East Asia Urban Working Paper Series Urban Poverty in East Asia a review of Indonesia, the Philippines, and Vietnam September 2003 Working Paper No. 11 Urban Sector Development Unit East Asia Infrastructure Department The World Bank Acronyms ADB Asian Development Bank BPS Badan Pusat Statistik DHS Demographic and Health Surveys EAP East Asia and Pacific FIES Family Income and Expenditure Survey GDP Gross Domestic Product GRP Gross Regional Product IN Indonesia LSMS Living Standards Measurement Surveys LSS Living Standards Survey PA Poverty Assessment PH Philippines PPA Participatory Poverty Assessment PRA Participatory Rapid Appraisal PRSP Poverty Reduction Strategy Paper NGO Non-governmental Organization SUSENAS National Socioeconomic Survey SWRS Social Weather Report Survey TA Technical Annex UIP Urban Indicators Program UN United Nations UNCHS United Nations Centre for Human Settlements (Habitat) UNDP United Nations Development Programme VLSS Vietnam Living Standards Survey VN Vietnam WDI World Development Indicators WDR World Development Report WUP World Urbanization Prospects Table of Contents Acronyms Foreword Acknowledgments EXECUTIVE SUMMARY i PART ONE: THE CONTEXT OF URBAN POVERTY 1 A. Purpose and Approach of this Review 1 B. Trends in Urban Demographics 1 The Urban Transition 1 The Sizes and Growth of EAP Urban Areas 4 Migration and Urban Growth Rates 6 C. The Urban Economic Context 7 Changing Economic Activities 7 Increased Concentration in Peri-Urban Areas 9 Susceptibility to Macroeconomic Shocks 9 D. Implications of Demographic and Economic Changes for Urban Poverty 11 Associations among Urban Growth, Poverty, and Total Poverty 11 The Relationship between Migration and Poverty 13 EAP Trends in the Incidence of Poverty, and Urban Poverty, in Particular 14 Implications of Urban Development for Reducing Poverty 16 PART TWO: AN ANALYSIS OF QUANTITATIVE AND QUALITATIVE INFORMATION 19 A. A Conceptual Framework for Urban Poverty 19 B. Methodological Issues in Quantitative Assessments of Urban Poverty 21 Limitations of National Household (Living Standards) Surveys 21 C. Demographic Profile of the Urban Poor 23 Age, gender, household size and composition 24 Migration Status 24 D. Labor and Human Capital Assets 25 Education Status 25 Health Status 25 Employment Status and Types of Work 27 E. Physical Assets: Housing, Land and Infrastructure 29 Housing and Land Tenure 29 Electricity and Durable Consumer Assets 31 Water supply 31 Sanitation 33 Transport 35 F. Insecurity and Urban Poverty 36 Insecurity of Home and Place 36 Personal Insecurity 37 Financial Insecurity 37 G. Empowerment, Social Capital and Urban Poverty 38 Relations with Government 38 Social Networks among the Urban Poor 39 PART THREE: IMPLICATIONS FOR POLICY, INSTITUTIONAL DEVELOPMENT, AND RESEARCH 41 A. Summary: Urban Poverty in the EAP Region 41 B. Policy and Institutional Implications of Urban Poverty 43 C. Suggested Priorities for Research into Urban Poverty in the Region 45 REFERENCES 47 ANNEXES 51 LIST OF TABLES Table I.1. Total Urban and Rural Populations, East Asia and Pacific Region 2 Table I.2. Historic and Recent Trends in Urbanization: East Asia and Other Countries Compared 2 Table 1.3. Comparative Structural Indicators of Urbanization, by Country and Subgroups 3 Table I.4. Standard Measures of Income (Expenditure) Poverty in Rural and Urban Areas 14 of the Philippines and Vietnam, 1985, 1991, 1998 Table I.5. Indonesia: Comparisons of Rural and Urban Poverty Incidence Derived 15 with Different Measurement Methods Table I.6. Cumulative distribution function of per capita expenditure, by type of settlement (Indonesia) 15 Table II.1. Health Indicators by Urban-Rural Residence 26 LIST OF FIGURES Figure I.1. Trends in rates of urbanization: Indonesia, Philippines, and Vietnam 4 Figure I.2. Comparison of urban growth in the Philippines, Indonesia, Vietnam 4 Figure I.3. Differences in City size patterns between the East Asian subregions, 5 as compared to other developing countries Figure I.4. Increase in the urban share of the total poverty in East Asian countries, 12 1990-2000 and projections to 2030 Figure II.1. Indonesia ­ Urban Household Demographics: Female-headed Households, 24 Elderly Households and Large Families, by per Capita Expenditure Deciles Figure II.2. Philippines ­ Urban household demographics: Female Headship and Elderly Headship, 24 by per Capita Expenditure Deciles Figure II.3 Vietnam ­ Urban Housing: Types of Temporary structures used, 29 by per capita expenditure deciles Figure II.4 Indonesia ­ Urban housing status: Private ownership, rent, lease and "other," 30 by per capita expenditure decile. Figure II.5. Philippines and Manila ­ Urban Housing Quality, by per Capita Expenditure Deciles 30 Figure II.6. Vietnam ­ Urban sources of drinking water, by per capita expenditure deciles 32 Figure II.7. Philippines ­ Urban access to "own faucet" rises with per capita expenditure status 32 Figure II.8. Indonesia ­ Urban sanitation (final disposal): Septic tank use and unsafe methods, 33 by per capita expenditure decile Figure II.9. Vietnam ­ Urban type of toilet: by per capita expenditure decile 34 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam Foreword The East Asia and Pacific Region is experiencing an urban transition of breathtaking scale. Between 2000 and 2030, the urban population will increase by half a billion people, almost doubling the current level to a total of 1.2 billion, and the growth rate of the urban population is likely to be 130 percent of the Region's total population growth. As