Location of Jobs in Developing Cities
Author: Kyu Sik Lee

For the latest thinking about the international financial system, monetary policy, economic development, poverty reduction, and other critical issues, subscribe to Finance & Development (F&D). This lively quarterly magazine brings you in-depth analyses of these and other subjects by the IMF’s own staff as well as by prominent international experts. Articles are written for lay readers who want to enrich their understanding of the workings of the global economy and the policies and activities of the IMF.


For the latest thinking about the international financial system, monetary policy, economic development, poverty reduction, and other critical issues, subscribe to Finance & Development (F&D). This lively quarterly magazine brings you in-depth analyses of these and other subjects by the IMF’s own staff as well as by prominent international experts. Articles are written for lay readers who want to enrich their understanding of the workings of the global economy and the policies and activities of the IMF.

For predicting and managing future urban growth, it is essential to understand the effects of development policies on location patterns of employment

The rapid urbanization in developing countries has generated enormous pressure for policymakers to control or direct urban growth and to expand public services. In the middle-income countries of Latin America and East Asia (including Brazil, Mexico, Venezuela, the Philippines, and the Republic of Korea) and in a number of countries in Africa and South Asia (such as Egypt, Nigeria, India, and Pakistan), policymakers have used many approaches to cope with the rapid growth of cities. Often the governments of these countries have initiated specific policies to decentralize economic activity and move it to the periphery of metropolitan areas.

Managing large municipalities is a complex and difficult task, mainly because the process of rapid growth and its effects on the functioning of large cities are not well understood. To deal efficiently with the problems of a rapidly growing city requires a policy framework that links a sound urban development strategy with appropriate investment programs. Such a policy framework, in turn, requires a solid understanding of the trends in urban development and how the workings of the land and labor markets and the location choices of individual households and firms contribute to such trends. Government policies and programs tend to be inefficient and costly if they attempt to reverse observed trends and behavior.

This article is based on the author’s book, The Location of Jobs in a Developing Metropolis: Patterns of Growth in Bogota and Cali, Colombia, published for the World Bank by Oxford University Press, 1989. $18.95.

Many urban policies aim to influence the location pattern of employment. This is not surprising since urbanization and the concentration of economic activities in most developing and developed countries are dependent on the location of employment opportunities. Contributing to the location of jobs are the nature and pattern of industrialization, the pace of agricultural development, and the growth of transportation and communication networks.

In most countries, employment location policies for cities have taken various forms: strict zoning regulations, outright prohibition of certain economic activities in particular areas, and various fiscal and financial incentives to induce industries and population to particular areas. The impact of these policy measures is uncertain mainly because of the lack of information on trends of employment location patterns within developing cities and the lack of understanding of how individual firms respond to market forces in selecting their locations.

The City Study

To fill this information gap, a World Bank research project known as the City Study investigated the workings of five urban sectors—housing, transport, employment location, labor markets, and public finance—in Bogota and Cali, Colombia, to help understand and better assess the impact of urban policies and investment programs in developing countries. This article presents the results obtained from research on employment location.

Bogota, with a population of nearly 4 million in 1980, was representative of large Latin American cities. Cali, with about one million people, was included in the study as a comparator to help test the research hypotheses. During the period under review (the mid-1970s), the Colombian economy was growing, and both cities were growing rapidly at about 5 percent per year.


The City Study analyzed several large data files on individual decision-making units, such as households and firms in Bogota and Cali. The analysis of the data confirms the stylized facts on employment location patterns obtained for North American cities; they are also consistent with, and reinforced by, similar results obtained for Seoul in a later Bank study that evaluated industrial location policies in Korea. This suggests that the development of cities in general in both developing and developed countries follows regular patterns, and that the research findings summarized below are applicable to cities in other countries.

Location patterns of employment are changing rapidly. The extent of changes in the location of jobs in cities in developing countries was unknown until this study of Colombian cities. In terms of birth (establishment of a new firm), death (termination of an existing firm), and relocation of firms, Bogota and Cali experienced a much higher degree of change than those of large US cities. The annual birth rate of manufacturing firms in Bogota at 8.8 percent was about twice that of large US cities. The birthrate of 7.6 percent for Phoenix came closest to that of Bogota and is comparable to that of Cali. The annual moving rates of firms were similar for North American and Colombian cities, ranging from 3 to 5 percent. This means that in a ten-year period, as much as half of all manufacturing firms in these cities would be relocated.

