Although urbanization began millennia ago, for most of human history cities were home to only a few. At the start of the first industrial revolution in 1765, less than 10% of the population lived in urban areas and economic growth depended largely on the productivity of rural labour. Since then, urbanization has grown at a steady and revolutionary rate in all regions, fueled by a growing demand for urban workers and technological advances such as the construction of high-rise buildings.

Today, more people live in urban areas than in rural areas, and growth increasingly depends on the productivity of the urban labor force. Urbanization rates are around 70% in the developed world and middle-income countries in Latin America, Europe, the Middle East and North Africa, and are increasing rapidly in other developing economies . What are the forces shaping urban employment? Why do cities of the same size located in countries with similar income levels differ in terms of employment composition? What are the implications of these differences for development?

Economic theory provides some insight into the differences in the composition of employment along a country’s urban hierarchy by highlighting the importance of agglomeration economies, which capture the benefits to firms from the presence of other firms in the same area. This allows for the growth of cities that are competitive in the production of tradable goods, such as clothing, electronics, and pharmaceuticals. The same agglomeration forces that allow production towns to grow also make it difficult for new market activities to grow elsewhere. Meanwhile, adjustment to adverse shocks increases the number of consumption cities in which urban residents work in retail and other non-market sectors. However, it is not known how the cities of the same size of the population located in countries with similar income levels differ in terms of employment composition.

The cities of the world: consumption, production or neutral?

We answer these questions in a recent article, which documents the employment composition of 6,865 functional urban areas, using newly available microdata for 74 countries for the period between 1960 and 2015. The selected urban areas include 3 billion people and represent three-quarters of the world’s urban population. We classify each urban area either as a production city with a disproportionately high share of employment in urban tradable goods, a consumer city with a disproportionately low share of employment in urban tradables, or a neutral city where the share of urban tradable goods in employment is neither too low nor too high. The distribution of producing (blue), consuming (red) and neutral (grey) cities, shown in Figure 1, indicates that the employment composition of cities varies significantly between cities of similar sizes in countries at different similar levels of development. Production cities are located primarily in China, Europe, India, and parts of the United States, Mexico, Central America, and Brazil. Apart from a few major production cities in Malaysia, Vietnam and South Africa, other cities in Asia and sub-Saharan Africa are either consumption cities or neutral cities.

Figure 1. Global distribution of production, consumption and neutral cities, circa 2000

Source: Jedwab, Ianchovichina and Haslop (2022) using IPUMS census data and the Global Human Settlements Layer database. Note: Lighter shades of each color indicate lower values ​​for how well a city can be classified into each specific type.

More paths to a consumer city

There are more paths to a city of consumption than to a city of production. Production cities emerged with industrialization, while the “origins” of consumption cities can be traced to (i) resource rents, (ii) rents from agricultural exports in countries with sufficiently high agricultural productivity, and (iii) “premature” deindustrialization. The latter has not led to deurbanization, but to the deindustrialization of cities, especially the largest ones, as shown in Figure 2 for the case of Latin America.

Figure 2. Share of employment in tradable goods by city size and decade in Latin America

Figure 2. Share of employment in tradable goods by city size and decade in Latin AmericaSource: Jedwab, Ianchovichina and Haslop (2022) using IPUMS census data and the Global Human Settlements Layer database. Note: MFGFIRE refers to manufacturing, financial, insurance and real estate services.

The “origins” of urbanization matter for the largest cities

Compared to cities in industrialized countries, cities of similar size in resource-rich and deindustrializing economies have lower employment shares in manufacturing, market services, and the formal sector, and employment in the non-market and informal sectors. In industrialized countries, the share of tradable goods in employment is high while that of non-tradable goods is low in cities of all sizes. In both resource-rich and deindustrializing countries, large cities have significantly higher shares of nontradables than small cities, but for all city sizes, these shares are much larger in non-tradable countries. deindustrializing than in resource-rich countries. Thus, the “origins” of urbanization matter for the largest cities, which are the main locations and engines of growth for nations (Figure 3).

Figure 3. Share of employment by city size and origin of urbanization, c. 2000

Figure 3. Share of employment by city size and origin of urbanization, c.  2000Source: Jedwab, Ianchovichina and Haslop (2022) using IPUMS census data and the Global Human Settlements Layer database. Note: MFG+FIRE refers to manufacturing, financial, insurance and real estate services; NTR2 includes wholesale and retail trade and other domestic trade-related activities.

Trapped in consumer cities?

Having mainly consumer cities would affect the potential of developing countries to catch up with the living standards of advanced economies and could even explain why some countries find themselves in a middle-income trap. Agglomeration economies may be weaker in countries where the share of urban employment in nontradables is disproportionately high (Venables, 2017; Burger et al., 2022). Agglomeration economies can even besterilein these countries due to congestion, which may be more detrimental to firms in sheltered sectors that benefit relatively less from supply-side agglomeration economies (Burger et al, 2022). While consumption cities do not have significantly less human capital than production cities, for a given level of human capital, consumption cities have more informality and more jobs in urban non-traded goods than production towns. We also find higher wages in urban tradables than in nontradables and lower returns to experience in countries with more consuming cities. This implies that human capital in consumption cities is employed in less productive sectors and jobs.

History also teaches us that the composition of employment in cities can change dramatically over time. The growth of many cities was initially fueled by agricultural rents and resource exports. Industrial revolutions transformed many of these consumer cities into thriving centers of production. Today, countries can seize new opportunities to leverage their human capital and Fourth Industrial Revolution technologies and transform their cities into thriving centers of production. To succeed, they will need to address technology skills gaps, increase investment in digital infrastructure, and foster innovation and reforms that strengthen their private sectors.