the Region is urbanizing, so too, to a considerable extent, is poverty. Urban areas offer greater opportunities for work, lower unit costs of service provision, and access to new ideas and social exchange. The urban transition is therefore a key element in the process of increasing income and welfare for the country as a whole. But countries facing rapid urban growth confront rising demands for housing, land and urban services, which local governments and related institutions are often ill-prepared to meet. Official policies have often denied recognition to urban migrants as legitimate claimants for city services. At the same time, urban residents are particularly vulnerable to macroeconomic shocks that affect demand for their labor and raise prices for essential goods and services. They are also vulnerable to environmental and public health hazards that arise when dense settlements are not well managed. The individuals least able to compete in the face of such constraints are the poor. The present study was motivated by the fact that the nature and dimensions of urban poverty in the East Asia and Pacific Region have not heretofore been subjected to much direct or systematic analysis, either quantitative or qualitative. Discussions of overall poverty in the Region have tended to gloss over its urban manifestations, while urban operations suffer from the lack of a strong grounding in relevant poverty knowledge. Poverty surveys and analyses in the Region have examined urban poverty to a very limited extent and without sufficient detail to permit conclusions about the factors that define and affect it. This review is a first attempt to analyze available quantitative and qualitative information on urban poverty across three large countries in the Region--Indonesia, the Philippines and Vietnam. It aims to draw out observations and issues concerning the substance of urban poverty in these three countries as well as the methodologies of urban poverty analysis. The analysis is based on a review of existing literature and household survey data. It is useful not only for what it reveals but also for what it cannot reveal due to the limited data on urban poverty that currently exists, and the methodologies used which often fail to capture the complex and variegated nature of urban living. It is therefore an important contribution to our knowledge on poverty. As a preliminary analysis, it lays the groundwork for determining priorities for further work ­ for both operations and research ­ on urban poverty in the Region. Christian Delvoie Sector Director, Infrastructure Department East Asia and Pacific Region Acknowledgments The study was Task Managed by Aniruddha Dasgupta, under the overall guidance of Keshav Varma. Michael Lokshin and Vijayendra Rao carried out the analysis of urban data from the living standards surveys for the three focus countries. Christine Kessides drafted the main report and Vijayendra Rao drafted Annex 1. Shareen Joshi contributed to the literature review and background data collection. Laura de Brular and Barbara Gregory helped to prepare the graphics and tables. The document was edited by Dorothy Silvers. Bobbie Brown, Anne Harrison, Vernetta Hitch, and Socorro Manila helped with document production. The cover design and layout was done by InfoKreasi Ltd in Jakarta. Rumana Huque was in charge of supervising the editing and publication of the document. The work has benefited from the comments of Peer Reviewers Judy Baker, Marianne Fay, Jesko Hentschel, and Kinnon Scott; from written comments from Bhuvan Bhatnagar, Edward Dotson, Teresa Ho, Tamar Manuelyan Atinc, Menno Prasad Pradhan, Rob Swinkels and Carolyn Turk; and from views expressed by participants at a World Bank review meeting on October 17, 2001, chaired by Homi Kharas. The Executive Summary was presented to participants including 11 mayors and other officials from Indonesia, the Philippines and Vietnam, as well as from Cambodia, China, and Mongolia at an Urban Poverty Learning Workshop held in Singapore from June 10-11, 2002. Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam EXECUTIVE SUMMARY This study reviews available quantitative and qualitative information on urban poverty Purpose and issues and trends in the East Asia and Pacific (EAP) Region, with particular focus on Approach of the Indonesia, the Philippines, and Vietnam. The effort is motivated by a concern that Study much of the poverty analysis undertaken in the Region does not adequately represent the urban realities there, and by the conviction that a strong grounding in relevant, in- depth knowledge about poverty could improve the effectiveness of urban operations. The aim of this preliminary assessment is to reach a fuller and more accurate understanding of the evolution of urban poverty in the Region and especially in those three large countries, in order to inform both the Bank's research and its operations, and to enhance our dialogues with local and national clients on strategies to reduce poverty. The review is a desk study--that is, it is limited to material accessible to the World Bank in Washington, drawing mainly on reports of field work and other published and unpublished papers. The empirical analysis focuses on the household poverty surveys and Bank-sponsored "poverty assessments", which are the principal data sources used by the Bank and by the national governments in designing poverty-related activities. The study focuses on Indonesia, the Philippines, and Vietnam because each has carried out a recent Bank-assisted exercise in