The locations chosen by relocating and new firms influence the spatial pattern of an urban area because the location of employment influences residential locations and travel patterns. Therefore, understanding the location patterns of new, discontinued, and relocating firms helps policymakers predict trends correctly and avoid wrong public investments.

Jobs are moving outward from central areas. The data revealed evidence of employment decentralization in Bogota and Cali. For analytical purposes, Bogota was divided into six concentric rings around the central business district (CBD) and Cali into five rings. In both cities, manufacturing employment grew at an accelerating rate with the distance from the center. In Bogota, the outer ring gained employment in manufacturing by 16 percent a year, while the center lost by 2 percent a year. The strong decentralization of manufacturing employment in these cities is similar to the trend observed for large US cities during the past several decades. Similar trends were observed for Seoul, although the magnitude of change was four times higher there, possibly because of its faster rate of industrialization. The center in the two Colombian cities, however, continues to retain retail and service employment and attracts large commercial and financial establishments. Consequently, the absolute number of jobs in the CBD has remained roughly constant, although the center’s share of all employment has fallen over time. In 1978, the CBD in both cities had about a sixth of all city employment. This pattern indicates that CBD-directed travel is decreasing as a share of total travel and that the existing transit system, linking the periphery to the center, should be supplemented with services connecting sites around the city.

New, small firms create more jobs than large, established ones, and do so in central areas. In Bogota and Cali, the birth of new firms created many more jobs than did the growth of mature firms (60 percent more in Bogota and 80 percent more in Cali). This is consistent with findings on North American cities. In both Bogota and Cali, small, new firms tend to locate in central areas, a tendency that supports the incubator hypothesis: certain central areas of large cities serve a special function in promoting the birth of small new firms that is vital to the economy and not easily transferable to outlying areas or to smaller cities. These firms, whose space needs are still modest, choose to locate in established areas despite higher rents in order to benefit from the availability of skilled labor, markets and business services, and shared delivery services. As these new firms grow and expand, space constraints become more significant, and they tend to move out to where more space is available.

Three criteria were used to identify such areas that acted as incubators: (1) the area must have a substantial share of the total number of jobs—5 percent or more; (2) the area’s share of new jobs must be higher than its share of total jobs; and (3) the average size of new firms must be small—not more than 25 or so employees. Indeed, an analysis of the data revealed that areas close to the CBD showed the characteristics of an incubation area in both Bogota and Cali. The data on Seoul also identified three such areas: one was the central business district, another was the oldest, largest market in the eastern part of the city, and the last was the fast-growing industrial district of Bucheon. This suggests that crowded central areas are good at “hatching” new industries in these cities.

Firms move, but not very far. In both Bogota and Cali, 40 percent of relocating firms stayed within their original subareas; those in Bogota moved an average distance of less than two kilometers. The majority of relocating firms were small, moving only a short distance from their previous location. Only a few large firms moved long distances. The 1982 national survey of industries in Korea showed that only 7 percent of those firms that relocated moved out of the Seoul region. Although the moving distance increases with the firm size, firms tend to move in such a way as to leave unaffected the distance involved in delivering inputs and outputs of their product and the commuting distance of their employees.

Firms respond differently to location attributes. The survey data were analyzed to determine location choices made by different types of firms. The results show that accessibility to the local input and product markets and the commuting distance of production workers are the most important factors in the location choice of small manufacturing firms. For these firms, the benefits from various externalities tend to compensate for high rent and congestion costs in the central area, as the incubator hypothesis suggests. Large firms tend to be more export-oriented (from the region) and require more space for modern assembly-line production technology. For these firms, land and plant space, available at lower cost in outer areas, are more important than access to local markets. Such locations also offer better access to the road network. It was striking to find from the survey results that almost all shipments of inputs and final products are made by truck; rail is seldom used. This is consistent with the observed trend of decentralization of manufacturing establishments. The central location near the rail terminal no longer has comparative advantages as highway interchanges have developed around the city.

Firms respond to the land market. Analysis of the survey data indicates a strong relationship between the price of land and intensity with which labor and capital are used; that is, land is used with greater intensity when its price is high. This implies that in rapidly growing cities in developing countries, manufacturing firms weigh the price of land versus other inputs in making location decisions. The survey also confirms that firms in trade and services choose a location that maximizes their access to customers with relatively high purchasing power. Changes in the location pattern of employment in cities such as Bogota and Seoul are therefore by no means random.

