Consider electric vehicle batteries. Traditional industrial policy encourages countries like the Democratic Republic of Congo to process cobalt domestically, rather than exporting the raw mineral. But domestic production alone no longer guarantees that countries will capture the most profitable parts of the value chain.
By Carlos LOPES
Economist Dani Rodrik has recently reopened one of the most enduring debates in development economics, arguing that developing countries need to shift their focus from manufacturing to productivity-enhancing services. In doing so, he has challenged the long-held assumption that manufacturing remains the primary engine of development. But the deeper question is not whether services should replace manufacturing. It is whether structural transformation should still be understood primarily as the movement of resources between sectors. This framework, which guided development economics for more than half a century, is increasingly out of step with how modern economies actually create value.
From W. Arthur Lewis and Raúl Prebisch to Albert Hirschman and Nicholas Kaldor, structural transformation was understood as the reallocation of labor and capital from low-productivity to more productive activities. Manufacturing took center stage because it combined economies of scale, technological learning, and sustained productivity growth. But that framework reflected the realities of an era in which factories were at the heart of technological progress and income growth. The goals of development have not changed since then, as productivity, technological learning, and economic complexity remain the foundations of long-term prosperity. What has changed is where—and how—those capacities emerge. Three factors are driving this shift. Artificial intelligence is rapidly reshaping the relationship between labor and production, while geopolitical fragmentation has put industrial policy back at the center of economic strategy. At the same time, the transition to clean energy is reorganizing production around new technologies, critical minerals, and low-carbon industries. Although none of these developments diminishes the importance of commodity production, together they suggest that it no longer determines structural transformation.
Increasingly, what matters is not the sectors themselves, but the productive capacities they encompass. Modern manufacturing depends on software, logistics, finance and digital platforms, while agriculture relies on biotechnology, satellite imagery and precision engineering. Sophisticated services now drive the technological learning, innovation and export dynamics that were once almost exclusively associated with manufacturing. As a result, the boundaries that once separated these sectors are gradually blurring. In many ways, capacities have always been the real engine of structural transformation. And as productive services increasingly take on the role that manufacturing has historically played, the most important question is which activities consistently expand an economy’s productive capacity. The answer lies in ecosystems that integrate manufacturing and engineering with research, digital infrastructure, finance and public institutions.
Consider electric vehicle batteries. Traditional industrial policy encourages countries like the Democratic Republic of Congo to process cobalt domestically, rather than exporting the raw ore. But domestic production alone no longer guarantees that countries will capture the most profitable parts of the value chain. Even if the DRC succeeds in producing battery components—or even entire batteries—most of the value will flow elsewhere as long as the design, specialized equipment, software, technical standards, and distribution networks remain concentrated abroad. China illustrates this dynamic. Despite its vast manufacturing capabilities, the country’s competitive advantage stems less from its scale of production and more from its control over the technologies, know-how, intellectual property, and industrial ecosystems in which manufacturing operates. The lesson is clear: structural transformation requires moving beyond production to ownership of the manufacturing ecosystem itself.
For Africa, however, this lesson strengthens the case for industrialization, as building technological capacity and capturing more value locally remain key priorities. If anything, artificial intelligence (AI) and the growing importance of manufacturing services are strengthening the role of manufacturing as a powerful source of learning, technological advancement, and capacity building. Yet Africa’s comparative advantages are also evolving. Unlike previous late developers, African countries are pursuing industrialization in a world where demographic change, climate transition, and digital technologies are transforming the development landscape. As advanced economies age, Africa is projected to account for the majority of the world’s growth in working-age populations and consumer demand. Because African countries are not constrained by legacy production systems, they can build renewable energy systems, low-carbon industries, and digitally integrated infrastructure from the ground up, rather than adapting carbon-intensive systems built over more than a century. But whether they can take advantage of these advantages will depend on their ability to build institutions that foster learning and innovation.
Industrial policy must evolve accordingly. Rather than simply promoting manufacturing or targeting strategic sectors, governments must create the conditions for productive capacity to expand. This means cultivating ecosystems that connect firms to universities, research institutions, financial systems, digital infrastructure, and Regional markets. Industrialization in Africa has always been a matter of building the institutional foundations of a modern economy. Industrial societies rely on formal markets, modern logistics, financial intermediation, enforceable contracts, and reliable data, all supported by institutions capable of managing complex economic interactions. Factories are just one part of that larger architecture. Here, emerging technologies offer new opportunities that previous latecomers did not have. Digital payments, identity systems, interoperable public platforms, and artificial intelligence can accelerate formalization, reduce transaction costs, improve tax administration, expand access to finance, and strengthen state capacity.
Artificial Intelligence (AI), in particular, has the potential to modernize the institutional architecture within which markets operate. Properly governed, it can help countries bridge the gaps in industrial development while creating new opportunities to develop and protect intellectual property. By challenging the exclusivity of production, Rodrik has done development economics a valuable service. The next step is to question sectoral exclusivity itself and rethink the way production is structured. In a world organized around digital networks, intangible assets, and industrial ecosystems, prosperity will depend on the ability of economies to accumulate and renew productive capacity. Rather than following the path of previous industrializers, Africa must seize this moment to create a new model of structural transformation.