poverty assessment and has had an active dialogue with the Bank on urban strategy and operations. The report identifies certain shortcomings and gaps in conventional poverty measurement and recommends future refinements and priorities. The urban population of the East Asia and Pacific Region will almost double between Part One: 2000 and 2030, from 665 million to 1.2 billion. The annual rate of increase in the last The Context of 25 years (3.75 percent) is exceeded only by that in Sub-Saharan Africa. Over the next Urban Poverty generation, the urban increment alone will account for 130 percent of the total population growth in the Region, because of the absolute decline in rural populations. Trends in Urban Vietnam, Indonesia and the Philippines are at distinctly different points in their urban Demographics transitions. According to the official indicator of urbanization, the Philippines appears in the upper-middle rank, Indonesia in the middle rank, and Vietnam in the low rank among countries of the Region. However, in many countries of the Region the measurement of what is "urban" is distorted by omission of the extensive peri-urban areas that are in fact the fastest growing areas. Urban growth results from a combination of natural increase in the urban population, net migration from other areas of the country, and reclassification of rural areas as urban. Rural-to-urban migration appears to be the major factor explaining urban growth in the least urbanized countries (e.g., Vietnam), but becomes much less important than natural increase and reclassification in more urbanized areas (e.g., in the Philippines). Countries undergoing rapid urban growth, whether due to natural increase or to in- migration, confront rising demands for housing, land, and urban services. The urban residents least able to compete for such constrained supplies are the poor. Their plight has been exacerbated by a distinctive phenomenon in Southeast Asia: migration not i into the city cores, but rather into peri-urban areas (beyond and not necessarily adjoining the city boundaries) that have recently attracted much investment. The exclusion of these peri-residents from local government jurisdictions means that the households lack access to social services and are at greater risk of poverty even in the midst of a relatively strong local economy. The transition countries in the region, notably China and Vietnam, have strictly controlled rights to reside in the major cities (although China no longer enforces the policy strictly). Such measures exacerbate hardships for migrants who may remain illegal "unregistered" residents for years. The absolute size of urban areas in EAP is dramatic. The population of cities in the Region with over one million residents will increase by half, from 330 million to almost 500 million, between 2000 and 2015; similarly, that of "mega cities" (over 10 million) will increase from about 80 to 120 million. In Eastern Asian countries, population is more evenly spread across city-size categories than is typical in the developing world, although with cities of 1-5 million residents slightly dominant. The populations in Southeastern Asia, however, are heavily clustered in cities of under 500,000 residents, and also more concentrated in the mega cities than is typical in other developing countries. The Southeastern Asia subregion is thus especially challenged to manage the very high growth rates in the largest cities. Size of cities is not in itself a development issue, but size together with rapid growth clearly present a serious challenge. The key to residents' welfare lies with the city's capacities, both financial and managerial, to ensure essential services and to prevent or counteract crippling land costs and negative externalities: traffic congestion, pollution and crime. Such problems tend to worsen as cities grow and they burden the poor disproportionately. Unlike higher-income groups, the poor lack the resources to find alternative services and protection, or to negotiate with authorities and service providers. The Urban Economic Simultaneously with the urban transition, the Region is undergoing a structural Context transformation of its economy. These changes can be seen best within a broad spatial perspective encompassing not only the "city proper" but entire city-regions. As economic development changes through shifts in production and employment from largely rural agriculture to largely urban industry and services, the transformation extends into hinterland areas before they are officially classified as urban. To find cheaper land, traditional manufacturing shifts outward from the large cities on to smaller cities. However, the more information- and technology-intensive industry and services remain in the central cities and metro areas (and also extend into some new periurban areas), because their needs for skills, infrastructure and amenities are more important considerations than land costs. For many unskilled workers and for most of the poor, the main sources of income are in the small-scale and "informal sector", where employment is related to infrastructure and other services, construction, trade and small-scale manufacturing, and urban agriculture. Those sources of employment exploit the multiplier effects of the registered "formal" economy. In rural areas too, such diverse economic activities are increasingly important. Non- farm employment there generally benefits from proximity to urban areas (as does agricultural production) because of access to markets, information, and