it created one. This required a state willing to absorb short-run losses for long-run structural transformation, and institutions capable of insulating investment decisions from short-run political pressures.
Crucially, the government did not simply throw money at a sector. Support was conditional, time-limited, and linked to export performance. This is Dani Rodrik's "embedded autonomy" — the state close enough to industry to direct it, but insulated enough to discipline it.
British Leyland — The Nationalised Car Industry
Click each event to expand the detail
❌ Factors of Failure
- Intervention was reactive (saving a dying industry) not strategic (building a new one)
- Political logic drove decisions — union votes, regional employment — not economic performance
- No conditionality: subsidies flowed regardless of productivity or quality improvement
- Management and unions both captured by rent-seeking — neither had incentive to compete
- Fragmented brand portfolio — multiple competing models from the same plants wasted scale
- Government changed strategy repeatedly across different administrations
🔄 Contrast with POSCO
- POSCO was building something new; BL was propping up something failing
- POSCO had export-oriented discipline; BL focused on domestic protection
- POSCO had performance targets; BL had unconditional support
- POSCO had insulated management; BL was politicised at every level
- POSCO built downstream industries; BL created no positive spillovers
- POSCO was privatised when competitive; BL was never competitive enough to privatise
📖 The Lesson for Economists
British Leyland illustrates the danger of conflating industrial policy with politically motivated bailouts. The key distinction economists draw is between proactive industrial policy (building new capabilities, correcting market failures, developing infant industries) and reactive industrial policy (rescuing incumbent industries to avoid short-run political pain).
The public choice critique of government failure is most powerful here: once subsidies begin, organised interests (unions, managers, regional politicians) fight to maintain them regardless of economic merit. The "soft budget constraint" removes all incentive for efficiency. Chang himself acknowledges this — the argument is not that all government intervention succeeds, but that the possibility of failure does not negate the case for trying. The real question is the institutional design of intervention.
🃏 The Card Battle
Choose your side, then play argument cards one at a time. Each card you play is countered — the persuasion meter shifts with every exchange. Play all 5 cards to reach a final verdict on how well your case held up.
You will play 5 argument cards. Each will be countered. The meter judges the overall strength of your case.