For most of human history, nearly everyone farmed. Today, in wealthy countries, agricultural workers make up just 1-3% of the labor force. This massive shift—workers moving from farms to factories to offices—is the story of how countries become rich.

Economists call this process structural transformation. It sounds technical, but the core idea is simple: when workers move from low-productivity activities to high-productivity ones, the entire economy grows without anyone necessarily working harder or smarter at their current job.

Yet something puzzling is happening in today's developing economies. Many countries are seeing their manufacturing sectors shrink before they ever really grew. Workers are moving directly from farms to informal services—street vending, domestic work, small-scale retail. This matters because the path a country takes through structural transformation shapes whether it achieves sustained prosperity or gets stuck in a middle-income trap.

Consider a simple thought experiment. Imagine a country where half the workers farm, producing enough food to feed everyone, while the other half work in manufacturing, producing goods for export. Now imagine technology improves so that fewer farmers can feed the same population.

Where do the freed-up agricultural workers go? If they move into manufacturing—where productivity is typically much higher—the country's total output rises substantially. The workers aren't more skilled. The factories aren't more efficient. Growth happens purely through reallocation.

This is precisely what happened in East Asia's growth miracles. In South Korea, agricultural employment fell from 63% in 1963 to under 5% today. China has moved hundreds of millions from farms to factories since 1980. Studies suggest that labor reallocation between sectors accounted for 20-50% of these countries' growth during their transformation periods.

The mechanism works because productivity gaps between sectors are enormous in poor countries. A farmer in a developing economy might produce $1,000 worth of output per year. That same worker in a modern factory might produce $10,000. This gap—which barely exists in rich countries—is the engine of catch-up growth.

Takeaway

Growth isn't just about doing things better—it's often about moving people to where their labor is worth more. The biggest gains come from reallocation, not optimization.

Here's the troubling pattern: many developing countries are deindustrializing before they ever really industrialized. Manufacturing's share of employment is peaking at lower levels and earlier stages of development than it did for previous generations of industrializers.

Consider the contrast. Manufacturing employment in Britain peaked at around 45% in the 1960s. In the United States, it peaked at about 28% in 1953. But in India, manufacturing employment has never exceeded 13%, and it's already declining. Brazil peaked at 16% in the 1980s and has fallen since.

Economist Dani Rodrik has documented this pattern across countries, calling it premature deindustrialization. The causes are debated—automation making manufacturing less labor-intensive, competition from Chinese exports, declining relative prices of manufactured goods—but the consequences are concerning.

Manufacturing has historically been special for development. Factories absorb unskilled labor from agriculture, offer clear productivity gains, produce tradeable goods that earn foreign exchange, and create agglomeration effects that spread technology. When countries skip this phase, they may be skipping the escalator that lifted previous generations into prosperity.

Takeaway

Countries are deindustrializing at income levels where they should still be industrializing. The manufacturing escalator that lifted East Asia may be harder to board today.

If manufacturing is less available as a development path, can services-led growth substitute? This is perhaps the central question facing development strategy today.

The optimistic case points to India. Its IT and business services sector has created well-paying jobs, earned export revenue, and demonstrated that service exports can drive growth. Some economists argue that modern services—software, finance, telecommunications—share manufacturing's desirable properties: they're tradeable, benefit from scale, and can absorb educated workers.

The pessimistic case notes that high-productivity services typically require educated workers, meaning they can't absorb the masses leaving agriculture. A call center needs English speakers; a garment factory doesn't. India's IT sector employs about 5 million people in a country of 1.4 billion. Manufacturing in China has employed over 100 million.

The honest answer is that we don't know if services can power broad-based transformation. Traditional services—retail, restaurants, domestic work—absorb labor but offer limited productivity gains. Modern services offer productivity but limited absorption. Countries may need to find new hybrid paths, combining selective manufacturing with services where comparative advantage exists.

Takeaway

There may not be a single development path anymore. Countries face harder choices about which sectors to prioritize, with less certainty about what will work.

Structural transformation remains the fundamental process by which poor countries become rich. But the rules of this transformation may be changing.

The manufacturing-led path that worked for East Asia is narrower today. Competition is fiercer, automation reduces labor absorption, and global supply chains have concentrated production geographically. Countries entering industrialization now face headwinds that earlier industrializers didn't.

This doesn't mean development is impossible—but it likely means it's harder, and that the strategies that worked before won't simply transfer. Understanding why structural transformation matters is the first step toward figuring out what replaces it when the traditional path closes.