For much of the late 20th century, Africa was associated with economic stagnation. Easterly (2001) famously referred to this period as the ‘lost decades’. The past 20 years tell a different story. As seen in Figure 1, since 2000, several large sub-Saharan African economies have grown at rates that exceed 5% annually – among the fastest rates in the world.

Figure 1 GDP growth in selected sub-Saharan African economies, 2000–2019

Notes: Bars show annual GDP growth and GDP per capita growth at constant 2021 international dollars (PPP). Green bars denote the West African sample including Benin, Burkina Faso, Côte d’Ivoire, Ghana, Guinea, Sierra Leone, and Togo; blue bars denote the East African sample including Ethiopia, Kenya, Mozambique, Malawi, Rwanda, Tanzania, and Uganda; WA (West Africa) and EA (East Africa) are population-weighted regional averages. We exclude oil producers and countries involved in long-running civil conflicts. 
Source: World Development Indicators.

Growth without factories

Yet, growth has unfolded along a development path that looks strikingly different from the textbook model of structural transformation. In that model, workers first move from agriculture to manufacturing, productivity rises through scale and technical change, and services expand later. This was broadly the path of Western Europe, the US, and much of East Asia. In many African economies, however, labour has moved from agriculture directly into services. Manufacturing has expanded only modestly. We see no widespread industrialisation.

Figure 2 Structural change without much industrialisation, 2000–2019

Notes: Bars show changes in sectoral employment shares. Green bars denote the West African sample; blue bars denote the East African sample; WA (West Africa) and EA (East Africa) are population-weighted regional averages.
Source: World Development Indicators.

The services that have absorbed workers are not tradable, skill-intensive business activities such as finance, law, or software development. They are mostly local, consumer-facing services: retail trade, food preparation, hospitality, personal services, repair, health, and local transport. Many of these jobs are concentrated in cities, which increasingly look more like consumer hubs built around domestic demand, rather than production hubs integrated into global value chains. This echoes the distinction between ‘production cities’ and ‘consumption cities’ emphasised by Gollin et al. (2016) and Jedwab et al. (2022).

These observations have fed a pessimistic narrative. Retail stalls, restaurants, motorbike taxis, and informal personal services are often associated with low productivity and low earnings. They may look like disguised unemployment rather than genuine development. From this perspective, the dominance of service-led growth is a symptom of failed industrialisation.

Our evidence suggests that this diagnosis is too bleak. In our earlier work on India (Fan et al. 2023) and new work on sub-Saharan Africa (Peters et al. 2026), we find substantial productivity growth in consumer services. These service jobs also do not appear to be uniformly ‘bad’ jobs. In household microdata from African censuses and surveys, workers in consumer services have access to electricity, piped water, durable housing materials, and basic amenities that are broadly comparable to those of manufacturing workers, and much better than those of agricultural workers. Informality remains widespread, but consumer services are not simply a residual sector of last resort.

Can consumer services raise productivity?

The key question is whether services are merely expanding because other sectors have made people richer, or whether services themselves are contributing to growth. We distinguish between service-biased growth and service-led growth. Growth is service-biased when productivity rises in agriculture, mining, and manufacturing or when educational attainment increases, raising incomes and therefore demand for services. Growth is service-led when productivity in services themselves improves, raising real incomes and expanding both demand and employment in services.

Using a spatial general-equilibrium framework, we infer service productivity from observed employment patterns, income, education, and geography. This avoids relying on local service price indices, which are often unavailable, poorly measured when available, and in any case hard to interpret because changes in service quality are difficult to capture. In our African sample, estimated consumer-service productivity grows by roughly 3%–4% per year, comparable to or faster than productivity growth in tradable activities. This indicates that these services were not merely expanding as a passive response to higher incomes elsewhere in the economy, but contributed to productivity growth in their own right.

Vietnam provides a useful contrast because its structural transformation followed the more conventional pattern of development that characterised growth in East Asia. McCaig and Pavcnik (2017) show that export opportunities encouraged movement out of agriculture and into formal manufacturing. In this respect, Vietnam looks less like the African and Indian trajectories, and more like the manufacturing- and tradables-centred reallocation associated with East Asian industrialisation (as in Song et al. 2011; see also Chen et al. 2023 on China’s later tertiarisation). Consistent with this interpretation, applying our methodology to Vietnam points to a much larger role for productivity growth in tradable sectors and a smaller role for consumer services.

How can local consumer services become more productive in developing countries? We highlight three mechanisms. First, market size matters. As cities grow and incomes rise, demand can sustain more specialised shops, restaurants, repair providers, and transport services. Greater density improves matching between buyers and sellers, encourages division of labour, variety, and quality, and raises effective productivity even without dramatic technological change. Recent evidence on African cities also shows that mobility can extend the reach of non-tradable services by allowing consumers to travel to where such services are supplied (Blanchard et al. 2025). 

Second, digital and basic technologies can transform informal services. Mobile phones, mobile payments, cheaper motorbikes, and simple platforms reduce search costs, improve inventory management, and enlarge markets. 

Third, formalisation can matter when modern service firms enter and workers move from very low-productivity informal activities into more productive establishments (Schwartzman 2025). This channel is powerful in some middle-income countries, but in much of sub-Saharan Africa formalisation is still limited, so productivity growth within informal services may be especially important.

