Share This
September 2007 Cover Art

Search Finance & Development

Advanced Search
About F&D


Back Issues

Write Us

Copyright Information

Free Email Notification

Receive emails when we post new items of interest to you.

Subscribe or Modify your profile

Finance & Development
A quarterly magazine of the IMF
September 2007, Volume 44, Number 3

Urban Poverty
Martin Ravallion

Are poor people gravitating to towns and cities? Yes, but maybe not quickly enough

The developing world is becoming more urban. Some observers see this as the unwelcome precursor to new poverty problems, such as urban slums blossoming in congested cities. Yet others see it as a force for poverty reduction, as economies shift gradually out of agriculture to more remunerative activities, such as better-paid jobs in other sectors.

What does the evidence suggest? Is the weight of the world's enormous poverty problem, historically a predominantly rural phenomenon, shifting to urban areas? Is urbanization good or bad from the point of view of fighting poverty?

The data to help answer these questions have improved greatly over the past 20 years, thanks in large part to the efforts of national statistical agencies throughout the world, often with the support of the donor community and international development agencies. Today, one can find reasonably well-designed and well-executed household surveys for the majority of developing countries. Although many data problems still need to be ironed out, there has been an undeniable advance in our knowledge about poverty in the world.

At the World Bank, we have drawn on these household surveys and other sources to create a data set that aims to shed new light on the extent to which poverty is becoming an urban phenomenon in the developing world and on the role urbanization has played in reducing poverty overall. It draws on more than 200 household surveys for about 90 countries representing more than 90 percent of the population in the developing world (Ravallion, Chen, and Sangraula, 2007).

Our findings show that 75 percent of the developing world's poor still live in rural areas, although there are some marked regional differences. But the share of the poor living in urban areas is rising, and more rapidly than for the population as a whole. We believe that, by facilitating overall economic growth, population urbanization has helped reduce overall poverty—however, the process of urbanization has affected rural poverty more than urban poverty.

Measuring urban and rural poverty

Why is this new data set so important? For the first time, the Bank's international poverty counts (see box) have been broken down into their urban and rural components and account for the higher cost of living typically facing the urban poor. The best available source of data on the cost-of-living differences facing the poor appears to be the World Bank's country-specific poverty assessments, which have been done for more than 100 developing countries. These reports, a core element of the Bank's country-level analytic work, describe the extent of poverty and its causes in each country.

Keeping track of poverty

The World Bank bases its poverty measures for the developing world as a whole and its main regions on two international poverty lines, of about $1 and $2 a day (or, more precisely, $32.74 and $65.48 a month) at 1993 purchasing power parity. The $1 a day line is a deliberately conservative definition of poverty in that it is anchored to the poverty lines typical of low-income countries. Richer countries naturally tend to have higher poverty lines. One could not reasonably argue that there are fewer poor in the world when the count is based on the standards of what poverty means in the poorest countries. The $2 a day line is more typical of middle-income countries. But good practice is to look at both or, indeed, a wider range of lines in any given country.

The calculations aim for consistency across countries: the international poverty line is intended to have the same real value in different countries and over time. The international poverty line is converted to local currencies using the Bank's purchasing power parity exchange rates for consumption in 1993 and then updated over time using the best available consumer price index for each country. The poverty lines, expressed in local currency amounts, are applied to the available nationally representative household survey data on household consumption or income per person to determine how many people in each country fall below the $1 and $2 lines.

The new data set uses the international "$1 a day" line as the rural poverty line and then applies to survey-based distributions of consumption or income the country-specific estimates of the difference between the urban and rural poverty lines found in the Bank's poverty assessments. For example, $1.08 a day at 1993 purchasing power parity represents 1.53 yuan a person in China, which is equivalent to 2.42 yuan at 1999 prices; this is the rural poverty line used for the 1999 household survey data for China. The poverty lines used by the Bank's poverty assessment for China imply an urban-rural differential of 1.37, so the urban poverty line for 1999 is 3.32 yuan a person.

By estimating everything from the primary data, we have achieved greater internal consistency than most other compilations of distributional data. But, of course, some comparability problems cannot be resolved. For example, different countries have different definitions of "urban" and "rural." The methods used to construct poverty lines for urban and rural areas also vary across countries. And the urban-rural cost-of-living differential may vary by income, so the differential reported in the Bank's poverty assessments for middle-income countries may not be the right one for the $1 a day international line used, which is based on notions of poverty prevailing in low-income countries.

The state of urban and rural poverty

These data throw new empirical light on the changing face of poverty in the developing world, giving us four main findings:

The incidence of absolute poverty is appreciably higher in rural areas. The poverty line in urban areas is, on average, about 30 percent higher than in rural areas. In poorer countries, the ratio of the urban line to the rural line tends to be higher than in middle-income countries. This is to be expected because transport infrastructure and internal market integration generally improve as countries climb the income ladder. However, even allowing for the higher cost of living facing the poor in urban areas, the $1 a day rural poverty rate in 2002 of about 30 percent was more than double the urban rate (see table). Similarly, while 70 percent of the rural population lived on less than $2 a day in 2002, the proportion in urban areas was less than half that figure. About 75 percent of the developing world's poor still live in rural areas.

