Journal Issue

Re-Examining Urban Poverty and Employment: Improved Incomes and Labor Productivity Hold The Key to Better Results

International Monetary Fund. External Relations Dept.
Published Date:
March 1986
  • ShareShare
Show Summary Details

Friedrich Kahnert

Urban poverty is a pervasive problem in many, if not most, developing countries. One fifth to one third of urban households in many middle-income countries earn too little to feed themselves adequately. In lower-income countries (or regions of countries), half or more of the urban population may fall into this category. Urban poverty has probably worsened in the recent recession years with many countries experiencing falling real incomes.

In these times of austerity, it is becoming even more urgent to raise urban incomes. Even in low-income countries, urban areas now account for half or more of the nation’s measured gross national product. The potential for raising public revenues ultimately depends on raising income, and urban incomes form an increasingly important part of the nation’s tax and savings base. Enhancing urban employment and labor productivity thus become key objectives both for alleviating poverty and for improving local resource generation.

When the World Bank first focused its attention on the relief of urban poverty in the early 1970s, a two-pronged approach was proposed: (1) improving the access of the poor to housing and related services, and (2) enhancing employment opportunities for the poor. Given the Bank’s lack of experience with poverty-oriented urban development, both approaches were conceived as experimental, intended to foster “learning-by-doing.”

It was possible to demonstrate that low-cost housing and services, adapted to local conditions, were in demand by the poor and could be made affordable to them. Sites-and-services schemes designed to provide land and ancillary facilities for housing (instead of construction of completed houses) and slum-upgrading became the main operational tools in this effort.

No consensus existed on how to promote job creation. Indeed, experts disagreed on basic questions. For example, was it desirable to increase jobs in urban areas or would this simply accelerate rural-to-urban migration and swell the ranks of the unemployed and underemployed urban poor? Would it be better to try to reduce “excessive” migration primarily by promoting rural development? Many similar questions remained unresolved with the result that the Bank’s operational pursuit of urban job creation remained tentative and modest.

Issues in retrospect

A decade of experience permits us to take a fresh look at the problem of urban poverty and its likely solutions. This evidence suggests that the primary emphasis on creation of urban jobs per se may be misdirected and should be complemented or perhaps replaced by an emphasis on labor productivity and income generation. A consensus may now be developing on what should be done and how.

Job generation in urban areas is still important. Urban population in developing countries has grown very rapidly over the last 30 years, averaging 4.3 percent a year compared with 2.1 percent in developed countries. In 1980 the urban areas of developing countries had about 3½ times more people than in the 1950s, and African urban populations more than quadrupled over these three decades. Even where urban growth rates are declining, unprecedented numbers of people are being added to urban populations. Growth rates of large metropolitan areas, particularly in Latin America, may have peaked in the 1940s and 1950s and they will probably decline further. The population of São Paulo, for instance, grew by about 2 million, at an annual rate of 6.5 percent in the 1950s; in the 1970s, the city’s annual growth rate fell to 4.4 percent, but even that meant 4.3 million people were added during that decade. São Paulo will still add the same number over the 1990s, even if its growth declines to 2.6 percent a year.

Urban jobs will have to grow just as rapidly and perhaps faster than the urban population, because declines in natural population growth rates (which are now occurring in many countries) affect the growth of the labor force only after a lag of about 15 years and because female participation in the labor force is increasing in many countries. Rural-to-urban migration is undeniably an important contributor to urban growth, but it has been less important than natural population growth for over 20 years. Moreover, notwithstanding differences in regional experience, the contribution of migration to urban growth is declining almost everywhere, even though the actual number of migrants may still be rising.

The intense concern with urban job creation originated from widely held fears that urban areas would be unable to provide income-earning opportunities for their rising populations. These fears have proved unfounded. In fact, urban areas seem generally to have done well in productively absorbing additions to their labor force, even though often at low income levels. This is prompting the shift away from an exclusive emphasis on job creation to improving labor productivity and incomes.

Urban labor markets

Much of the vast literature generated by the concern with job creation saw extensive imperfections in urban labor markets that impeded the matching of labor supply and demand, causing large and unjustified wage differentials. Clearly, wages in urban areas differ greatly between high- and low-wage sectors and subsectors, large and small firms, the public and the private sector, and so on. But these differences in themselves do not tell us that the market is not functioning properly. This must be judged by whether or not:

  • There is a serious and growing problem of urban labor surplus, in the form of open unemployment or underemployment;

  • Wages for workers with similar characteristics tend to be equalized across sectors and firms;

  • Wage differentials respond over time to changes in labor supply and demand.

The perception that developing countries were experiencing a growing urban labor surplus, in the form of open unemployment, underemployment, or employment hidden in tertiary (trade and services) activities or in the informal sector (including construction, manufacturing, etc.), was a key reason for the growing concern with urban job generation. The evidence, while partial and far from flawless, consistently challenges this perception.

