At a recent gathering of academics and policymakers at Harvard, a visitor reminded Dani Rodrik of a column he wrote 15 years ago titled “The Manufacturing Imperative.” For the policymaker, an official from Africa, the piece was a North Star—a definitive argument that industrialization was the only reliable engine for driving economic growth, creating stable jobs, and building a robust middle class in developing nations.
But the world has changed since that column was published, and so has Rodrik. The Harvard professor, long one of the most influential voices on global trade and economic development, has moved from being a champion of the factory floor to what he now calls a “manufacturing skeptic.” This proves not that manufacturing is obsolete, but rather that the “economic escalator” it once provided is slowing down, and the path to prosperity is diversifying.
For decades, the playbook for emerging economies was simple: move workers from low-productivity agriculture into high-productivity manufacturing. Factories could absorb vast numbers of low-skilled workers without requiring the sophisticated governance or high-tech infrastructure that modern service economies demand. This transition created the middle classes of the East Asian tigers and, more recently, China.
However, Rodrik suggests that this traditional route is no longer the only—or even the most viable—option for many low-income countries. The rise of automation, the shifting nature of global supply chains, and the digitalization of the economy have fundamentally altered the math of industrialization.
The breakdown of the industrial escalator
The primary appeal of manufacturing was its ability to scale. A garment factory or an electronics assembly plant could employ thousands of people with minimal training, providing a steady wage that lifted families out of poverty. This “escalator” effect allowed countries to build technical capacity and institutional strength incrementally.
Today, that escalator is stuttering. The “low-skill” advantage that once attracted investment to developing nations is being eroded by robotics and artificial intelligence. When a robot in a high-income country can assemble a circuit board more cheaply than a human in a low-income country, the incentive to outsource manufacturing vanishes. This phenomenon, often described as “premature deindustrialization,” means many countries are seeing their manufacturing share of GDP peak and decline before they have ever reached high-income status.
the infrastructure requirements for modern manufacturing have grown. The simple warehouses of the past have been replaced by “just-in-time” logistics chains that require world-class ports, reliable electricity, and seamless digital integration—assets that many developing nations still struggle to implement.
The rise of tradable services
If manufacturing is no longer the guaranteed path, where does the growth come from? Rodrik points toward the services sector, but not the kind of services that typically keep a country poor—such as subsistence street vending or low-end domestic work.
The key is “tradable services.” These are high-value activities that can be exported globally via digital platforms. This includes everything from software development and architectural design to business process outsourcing (BPO) and specialized consultancy. Unlike traditional services, which are consumed locally, tradable services allow a worker in Nairobi or Manila to sell their skills to a client in New York or London.
The potential for productivity growth in these sectors is significant. When a service is digitized and scaled, it can mirror the productivity gains once reserved for the factory. The challenge, however, is that the “barrier to entry” is higher. While a factory worker might need a few weeks of training, a tradable service worker needs a level of education, English proficiency, and digital literacy that requires a more sophisticated national education system.
| Feature | Manufacturing-Led Model | Services-Led Model |
|---|---|---|
| Primary Labor Force | Low-skilled, mass employment | Medium-to-high skilled, specialized |
| Key Requirement | Physical infrastructure (Ports, Power) | Digital infrastructure (Broadband, Education) |
| Growth Driver | Economies of scale in production | Productivity gains via digitalization |
| Risk Factor | Automation and robotics | Digital divide and skill gaps |
A new blueprint for policy
This shift in perspective necessitates a shift in policy. For years, governments in the Global South have obsessed over “attracting FDI” (Foreign Direct Investment) in the form of massive industrial parks and tax breaks for multinational manufacturers. Rodrik’s skepticism suggests that these resources might be better spent elsewhere.
Instead of chasing factories that may be automated out of existence within a decade, policymakers may find more success by investing in “human capital.” This means prioritizing vocational training in digital skills, improving higher education, and ensuring that the electricity grid can support a million home offices rather than ten massive plants.
The transition is not without risk. The services sector can be more fragmented than manufacturing, and it often lacks the strong labor unions that historically helped industrial workers secure a larger share of the profits. There is a danger that a services-led economy could create a “bifurcated” society: a slight, wealthy class of digital professionals and a large, precarious class of gig workers.
Despite these risks, the alternative—clinging to an industrial model that no longer fits the global reality—is far more dangerous. The goal is no longer simply to “industrialize,” but to “modernize” in a way that leverages the strengths of the 21st-century economy.
Disclaimer: This article is intended for informational purposes and does not constitute financial, investment, or legal advice.
The global economic community continues to monitor these shifts, with the World Bank and IMF increasingly focusing on “digital transformation” as a core pillar of development. The next major checkpoint for this discourse will be the upcoming World Economic Forum meetings and subsequent UN development reports, where the efficacy of services-led growth models in Sub-Saharan Africa and Southeast Asia will be further scrutinized.
Do you think the “factory model” is truly dead for developing nations, or is it still the most reliable path to the middle class? Share your thoughts in the comments below.