The changing face of poverty

The urban sector's share of the poor is rising over time. Among those living on no more than $1 a day, the proportion found in urban areas rose from 19 percent to 24 percent between 1993 and 2002; over the same period, the urban share of the population as a whole rose from 38 percent to 42 percent. Nonetheless, it will be many decades before a majority of the developing world's poor live in urban areas.

The poor are urbanizing faster than the population as a whole. This reflects a lower-than-average pace of poverty reduction in urban areas. Between 1993 and 2002, 50 million more people joined the ranks of those living on less than $1 a day in urban areas. However, the aggregate number of poor people fell by about 100 million, thanks to a decline of 150 million in the number of rural poor.

There are marked regional differences. The fastest urbanization of poverty occurred in Latin America, where the majority of the poor now live in urban areas (see Chart 1). By contrast, less than 10 percent of East Asia's poor live in urban areas, largely because absolute poverty in China is overwhelmingly rural. There are also regional exceptions to the overall pattern of poverty's urbanization; indeed, there are signs of a ruralization of poverty in China and in Eastern Europe and Central Asia.

Chart 1. Big regional differences

This new evidence implies that an extra 130 million people lived on less than $1 a day in 2000. We also find a somewhat slower pace of overall poverty reduction over time than in past work. These differences stem from the higher cost of living and the slower pace of poverty reduction in urban areas.

Urbanization is a generally positive factor in overall poverty reduction. There is no denying that new urban problems emerge in poor and rapidly urbanizing countries. However, countries' experiences are generally consistent with the view that a rising share of the population living in urban areas plays a positive role in overall poverty reduction—the one regional exception being sub-Saharan Africa. A pooling of countries and dates reveals lower overall (urban plus rural) poverty rates at higher urban population shares (see Chart 2). Furthermore, higher rates of increase in the urban population share tend to be associated with steeper rates of overall poverty reduction.

Chart 2. A force for good

Why is this happening? In principle, urban population growth can help reduce overall poverty in two ways. First, urbanization can come with higher mean income, which reduces poverty even if the distribution of incomes relative to the mean does not change. Second, it might entail improved distribution even if there is little or no economic growth. The evidence suggests that the first factor has been far stronger than the second; in other words, population urbanization and poverty reduction are linked mainly through economic growth.

A number of economic forces appear to underlie the link between urbanization and poverty reduction. On the one hand, urban economic growth often provides new opportunities for those migrating out of rural areas, some of whom escape poverty in the process. On the other, there can be important indirect effects of urbanization on the living standards of those who remain in rural areas. Positive effects stem from the fact that those moving to urban areas often send remittances back to the rural areas and (importantly) that fewer people are left in rural areas to compete for available jobs. There can be negative effects too, such as when large out-migrations lead to long-term deterioration in the physical and human infrastructure of rural communities.

The positive indirect effects, through higher rural living standards, appear to be more important than has generally been thought. Indeed, the evidence suggests that population urbanization has done more to reduce rural "$1 a day" poverty than to reduce urban poverty. Rural poverty measures tend to fall more rapidly in countries with higher rates of population urbanization. Urbanization appears to be having a compositional effect on the urban population, in that the new urban residents tend to be poorer than the previous urban population. Naturally, this slows the pace of urban poverty reduction, even though poverty is falling in rural areas and for the population as a whole.

What does this mean for policy?

Given that the bulk of the developing world's poverty remains in rural areas, policies to promote agricultural and rural development will continue to play a crucial role (World Bank, 2007). But these research findings also point to the potential importance of urban policies for rural poverty reduction. The pace of urbanization and the gains it brings to the poor, including the rural poor, depend in part on the actions of urban governments, such as providing services and legal protection, improving transportation, and strengthening land-use regulations and land titling.

However, urban governments are typically answerable to their urban constituents alone. A city government, on its own, will probably devote too few resources to actions that yield spillover benefits to its rural hinterland. Yet many urban policies (in land-use and housing regulations and service provisioning) are likely to affect the pace of urbanization. The issue then arises as to whether these policies are socially optimal, taking account of urbanization's external gains to rural areas. Indeed, some incumbent urban residents may expect to become worse off from policies that attract rural migrants. It is thus not surprising that past urban policies have often ignored migrants' needs and even burdened them with extra costs (both pecuniary and nonpecuniary).

The bottom line? Yes, the poor are gravitating to towns and cities, but more rapid poverty reduction will probably require a faster pace of urbanization, not a slower one—and development policymakers will need to facilitate this process, not hinder it.


Ravallion, Martin, Shaohua Chen, and Prem Sangraula, 2007, "New Evidence on the Urbanization of Global Poverty," Policy Research Working Paper No. 4199 (Washington: World Bank);

World Bank, 2007, World Development Report 2008: Agriculture for Development (New York: Oxford University Press for the World Bank).

Martin Ravallion is Director, Development Research Group, World Bank.