The Quiet Revolution In Services

By Dani Rodrik, Project Syndicate

While economists debate whether industrialization can still play its traditional growth-promoting role in the current global and technological context, services have unexpectedly become an engine for economic development around the world. For developing and developed economies alike, the timing could not be better.

CAMBRIDGE—In 2023, the University of Chicago economists Chang-Tai Hsieh and Esteban Rossi-Hansberg published a study showing that services in the United States were undergoing a radical transformation. Digital technologies and organizational innovations in retail, hospitality, personal services, and medical services, they argued, had enabled some firms to become much more productive and expand their presence in local markets previously served by smaller, less efficient establishments. It was an “industrial revolution in services.”

Although Hsieh and Rossi-Hansberg suggested that much of the productivity gain from this trend did not show up in official statistics, even the measured upsurge in services productivity has been impressive. Between 2000 and 2024, labor productivity in the US grew at an annual average rate of 3% in food-service establishments, 2.1% in accommodation services, 2.9% in retail trade, and 4.4% in wholesale trade. Compared to manufacturing, where labor productivity growth has essentially stagnated at an annual rate of 0.2%, the record in services is nothing short of extraordinary.

This performance is as unexpected as it is welcome. It is unexpected because the traditional view holds that services are a drag on the economy’s productivity (what economists call Baumol’s cost disease); and it is welcome because services now dominate economic activity and provide the bulk of jobs. Higher productivity is a necessary (if not sufficient) condition for turning precarious jobs in services into the good jobs that inclusive prosperity requires.

Even better, the services revolution appears to be a global phenomenon that has already spread to developing countries. While economists and policymakers debate whether industrialization can still play its traditional growth-promoting role in the current global and technological context, services have quietly turned into an engine of economic development around the world.

When services are discussed in the context of economic growth in developing countries, the focus tends to be on tradable, skill-intensive activities such as information technology, business processing, or financial services. But by their very nature, these services are not well suited to low-income contexts. They can absorb only a small share of the labor force—the same disadvantage (from a development standpoint) as manufacturing today.

Instead, the current revolution is happening in the considerably less glamorous, but more consequential services: restaurants, deliveries, ride-hailing, grocery stores, gig work, and even care. Recent academic research has shown that the expansion of middle-class services has been an important driver of economic growth in India, Sub-Saharan Africa, and Latin America.

In India, for example, advances in IT and business process outsourcing have certainly been important for graduates of the country’s technology institutes. But the vast, less-educated majority of the workforce has benefited primarily from the expansion of what Yale economists Tianyu Fan, Michael Peters, and Fabrizio Zilibotti call consumer services sold on local markets.

Moreover, a November 2025 International Monetary Fund report finds that the marginal product of labor in India’s services now exceeds that in manufacturing, partly because labor productivity in the manufacturing sector has stalled in recent years. Taken at face value, this means that moving a worker from manufacturing to services would increase India’s overall income.

To be sure, even if manufacturing firms do not directly generate much employment, they can have indirect effects on the economy through linkages to other firms and the technological spillovers they often generate. This point is often used to counter the argument that services offer a viable development model. But the manner in which global value chains are organized today almost guarantees negligible spillover benefits from export-oriented manufacturing. Since the bulk of intermediate and capital inputs are imported, only a thin segment of value added is generated domestically, and any backward spillovers are minimized.

By contrast, modernized service establishments can offer significant linkage opportunities. For example, discount retail chains typically operate under private label, which means that they offer predominantly locally manufactured goods. In Colombia, retailers such as D1 and Ara rely almost exclusively on local sourcing. They work with hundreds of local suppliers across the food and grocery value chain, providing them with the benefits of scale and access to larger markets. Importantly, these supply chains enable technological upgrading in areas such as packaging, traceability, and cold chains, as well as stronger enforcement of quality standards. They have been shown to increase formal employment in manufacturing and agriculture by several percentage points, without adverse effects on employment in general.

Unpublished work by Victo Silva, a visiting researcher at Harvard’s Center for International Development, documents a similar phenomenon in food delivery services in Brazil. Consider the platform firm iFood, which serves 400,000 restaurants of all sizes. Beyond delivering meals, it provides financial, logistical, customer-service, and market-analysis tools to food service establishments, enabling productive upgrading throughout the sector.

As this evidence shows, the potential for productivity gains in labor-absorbing services is real and significant. The risk is that the benefits will be appropriated by employers, especially large chains and platforms that can exert market power over their employees, customers, and suppliers. In the US, Uber drivers or Amazon warehouse workers have not experienced major improvements in wages and working conditions despite the remarkable productivity gains these platforms have enabled.

This is no different from what happened in manufacturing industries in their early days. For the gains of workplace productivity to be widely shared, corporations must face countervailing power. Competition policy, regulation, and a stronger voice for workers will be as important in services as they were in manufacturing. In the absence of productivity gains, demands for higher labor standards may conflict with overall employment. But if the productivity revolution in services can be sustained, it will help developed and developing countries alike avoid this cruel tradeoff.

Dani Rodrik, Professor of International Political Economy at Harvard Kennedy School, is Past President of the International Economic Association and the author of Shared Prosperity in a Fractured World: A New Economics for the Middle Class, the Global Poor, and Our Climate (Princeton University Press, 2025).