Picture the room. A young economist in Dhaka or Addis Ababa, a projector humming, slides that took three weeks to perfect. The logic is airtight. The historical precedent is, by any reasonable measure, overwhelming: build up your manufacturing base, keep wages competitive, orient production toward export markets, accumulate foreign exchange, reinvest in infrastructure and education. It worked for Japan. It worked spectacularly for South Korea. It worked for Taiwan, for Malaysia, for coastal China on a scale the world had never seen before and may never see again. The slides practically write themselves.
The problem is that the strategy those slides describe is increasingly a period piece, not a playbook.
The original machine and why it hummed
Export-led growth, in its classic postwar form, exploited a specific arbitrage. Rich countries had high wages and high consumer demand. Poor countries had low wages and underemployed rural populations. The gap between those two facts was enormous, and manufacturing slotted neatly into it. A country like South Korea in the early postwar decades could produce textiles, steel, and eventually electronics at a cost that American or European producers could not match, selling into consumer markets that were growing fast and protected by stable institutions.
The mechanism was self-reinforcing. Export revenues funded imported capital equipment. Workers who moved from subsistence farming to factory floors gained skills. Firms learned by doing. Wages rose, but productivity rose faster, at least for long enough to accumulate the reserves and institutions needed for a subsequent stage. South Korea's export-to-GDP ratio climbed from roughly 3 percent in its early industrialisation phase to above 30 percent within a generation. That is not a marginal shift. It is a structural transformation of an entire economy, accomplished before the children of those first factory workers had finished school.
Taiwan ran a nearly identical script with distinct variations. So did Singapore, which compressed decades of industrial development into roughly twenty years by making itself indispensable to global supply chains for electronics. These are genuine miracles of policy and circumstance, and any serious account of them has to credit both. The mythology, though, is where the trouble starts.
What changed in the room
Three structural shifts have rewritten the conditions under which that arbitrage operates, and they compound each other in ways that no individual policy can fully offset.
Automation has compressed the labour cost advantage. A garment factory in Bangladesh still employs hundreds of workers per floor, yes. But in footwear, electronics assembly, and increasingly in basic textiles, robotics and computer-controlled machinery have reduced the number of unskilled labour hours in each unit of output. When a Nike supplier in Vietnam can be undercut not by a cheaper human workforce in Ethiopia but by a partially automated facility in South Carolina, the wage differential that once made export manufacturing so attractive narrows dramatically. The International Labour Organization has tracked this compression across multiple sectors; the precise figures vary by industry, but the direction is consistent and the trend shows no sign of reversing.
Global value chains have matured and concentrated. In an earlier era, a country could enter manufacturing by making a complete, simple product, a shirt, a bicycle, a transistor radio, and selling it whole. Supply chains were shorter, more national. Today, a smartphone contains components sourced from forty-odd countries, assembled under tight tolerances, governed by intellectual property arrangements and platform standards controlled by firms in the United States, South Korea, and Japan. Breaking into that system requires not cheap labour but certified quality, reliable logistics, intellectual property compliance, and often a pre-existing relationship with a Tier 1 supplier. Ethiopia's industrial parks have attracted some garment production, but moving up into electronics assembly requires a different kind of institutional infrastructure that takes decades to build, decades that earlier entrants, arriving when the system was still being assembled, did not have to spend.
And then there is demand. The wealthy consumer markets that absorbed East Asian exports across several postwar decades were expanding rapidly, trade regimes were liberalising, and major importers actively wanted to diversify their supply bases for geopolitical reasons. Today, import-absorbing capacity in the United States and Europe is not growing at anything approaching that pace. Protectionist sentiment, reshoring initiatives, and the political economy of deindustrialised regions have made rich-country governments less willing to absorb large new waves of manufactured imports from low-income competitors. A country trying to replicate South Korea's trajectory now faces buyers who are simultaneously more demanding about quality and more politically constrained in how much they can import.
Two friends, two outcomes
Consider a simplified but instructive comparison. Call them Country A and Country B. Both are lower-middle-income nations that pursued export-oriented garment manufacturing aggressively over a fifteen-year period, using special economic zones, preferential trade agreements, and wage restraint to attract foreign investment.