First, there is no evidence that open urban unemployment rates are generally rising. The available evidence suggests these rates may have been falling over the 1970s. Double-digit rates of open urban unemployment have become less common in a number of countries, such as Colombia, the Republic of Korea, the Philippines, and Syria. In others, for example Brazil, Egypt, Mexico, and Thailand, unemployment rates were not in the double-digit range to start with and they remained stable or even declined. Similarly, underemployment is nowhere near as large as frequently claimed, if defined as involuntary short-time work and correctly quantified as the difference between actual hours worked by those on involuntary short time and the standard work week in the country concerned. Crude underemployment rates of 20–40 percent or more are often cited; in reality, if all those who want to work more actually put in a full work week, total hours worked would not rise much. For instance, in Mexico City and Guadalajara, about 20 percent of the work force worked less than a 40-hour week in early 1979 but only 5.8 percent wanted to work more hours.

Table 1Indicators of the economic importance of some urban areas in developing countries
Urban areaPopulation

Output measure (In percent)
Brazil: Greater São Paulo (1970)8.636.0 of net domestic product
China: Shanghai (1980)1.212.5 of gross industrial production
Haiti: All urban (1976)24.257.6 of national income
Port-au-Prince (1976)15.038.7 of national income
India: All urban (1970/71)19.938.9 of net domestic product
Kenya: All urban (1976)11.930.3 of national income
Nairobi (1976)5.220.0 of national income
Mexico: All urban (1970)60.079.7 of personal income
Federal District (1970)14.233.6 of personal income
Pakistan: Karachi (1974/75)6.116.1 of GDP
Peru: Lima (1980)28.043.0 of GDP
Philippines: Metro-Manila (1970)12.025.0 of GNP
Thailand: Metro-Bangkok (1972)10.937.4 of GDP
Turkey: All urban (1981)47.070.0 of GNP
Source: World Bank.
Source: World Bank.

Another frequently expressed view holds that a growing proportion of the labor force is channeled into urban tertiary activities where they form a pool of “surplus” and “unproductive” labor. It is not clear why income received from tertiary activities should be considered inferior to income from the primary and secondary sectors; the income recipient certainly does not think so. In any event, the evidence, though scanty, does not support the hypothesis that the urban tertiary sector is absorbing ever larger proportions of surplus labor. Instead, there is some evidence that the share of tertiary activities in urban employment has remained quite stable over lengthy periods in the developing countries. In addition, average incomes in this sector compare well with those elsewhere in the urban economy. The best guess now is that the urban tertiary sector in developing countries is growing primarily because there is demand for its output and not because people are being forced into it for lack of alternative income-earning opportunities.

The preceding generalizations pertain to the 1970s, a period when developing countries, as a group, experienced rapid growth of aggregate output. More recently, per capita incomes in many developing countries have stagnated and even declined, particularly in Latin America and Africa. In parallel, the recession of the early 1980s has caused urban unemployment to increase but this is a cyclical change and is not necessarily part of a secular rise in unemployment rates. In any case, whether or not the past employment trends, analyzed above, continue is a separate question from that of the light they shed on the functioning of labor markets.

The view that high and low wages observed side by side in urban areas portray severe malfunctioning of the labor market is not supported by research. Attempts to verify the existence of segmented high-wage and low-wage sectors have largely failed to document the importance of institutionally induced wage differentials or rigidities. The major part of wage variance in urban activities, at least in the private sector, is explained by productivity factors. And informal sector earnings are not universally lower than those in the formal sector; a large portion of informal sector incomes are well above what labor with comparable skill levels can earn in the formal sector. No wonder that many informal sector participants have no intention of leaving the sector to take a formal wage job.

The urban labor market is best viewed as a continuum of employment and income-earning opportunities ranging from very low to very high income levels. Many imperfections can be demonstrated, particularly along ethnic or sexual lines, by barriers to entry into specific activities of a formal or informal nature, by government and union policies and restrictive practices, or by inadequate information flows, but they do not appear to have an overriding impact on the functioning of the labor market. This conclusion may not fully apply to public sector wages and salaries and, a fortiori, public employment policies, which may be the source of significant economic losses.

Urban poverty is by no means confined to the informal sector but occurs in all sectors of urban activity. Moreover, urban poverty is not synonymous with urban unemployment. Many unemployed are poor but many others are from families that were able to support their efforts to acquire education. More important, many of the poor in the labor force are not unemployed or on involuntary short-time work but put in long hours at low wages, or earn reasonable incomes but must use them to support large numbers of dependents. Thus, while more jobs and more working hours will help some of the poor, others can be helped by increasing their productivity and income per hour worked. Those not in the labor force, namely, the too young, the too old, and the infirm who are not part of households that have working members, cannot be helped by purely economic progress; their problems require other solutions.