The distributional catch

This more optimistic view does not imply that service-led growth is automatically inclusive. Demand for consumer services is local and income-elastic: these services are mostly available to urban households and appeal more to richer households who spend a larger share of their budgets on them. Productivity improvements in restaurants, retail, personal services, or urban transport therefore benefit richer urban consumers more than poor rural households. By contrast, productivity gains in tradable sectors such as agriculture, manufacturing, and producer services can lower prices economy-wide and spread gains more broadly. Service-led growth can raise living standards while also widening spatial and income inequalities.

Policy implications

The role (and the necessity of) industrialisation is a salient issue in the policy debate. While services have been the main lever of growth, many governments are investing in industrial parks, export zones, infrastructure corridors, and regional trade integration in the hope of replicating East Asia’s factory-led success. At the same time, global competition, automation, and geopolitical fragmentation have made the traditional route harder for latecomers. A series of VoxEU columns has framed this tension: Rodrik (2015) warned of premature deindustrialisation; O’Connell and Ghani (2017) argued that services can be a ‘growth escalator’; Sanfilippo et al. (2022) documented the growing importance of services jobs in Africa; and Hallward-Driemeier et al. (2022) argued that small service firms can still be productive.

Our take-home message is not that countries should abandon efforts to promote industrialisation. Manufacturing, agriculture, and producer services remain crucial development forces, especially in rural areas. But our evidence suggests that a development strategy focused only on factories could miss an important source of growth that is already underway. Policy should therefore treat consumer services as part of the productivity agenda, not merely as a residual sector that expands after industrialisation has taken place.

This is especially important because service-led growth may be less demanding in terms of institutional development than conventional manufacturing-led strategies. Successful industrialisation often requires a demanding bundle of complementary conditions: reliable infrastructure, coordinated investment, access to finance, export capabilities, regulatory capacity, and credible state support. Consumer services are not institution-free, but the threshold for productivity gains may be lower. They can expand through urban demand, better matching between firms and consumers, digital connectivity, improvements in management, and gradual formalisation, even where governance is weak and institutional development is still at an early stage.

The policy implication is not a laissez-faire approach to services, but a broader productivity agenda. That means investing in urban infrastructure, reliable electricity, digital connectivity, and payments systems; reducing entry barriers and harassment costs for small firms; improving management and skills; and making formalisation attractive rather than punitive. It also means linking cities and rural areas through transport, market access, and mobility, so that urban service growth does not simply create enclaves of prosperity. Better measurement is also needed: national statistics should track service quality, local prices, informality, and household welfare with the same seriousness traditionally reserved for manufacturing output.

In conclusion, Africa’s recent growth does not fit the classic factory-first narrative. That need not be a failure. ‘Skipping the factory’ can generate real persistent gains if consumer services become more productive. In this light, it is important not to let the perfect become the enemy of good. Policy should recognise service industries, including non-tradable consumer services, as an engine of growth while acknowledging that this path has limits. It is more local, and more likely to favour urban and higher-income households. The challenge for policymakers is to make service-led growth both more productive and more inclusive, while preserving opportunities in agriculture, manufacturing, and tradable services. The future of development may not be a choice between factories and services; it may depend on understanding how they can grow together.

References

Blanchard, P, D Gollin, M Kirchberger, and M Peters (2025), “Making non-tradables tradable: Human mobility and the spatial reach of African cities”, unpublished manuscript.

Chen, X, G Pei, Z Song, and F Zilibotti (2023), “Tertiarization like China”, Annual Review of Economics 15: 485–512.

Easterly, W (2001), “The lost decades: Developing countries’ stagnation in spite of policy reform 1980–1998”, Journal of Economic Growth 6(2): 135–57.

Fan, T, M Peters, and F Zilibotti (2023), “Growing like India: The unequal effects of service-led growth”, Econometrica 91(4): 1457–94.

Gollin, D, R Jedwab, and D Vollrath (2016), “Urbanization with and without industrialization”, Journal of Economic Growth 21(1): 35–70.

Hallward-Driemeier, M, G Nayyar, and E Davies (2022), “For services firms, small can be beautiful”, VoxEU.org, 12 January.

Jedwab, R, E Ianchovichina, and F Haslop (2022), “Consumption cities versus production cities: New considerations and evidence”, World Bank Economic Review 36(2): 327–52.

McCaig, B, and N Pavcnik (2017), “Moving out of agriculture: Structural change in Viet Nam”, in M S McMillan, D Rodrik and C Sepúlveda (eds.), Structural Change, Fundamentals, and Growth: A Framework and Case Studies, Washington, DC: IFPRI.

O’Connell, S, and E Ghani (2017), “Services as a growth escalator in low-income countries”, VoxEU.org, 15 June.

Peters, M, Y Zhang, and F Zilibotti (2026), “Skipping the factory: Service-led growth and structural transformation in the developing world”, NBER Working Paper 34692.

Rodrik, D (2015), “Premature deindustrialisation in the developing world”, VoxEU.org, 12 February.

Sanfilippo, M, M Fiorini, B Hoekman, and L Baccini (2022), “Services, jobs, and economic development in Africa”, VoxEU.org, 25 January.

Schwartzman, M (2025), “From street markets to shopping malls: The modern service multiplier”, unpublished manuscript.

Song, Z, K Storesletten, and F Zilibotti (2011), “Growing like China”, American Economic Review 101(1): 196–233.