infrastructure. To reap the potential productivity advantages of large urban labor markets, the extent to which the framework of economic policies and the business environment offer incentives for investment and private enterprise is clearly important. The Region's macroeconomic-financial crisis of the late 1990s demonstrated that although urban Executive Summary ii Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam areas generate economic activity, they also are vulnerable to ripple-through effects of cyclical and other macroeconomic shocks--effects that are particularly hard on those living at the margin of poverty. One result of extended periurban development in the Region is that the central cities no longer offer a wide range of formal sector jobs across skill levels and in both manufacturing and services. Manufacturing is moving outside the city limits; consequently "high end," modern services and the informal sector's very low-end production and services are left as the officially recognized urban economy. Thus, the urban poor may have less occupational mobility within the cities than has been available to urban residents in other countries. Policies to ensure a well-integrated internal labor market (with ease of migration and internal mobility) are called for if spatial segmentation of population and of jobs is to be avoided. The peri-urbanization phenomenon also calls for realistic policies of urban management that integrate the outer zones with the cities and provide them with adequate urban services. In both the Philippines and Vietnam, consumption, or expenditures, poverty has The Urbanization of declined over the 1990s for both urban and rural populations, whether measured by Poverty: Incidence headcount or as depth of poverty. Urban poverty headcounts cited in the recent and Problematic poverty assessments (16 percent of the urban population in Indonesia, 12 percent in Elements the Philippines and 9 percent in Vietnam) remain well below those for rural areas. However, measurements of urban poverty are subject to certain important caveats. Correcting for the omission of unregistered migrants, for example, could raise the urban poverty headcount in Vietnam to perhaps 15 percent. For Indonesia, differences in the measurement methodologies have an enormous impact on the ratios of poverty headcounts between rural and urban populations. Moreover, the low but improving indicator of poverty depth there implies that the urban population is close to the poverty line. Therefore, changes in estimations of this line--as well as actual adverse events that cause income shocks--can have a large (whether statistical or real) impact on the risks of individuals falling into poverty. For all three focus countries, economic inequality is higher in urban than in rural areas. In the Philippines and Vietnam, urban inequality is worsening. Intra-urban inequality in those countries is also evident from much of the data on access to basic services and on status across housing, land tenure, water and sanitation, and transport. Possibly reflecting the limitations of official measurements of poverty as well as attesting to real inequalities, the urban population in the Philippines rate their own poverty status well above official estimates, and above self-ratings by rural respondents. As measured by the conventional consumption (expenditure) measures of poverty, the urban poor as a proportion of the total poor populations of developing countries, including those in the EAP Region, are projected to increase significantly. Conservative assumptions would indicate that in 2025, urban poverty would average about 40 percent of total poverty in the overall Region, up from about 25 percent in 1998 (Hentschel and Bump, 1999). The urban share of populations in poverty could range from over half (Indonesia and the Philippines) to less than a fifth in China and Vietnam--although predictions for China, especially, vary greatly according to presumed poverty income thresholds and definitions of "urban" population. It is not a simple task to predict how the trends in urbanization and economic activity in the EAP countries will affect the magnitude and proportions of poverty in urban places. In theory, as the share of the total population that is urban increases, that should reduce the total poverty in a country over the medium term, because urbanization is highly correlated with rising national income, greater market depth, and other development indicators. In those terms, increasing urbanization would be expected iii to mean rising incomes for the urban residents and for in-migrants to the cities. Over time, too, urbanization should benefit the remaining rural population, by relieving pressure on rural land, enlarging markets for rural goods, and building up savings for public and private transfers to rural areas. However, the pace of income growth in urban and in rural areas does not necessarily remove the income inequalities between them. Indeed, the urban-rural gap may widen for a time, especially with respect to rural areas that are intrinsically under-resourced in natural or human capital, if governments do not introduce effective transfer policies and foster well-integrated financial markets. Although it is sometimes thought that rural-urban migration simply shifts poverty from rural to urban areas, in-migrants to cities are not necessarily motivated by poverty nor languishing among the poor there. How government policy and the incumbent population treat the migrants is critical in determining their welfare and