The experience of migrants does not conform to earlier expectations either. They are not nearly as grossly overrepresented among the unemployed, the underemployed, those in the informal sector, and the poor as is widely believed. Some evidence even shows them to do as well or better than urban natives both in formal employment and in earnings across the board, at all education and age levels.

What can be done?

If persistent urban poverty problems cannot solely or primarily be blamed on labor market performance, and earnings are indeed largely determined by labor supply and demand, one must act on these in order to raise employment, productivity, and incomes. Reform of overall national policies wherever these show pronounced pro- or anti-urban biases, whether explicit or inadvertent, could be a prime tool for encouraging efficient urban employment and income generation. However, other more specific actions are just as necessary and do not have to wait for overall policy reform to remove conditions that inhibit productivity and income growth. Removing or attenuating these inhibitions will often be beneficial even if labor markets actually function less well than the preceding discussion seems to show. For example, upgrading the quality of labor will pay handsome dividends, even if labor mobility falls short of equalizing wages for workers with similar characteristics across sectors and firms.

Table 2Proportions of population living in urban areas
Share of total population
1950196019701980 (estimate)
Developed regions52.558.264.770.2
Less developed regions16.721.925.830.5
Less developed regions without China and India21.026.031.337.1
Source: United Nations, Patterns of Urban and Rural Population Growth, 1980.
Source: United Nations, Patterns of Urban and Rural Population Growth, 1980.

Education and health. The best way to enhance the capacity of workers to earn better incomes is to improve their education and health. Towns and cities appear to be much better equipped to do this, in general, than rural areas. Yet there are serious gaps within cities; for example, access to primary education in low-income urban neighborhoods can be very substantially below that of many rural areas. Some extra health risks are also intrinsic to urban life, such as increased risk of traffic accidents or the easier spread of communicable disease in high density settlements. More generally, the incidence and severity of nutritional, especially caloric, deficiencies, is often higher in low-income urban areas than in many rural areas, and access of the urban poor to health services is often limited.

The impact of education on labor productivity and earnings is well studied, clear, and direct; returns are often highest for primary education. The impact of health is more difficult to prove statistically but is not in doubt. It is most obvious through the intimate connections between health, education, and nutrition, and through the direct impact of debilitating and endemic diseases on the earnings of casual workers and on the level of sustainable effort when at work.

Programs to improve the income-earning capacity of the urban poor must improve access of the poor to education and health services. They must be concerned with not only the provision of physical infrastructure for primary education, basic health centers, and vocational training but also what is being taught and how. Many of the potential customers for these services, particularly if they are the breadwinners, cannot leave their jobs to improve their education or get health advice. Even primary school children have important functions in their families, such as helping in the family business, earning occasional income, or baby-sitting to allow the mother to work. The costs to the households of family members attending traditionally run educational institutions can thus be prohibitive and delivery systems must be adapted to reduce them to acceptable levels—for example by providing instruction and health care in the evening near the places of residence. Similarly, vocational schools often produce skills that are not in demand in the local economy or teach the use of tools not generally available. This argues for involving the business community in deciding on how and what to teach. Finally, health and nutrition education are important complements to physical delivery facilities. For example, the health effects of clean drinking water are lost if the water is later contaminated by unsanitary practices. Such counseling should include family planning advice, because many poverty studies highlight the importance of family size as a cause of urban destitution.

Land, infrastructure, and public services. The growth of employment and income-earning opportunities depends on the availability of inputs other than labor. If access to land, work space, and basic infrastructure is insufficient or these are available in uneconomic locations, entrepreneurs will be unable to expand businesses or create new ones. The same applies to essential inputs like electricity, water, waste disposal, and transport. If these are in short supply or are unreliable, the cost of doing business will be higher than it need be, and both employment and the productivity of labor will be depressed.

For a variety of reasons, the public sector in developing countries has assumed a much wider role in the provision of land, infrastructure, and public services than is common in developed countries. An equally wide range of reasons explains why it is not performing well in this role. A detailed examination of what is wrong and why would lead us into general questions of urban management that go well beyond the focus of this article. It is clear, however, that inadequate provision of these inputs and services is widespread and holds back employment and productivity growth. This affects the smaller enterprises more severely than large ones, because they are much less able to compensate for such deficiencies on their own. The small and the very small enterprises are also the ones that create most new jobs, even in developed countries. This makes remedial measures all the more important.

To be successful, such remedial measures will have to be tailored specifically to the needs of individual cities and towns; generalizations are not very helpful. From a large number of diagnoses in specific urban areas, it appears, however, that the following should be part of any program to improve the situation (in addition to more efficient management, cost recovery, and pricing of public services):

  • Operation of public entities with a clear commercial purpose or promoting private initiatives in land development, factory and workshop construction, public markets, and so on;

  • Systematic consultation with the business sector on issues such as investment planning for infrastructure and services, the location of commercial land, etc; and

  • Helping businesses adapt to service deficiencies.