Country A started earlier, when global value chains were still relatively open and automation in garment manufacturing was limited. It accumulated foreign exchange, invested in technical education, and used the breathing room to develop a domestic supplier base for fabrics and fasteners. By the time automation pressure intensified, it had moved up the value chain enough to absorb the shock.
Country B started later, facing a more automated global industry, a more consolidated buyer landscape, and trade agreements already negotiated by its predecessors. It attracted investment, employment rose, and export revenues climbed. But the productivity spillovers were smaller, wages stayed lower for longer, and when a major buyer decided to near-shore production to reduce shipping costs, Country B lost 18 percent of its garment export revenue in two years without a domestic industrial base sophisticated enough to pivot. The strategy was identical. The outcome diverged sharply.
That gap is not a failure of execution. It is a failure of timing compounded by structural change, and no amount of better governance would have closed it entirely.
What people keep getting wrong
The persistent mistake in development economics circles is treating export-led growth as a technology, a set of policy levers that produce predictable outputs when correctly assembled. It was never that. It was an opportunity that existed within a specific configuration of global factor prices, trade politics, and technological capabilities. South Korea and Taiwan did not simply apply a formula. They exploited a window.
The window has not closed entirely, but it has narrowed. Countries that assume they can reproduce the East Asian trajectory by building SEZs, suppressing labour costs, and signing preferential trade deals are, in a sense, building a factory to manufacture a product for which peak demand has already passed. Think of it as arriving at a gold rush not a decade late but a generation late, when the easy claims are staked, the assay offices are crowded, and the merchants selling shovels have already made their fortunes. Some production will still move to low-wage locations, particularly in sectors where automation remains genuinely difficult. But the volume of employment and the pace of structural transformation that low-income countries can expect from that strategy is considerably smaller than East Asian precedent suggests.
There is also a currency problem that deserves more attention than it typically receives. Successful export-led growth generates large current account surpluses, which put upward pressure on exchange rates. Managing that pressure requires either capital controls, sterilised intervention, or tolerance of asset price inflation. South Korea and Taiwan had the institutional capacity and political will to manage it. Many current aspirants do not, and the IMF's own research suggests that premature currency appreciation has truncated industrialisation in several sub-Saharan African cases before manufacturing wages could rise far enough to generate meaningful middle-class consumption. That is a structural trap, not a policy error.
The paths that remain
None of this means low-income countries are without options. The options are simply different and, in most cases, more difficult.
Domestic demand-led growth, building consumer purchasing power through land reform, public investment in health and education, and small-enterprise development, is less glamorous and harder to measure with export statistics, but it builds resilience that export dependence does not. Regional trade, selling manufactured goods to neighbouring economies in preference to trying to penetrate saturated OECD markets, is underexploited and politically undervalued. Services exports, from software development to medical tourism, represent a genuinely new channel that did not exist for South Korea in its industrialisation years, though they require human capital investments that take a generation to pay off. How many development strategies seriously account for that lag before announcing a target?
Some countries will still find export manufacturing viable, particularly in sectors where labour intensity remains high and automation is genuinely limited, handcraft production, certain food processing, some construction materials. The employment numbers will be smaller and the transformation slower, but the path is not entirely blocked.
The real lesson from the miracle economies
What South Korea and Taiwan actually demonstrate, stripped of the mythology, is not that export-led growth is a reliable development algorithm. It is that development requires matching strategy to the specific conditions of a given moment. They succeeded partly by reading their era correctly and partly owing to circumstances that were unusually generous: cheap energy, expanding Western consumer markets, Cold War geopolitics that made rich countries willing to absorb imports from strategically important allies, a global trading system still under construction and not yet dominated by incumbents.
The young economist in Dhaka or Addis Ababa is not wrong to study those miracles. The study should produce humility about replication, not confidence in it. The past is not a template. It is a set of conditions that no longer fully exist, pointing toward a question that each generation of developing nations has to answer for itself: given the world as it actually is right now, what does your country make that anyone actually needs?
The slide deck, however clean, has never contained a satisfactory answer to that.