how well they integrate into city life. Despite the expectation for rising urbanization to reduce both total and urban poverty over the medium term, the transitional processes--the influx of migrants, as well as other changes that create new demands on cities for services, jobs, housing, and infrastructure can create hardship for many urban residents in the short term. Even with good conventional urban management, cities may confront a more deep-seated phenomenon: poverty that is more a manifestation of fundamental social, political, and institutional divisions, now heightened by the context of urban growth. Such problems are revealed where: · there are deep divisions between social groups (e.g., "legal" and "nonlegal" residents); · certain settlements within the city are spatially segregated from others, leaving residents burdened with many persistent disadvantages, including risk of eviction and social stigma; · many residents have little or no normal political voice, or access to legal redress; or · certain vulnerable groups are insufficiently protected by social networks and other institutions. To change such structural determinants of poverty requires more than simply efforts to improve incomes; fundamental reforms in governance, both local and national, would be necessary. There are thus possibilities that in urban areas, poverty could both be created faster and be resolved faster, according to the determinants involved and how they are managed. The Role of Urban It is of interest to this review to identify factors associated with, or contributing to, the Development in Region's increases or decreases in overall poverty (and of course, in urban poverty in Reducing Rural, particular). Causation of change and its implications for poverty strategies are not Urban and Nationwide Poverty easy to identify. In all three countries, the agricultural population has the highest incidence of poverty and the highest share of the total poor. For the Philippines and Vietnam, poverty assessments conclude that given the continuing preponderance of the poor in agriculture, better performance of this sector is crucial to alleviate poverty impacts, but so is a structural shift of population out of agriculture into activities with higher productivity. The key questions are then: in what locations do high productivity activities (including high-value agriculture) take place, and what are the conditions that best support them? Within all three countries, the regions with the highest incidences of poverty are often remote from population centers, and are those that depend on agriculture yet have Executive Summary iv Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam relatively poor natural resources. By virtue of their locations, such regions have less access to the economies of agglomeration, which are based on proximity to markets for goods and labor, to infrastructure that reduces production costs, and to networks for exchange of information and technology. Even within rural areas of Vietnam, for example, both nonfarm employment and agricultural activities benefit in peripheral areas close to sources of urban demand; their living standards are higher and there is less poverty. The urban characteristic of relatively dense and larger-scale settlement also permits activities with high fixed costs (increasing returns to scale) to be provided more efficiently--activities such as secondary and tertiary education and health services, and network infrastructure. In short, the common observation that more urbanized regions (e.g., the Red River Delta, the South East and the Mekong Delta in Vietnam) have lower incidences of poverty should not be attributed simply to chance or unfair policy advantage. Their advantage may arise from intrinsic elements of the general urban context that enable growth and poverty-reducing activities at lower cost and with higher returns than would be possible in a general rural context. The advantages hold not only for public investment in infrastructure, for example, but also for much private investment, which derives productivity advantages from the physical proximity of other producers, workers, consumers and suppliers. Nothing about the theory of agglomeration economies guarantees that population concentration alone will achieve economic growth or reduce poverty. What is critical is how well producers and workers are able to respond to the opportunities that urban marketplaces offer. The management of urban growth in the EAP countries will therefore matter greatly not only for the poor within the cities, but even more for the contribution urban areas can make to the economy and to the prospects for reducing poverty in the rural areas. Similarly, it will be important for strategies for reducing national and rural poverty to draw on the potential of urban agglomerations to enhance productivity. That is acknowledged quite explicitly in the poverty assessments for Vietnam and the Philippines, and in the latest national plan for China. The national poverty strategies in countries of the Region should incorporate two indispensable elements: i) policies and investments that strengthen the opportunities and abilities of the rural population to shift to activities allowing higher returns, and/or to move to locations offering greater opportunity; and ii) programs to help cities and towns become more effective in providing jobs and services. The present review uses a broad definition of poverty reflecting several dimensions: Part Two: opportunity and capability, security, and empowerment (WDR 2000/01). Three An Analysis of particular aspects of urban life directly affect how poverty is manifested (Moser, Quantitative and Gatehouse and Garcia, 1996): Qualitative · the characteristic reliance on cash income for all necessities (monetization); Information · environmental hazards, stemming in particular from the relative density of urban Methodological habitation; Approach and Issues · social fragmentation or "churning." Although urban social networks can be strong and highly functional, they differ from those in rural areas. The larger turnover and the absence of many of the social and especially, familial support structures common in rural areas can contribute to urban social stress. To incorporate the dynamic and contextual elements of urban poverty, it is also necessary to take into account the vulnerability to risks associated with poverty, and v the corresponding importance of the various assets, both formal and informal--labor, human capital, natural capital, physical productive assets, household relations, and social capital--that determine households' ability to manage risks (Moser, 1998). Evidence of how the poor cope with risks reveals that, although they draw on a wide range of such assets as best they can in response to shocks and crises, often they also cut back expenditures for food and other basic necessities. Urban populations face a high covariance of risks to the household when jobs or other sources of income disappear, because obtaining essential services requires cash; they face risks to the neighborhood community as well, of forced evictions when tenure is insecure. For the study, the databases of the living standards surveys for Indonesia, the Philippines and Vietnam were broken down to extract the responses from the urban and the rural population samples.1 Respondents were classified according to expenditure deciles and in some cases, according to poor and nonpoor groupings and their survey responses were compared accordingly. These (and most) living standards surveys limit the picture they can convey about urban poverty for several methodological reasons: · They may fail to capture fully the mixed sources of many households' livelihoods, drawn from both rural and urban activities and assets regardless of where they live. · The samples are typically too small to permit disaggregation among and within urban areas (i.e. different cities cannot be compared, nor can neighborhoods within cities.) · Given the rapid changes in urban population, the sampling frames may be outdated. · The survey instruments are often insensitive to the particularities of urban life. More discussion of the last two points is found below. Quantitative and Despite the caveats above, some preliminary observations can be made from the survey Qualitative Findings data, supplemented by limited review of qualitative studies in the three focus countries. Reviewed The following analysis identifies both what can and what cannot be known from those sources. Characteristics of the · The data from the living standards survey reveal that the poor and poorest among income-poor and their locations the urban populations of Indonesia, the Philippines and Vietnam are not the usually expected "vulnerable groups" (i.e. not female-headed households or the elderly), but they do include large families.2 Private transfers of urban origin may be a key to the "vulnerable groups" avoiding poverty. · The migrants surveyed are not shown to be less well off than longtime residents. However, a major shortcoming of the Vietnam survey is its failure to capture the nonregistered migrants who suffer official exclusion from services and benefits. · Although in general, indicators improve with the increasing size of settlements, the largest urban areas (Manila and Jakarta) are not necessarily the most favored. For example, with respect to water and sanitation in Jakarta, the poor there are worse off than those in smaller settlements. 1) For the Philippines, the 1997 Family Income and Expenditure Survey (FIES) dataset contains information about 38,000 households in 81 provinces and 16 regions; 47 percent of the total sample are urban households. For Vietnam, the 1998 Household Living Standards Survey (VLSS) was used, containing 6000 households of which 1730 (29 percent) are urban. The Indonesia poverty assessment is based on the SUSENAS 1999 dataset of 205,700 households (approximately 800,000 individuals), 31 percent of which are urban. Each survey provides for comparison of income and non-income characteristics across per capita expenditure deciles and across poor/non- poor groupings; it also breaks down the characteristics within each decile or grouping. Poverty lines are established separately in the surveys by region and by rural and urban zones. 