Direct business support and regulatory reform. Flaws in markets for capital and foreign exchange also hold back demand for labor. Capital and foreign exchange are often rationed and allocated by criteria that discriminate against medium- and small-sized businesses, even though these businesses could hire more people and produce more output with a given amount of capital or foreign exchange. Similar discrimination arises from overly burdensome loan approval, disbursement, and collection procedures of financial institutions and from the increasingly untenable perception that the small entrepreneur cannot service loans on commercial conditions.

Paradoxically, special support programs to small business often add to capital market flaws because of the way they are designed. Many of them provide subsidized credit, which keeps them small in relation to need and encourages their clients to move prematurely into capital-intensive activities and technologies. Also, they often do not provide working capital, because they are slow to recognize that working capital is often more important than investment finance for the regular operation and growth of small enterprises. Similarly, they frequently fail to recognize the importance of construction, trade, and services in urban employment, income, and growth, and they restrict their lending to manufacturing activities. This excludes financial support from businesses that typically employ two thirds to three quarters of the labor force active in microenterprises.

The regulatory environment also constrains the growth and development of urban businesses, particularly the small ones. It is often characterized by regulations that are without substantive merit or unenforceable. Many rules are also ill-adapted to urban socioeconomic conditions and discriminate against smaller businesses. For example, zoning regulations may prohibit business activity in low-income residential areas, but very small enterprises cannot afford work space away from home. They will thus be forced to operate illegally or not at all, and both income and employment growth will suffer.

A variety of actions can be undertaken to mitigate the effects of these constraints on business development. Microenterprise credit schemes have shown very high rates of return and could be more generally promoted. Considerable experience is now becoming available on what it takes to make such schemes financially viable, such as a motivated financial intermediary, preselection of borrowers by grass-roots voluntary organizations, mutual repayment guarantees by borrowers, quick disbursement of loans, frequent and convenient collection of repayments, and so on. More generally, financial institutions could be helped to resolve biases against small- and medium-sized businesses, and priority could be given to allocating foreign exchange for vital spare parts. Where the need is established, technical and managerial assistance to entrepreneurs could also be promoted.

Special assistance could be provided to help entrepreneurs deal with cumbersome bureaucratic procedures. More important, in cooperation with the business community, the regulatory environment should be reviewed and inappropriate regulations weeded out; regulations and procedures streamlined to minimize costs and maximize gains for the local economy while safeguarding essential public interests; and relevant staff should be trained to understand the socioeconomic implications of their regulatory activities.

Broad approach

Some common directions run through the various actions proposed above. First, their target is the business sector and its labor force as a whole. But many of them are likely to have their greatest value for the small and the very small enterprises, because these are, on balance, the least able to compensate for the imperfections and constraints they face. Second, the ideas presented argue for greater involvement of the business sector in such diverse areas as investment programming and coordination, financing of land development, factory and workshop construction, provision of some public services, the establishment of vocational training curricula and teaching methods, reform of regulations, and so on. Many of these activities are designed to improve the public response to the wide-ranging needs of the business sector. Particular care must be taken to ensure that, wherever appropriate, the voices of the small and very small enterprises are heard.

Third, numerous interactions exist between the actions proposed. For example, relaxing zoning regulations in low-income residential neighborhoods will reduce the need for workshops separate from people’s homes; lowering building standards will reduce needs for credit; locating low-income housing close to workplaces will reduce the need for and cost of public transport; and so on. Because of these interactions, the proposals for action must be seen as part of a program.

This broad program can be implemented by local authorities. Such authorities usually have substantial responsibilities in land, infrastructure, and service delivery and maintenance, and they are best able to exploit complementarities and to plan and coordinate investments in these fields. While they should not, themselves, become financial intermediaries, they can use persuasion, leverage, and funds to encourage lending to small businesses by existing financial institutions. They also have a clear role in the direct support of business at the local level, and in helping in other ways, such as in dealing with the bureaucracy (including their own), and in streamlining the regulatory framework and its enforcement. Many local authorities are also active in education and health services and more are entering these fields. Because responsibility for the various actions is likely to be diffused among a number of institutions and agencies, a city-level Economic Development Office with sufficient authority may need to be created to implement the program. Many cities in developed countries have been active in these areas for some time, yielding useful experience for their counterparts in the developing world.

Many of the actions proposed have been tried and much experience with them has accumulated. They have not, however, been tried in a coherent and cohesive package, organized around the major objective of urban employment and income promotion. This should now be done. It could produce significant progress in meeting the challenge of poverty.

Other Resources Citing This Publication