Policy implication

Although many jobs moved away from the centers of large cities during the past several decades, much of the movement resulted from the responses of firms to the price of land and other market forces rather than to explicit government policies. Even in the case of Seoul, where the government offered strong incentive schemes for decentralization, such as tax breaks, loan guarantees, and relocation grants to industries, a small proportion of firms moved in response to government actions. Newly established firms accounted for nearly 80 percent of new jobs created in Seoul, but they were little affected by the government incentive schemes.

In contrast to the United States, evidence from Bogota and Cali does not suggest an increasing concentration of low-income population in the central city, and there are few signs of central city decay. In fact, attempts to decentralize economic activity in developing countries stem mainly from the increasing concentration of economic activity in the central city, accompanied by congestion and pollution as the city’s population grows rapidly.

But policies to decentralize population and economic activity are probably not good substitutes for better internal management of city growth. For example, reducing the population or employment in a large city by a certain amount is likely to have very little effect on air pollution or traffic congestion. The principal question is how to guard against poorly conceived spatial policies when decentralization is already prevalent. Since the location of jobs can have a significant impact on the patterns of residential location, commuting, and hence overall urban development, inappropriate spatial policies could seriously reduce the welfare of the economy.

A commonly observed but potentially damaging spatial policy is the outright legal prohibition of new manufacturing activities within large cities. Several developing countries, including India, Korea, and Venezuela, have implemented such measures as part of their decentralization policy. Such government actions to slow further growth of the city may kill the incubation effect. Since governments tend to encourage young industries to start up away from large city centers, the important policy question is whether the incubation function could be replicated in outlying areas of large cities without becoming prohibitively costly. A related policy question is how to strengthen the incubation function of secondary cities where indigenous manufacturing industries could be “hatched” and helped to grow. This could be a potentially viable policy that would lessen the pressure of growth in the major urban centers.

The findings from the study on Seoul suggest that the costs of government subsidies to compensate small firms for moving long distances will be extremely high. In 1978, the Korean Government established a new industrial town at Banweol, less than 30 kilometers from Seoul, to draw small and medium-size firms away from the capital. One thousand plant sites were initially prepared, but the occupancy rate remained low until the size restriction on firms was removed several years later. The most serious problems faced by many firms that moved to Banweol were that product markets and input suppliers were not readily accessible, production workers were reluctant to move to Banweol or commute from Seoul, and day-to-day business information was difficult to obtain because of poor telephone services and the lack of person-to-person contacts.

That such a seemingly short distance thwarted the government’s attempt to develop Banweol was striking. In contrast, the government discourages firms from locating in Bucheon, a fast-growing industrial city between Seoul and Inchon. Firms have to pay a penalty of 500 percent of local taxes to locate there. Nevertheless, many are willing to pay this penalty because the opportunity cost of moving to government-supported areas such as Banweol is even higher. Analysis of the Bogota data helps explain the Korean experience: small firms prefer central locations. Accessibility to local markets and proximity to production workers are the most important attributes for them. Hence, government policies intended to influence employment location patterns generally tend to be ineffective; infrastructure investments alone may not attract firms to newly developed areas.

Trends in Developing Economies 1989

The Bank’s newest annual data collection provides concise textual and statistical profiles of some 90 developing economies. Each profile includes a comprehensive table of economic indicators with the latest data available. Most profiles also include an analysis of the prospects for the economy and the main elements of the country’s development strategy.

544 pages/ISBN 0-8213-1358-4/US$19.95/Order Stock #11358

Other recently published data collections:

World Bank Atlas 1989

The 1989 edition of this classic reference provides data for a wide range of economic and social indicators. Included are the latest 1988 data. Colorful and easy-to-read world maps, tables, and graphs highlight key relationships and trends in the development of some 185 countries and territories. Text in English, French, and Spanish.

30 pages/ISBN 0-8213-1354-l/US$5.95/Order Stock #11354

Social Indicators of Development 1989

Key social indicators for more than 130 developing countries. This 1989 edition, published by the Johns Hopkins University Press, contains the latest World Bank data available.

About 275 pages/ISBN 0-8018-4006-6/US$24.95/Order Stock #44006

Mark the number of copies desired in the boxes.

Coupon orders accepted only from the U.S.A. and France. To order from other countries, request a list of local World Bank Publications Distributors. Prices may vary by country.

In the U.S., return coupon to:

World Bank Publications

P.O. Box 7247-8619

Philadelphia, PA 19170-8619

In France, return coupon to:

World Bank Publications

66, avenue d’Iéna

75116 Paris, France