2) The finding that larger households have higher poverty risk was not checked against alternative assumptions about economies of scale in household consumption. Executive Summary vi Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam · The housing status of the urban poor is marked by a high degree of crowding and Housing, physical by tenure insecurity, or the risk of forced eviction. A large quality differential in assets and services housing exists across urban income groups, although in the Philippines housing conditions are surprisingly poor even for the urban middle class. · Although, overall access to education, health facilities, water and sanitation is predictably higher in urban than in rural areas, it would require a breakdown of effective access and quality of service, and of outcomes --across income groups and zones of the city, and among urban localities--to determine accurately the welfare status of different groups in either absolute or relative terms. Such disaggregations are not feasible from most of the survey databases. · Sources of water supply and quality of sanitation are highly divergent across urban income groups. In Jakarta and Manila, the poor are more likely than the rich to pay for water. The large proportion of the urban poor without basic sanitation or safe waste disposal in Vietnam and Jakarta poses major health risks for them and for the entire city populations. · Historically, the high density of East Asian cities has enabled use of nonmotorized transport and walking, but with urban growth, that is changing. The mobility and access of the poor are affected by problems in public transport, increased traffic congestion and accidents, ground level pollution, and transport-related crime, although these factors were not captured by the survey data sets. · The unemployed identified in the surveys are not always also among the urban Observations poor. The survey instruments are not well suited to capture the vagaries of urban Concerning Earning earnings patterns, especially in the informal sector. Status and Capabilities · The information available makes clear the large disparities in educational attainment among the urban population. In the urban setting, moreover, the intermediate levels of education do not necessarily translate into higher incomes. · Health status in urban areas is worsened by behaviors, multiple stresses and environmental risks. In the Philippines, infant and child mortality are higher for the urban poor than for the rural poor. Studies elsewhere of variations in health outcomes have shown them to vary even more across zones of the same city. However, the available living standards surveys for this study's focus countries do not permit such an analysis. The surveys reveal malnutrition and hunger in some urban areas of the Philippines and Vietnam. Although risky behaviors, disaster- prone living conditions, incidence of crime, violence and HIV-AIDS, and traffic accidents are likely to affect mortality and morbidity in urban areas, especially among the poor, those factors are not covered in the surveys reviewed. · The urban poor face a covariance of threats to their personal, financial and communal Characteristics security stemming from uncertain housing tenure, macroeconomic shocks (both to Affecting their earnings and to prices), crime, and other social pathologies such as drug use. Empowerment and Security · Usually, the poor must rely on private financial transfers rather than public transfers to mitigate their financial risks. · Despite their physical proximity to seats of political power, the urban poor report having little influence on policies or programs affecting them, unless t h e y organize. Generally the urban poor perceive themselves to be excluded by government, yet highly vulnerable to individual instances of official corruption. · The urban poor have many complex social networks which serve many functions: of social integration, mutual support, labor market facilitation, and collective action to obtain services and housing. It is this highly diversified social capital, rather than formal relations with government agencies, that helps the urban poor to manage. vii A Key Distinction Some urban poverty can be viewed as resulting from a temporary mismatch between and Its Implications the supply and demand for jobs and for services, especially when rapid in-migration for Policy and occurs. If institutions are both responsive and efficient, they can reduce such gaps by Programs providing more resources for services and by removing specific bottlenecks such as undue regulatory barriers. But in the three focus countries viewed here, and arguably in many others, much of the poverty appears due to deepseated political and institutional factors that shut certain groups out from the opportunities and protections that others can expect from either markets or from the government. The empirical finding of deep and pervasive inequalities in the urban areas suggests that the fundamental roots of poverty reside in the structures of governance. To relieve the other type of poverty--which may be characterized as waiting in a queue that moves by fair, well-known and accepted rules--governments and external donors can accelerate the "queue" by applying more financial resources and by identifying and establishing measures to accelerate supply, such as support for private sector participation; also, by providing information, they can strengthen the expression of demand for services. Such processes depend on government attention to policy reform and improved investment in infrastructure, and to private sector development. Thus, appropriate government efforts remain a priority in the Region, in particular, attention to reducing regulatory or other disincentives to providing services to low- income residents. Another high priority should be encouraging municipalities to change polices that obstruct land development, and to undertake flexible urban planning that can steer such development so as to forestall slums appearing concomitantly with population growth. To get at the deeper problems that disempower the urban poor and keep them from taking advantage of improved opportunities, it is necessary to act more directly on underlying institutional issues. The insecurities of housing and land tenure, livelihood insecurity and physical insecurity that plague low-income urban residents call for new policies and programs that will strengthen residents' legal protections and their rights for housing and land tenure, reduce official corruption and arbitrary acts, and foster the communities' own social capital. Community-driven programs to upgrade slum neighborhoods, which improve a wide range of physical and communal services through the residents' own actions, would reduce many sources of vulnerability. Such programs' success, however, requires local and national governments to support them as part of an evident commitment to better governance that recognizes the residents as full citizens with rights and responsibilities, rather than simply tolerating the programs as isolated actions. Also necessary would be raising the capacity of local governments to function more responsively, transparently and accountably in their basic responsibilities, and perhaps even to establish participatory strategic planning of their funding allocations and other activities. Broad benefits to the urban poor would flow from such a transformation of the relationship between the local government and citizens. The urban transition and the potential economies of agglomeration can raise productivity for both rural and urban residents, but only in the presence of certain basic mechanisms: a well integrated internal market for labor and goods, with ease of movement and good information flows, and low production costs due to shared infrastructure. Reforms of policies and programs to make urban economies function efficiently, and thus raise their returns to private investment, would be of value to the nation as a whole. Such policies and programs would, for example: · welcome internal migration--certainly, remove residence restrictions where they linger, and facilitate urban-rural remittances as a major source of private transfers; Executive Summary viii Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam · favor efficient domestic markets for goods and services and aid them by improving transport and telecommunications; and · correct distortions or missing segments in the financial markets, in particular by enabling credit and banking services for the poor. Greater attention to reforming the housing and land markets and to improving urban transport in the Region is crucial. The widespread inadequacies in housing and transport, which affect many urban residents but especially burden the poor, should be among the most urgent concerns of governments and of the Bank because those failures weaken the very heart of the urban economy, which is a fluid labor market. Finally, the findings on urban poverty and inequality argue for developing more detailed information on impacts and targeting within the urban population. Nationally representative household surveys cannot provide the necessary spatial disaggregation. What are needed are special purpose surveys capable of producing panel data. Similarly, analysis of public expenditures should go beyond aggregate attributions simply to rural or urban beneficiaries, to more accurately identify the distributional reach within each urban population and urban area. The present analysis raises important points about the accustomed methods of empirical Suggested Priorities research on poverty, the living standards surveys, as being ill-suited to the context of for Research on urban poverty, and therefore unlikely to delineate accurately its nature and the relevant Urban Poverty in the distinctions with respect to rural poverty. The most serious problems are likely to EAP Region stem from the fact that the research methodology, which has been developed and applied most extensively in rural settings and reflects rural notions of life, carries a bias inherent in the sampling design and survey instruments that works against an accurate representation of the urban poor. For one thing, because urban populations constantly change so that homes often contain several families, and because unregistered urban in-migrants are not counted, the decennial census-based urban sampling frames quickly become nonrepresentative. They are especially likely to miss those who are transient or without a fixed address. Although techniques to ensure more accurate sampling are known, they are not always applied. Another difficulty is that because the national survey instruments and methods are typically designed for rural households, work, and living conditions, they usually are not adapted adequately to capture the complexities of urban livelihoods and social relationships, or the multi-spatial nature of households rooted in both urban and rural life. It is necessary to combine qualitative and quantitative methods and information sources to overcome both these limitations. Alternative approaches to sampling and to survey design should be tested and applied. Given better methods of obtaining a more complete and accurate picture of urban poverty, a number of specific issues and questions should be explored. The preceding review suggests some highly relevant research topics: · The effects of interspatial mobility (migration and multi-spatial livelihoods) on poverty in both rural and urban areas. This research would look at circular and temporary rural-to-urban migration as well as at longer-term movement, and would examine both the role of private financial transfers and the mechanisms by which rural migrants become integrated into the urban society and economy. · The dynamics of informal employment in urban areas--the quality of livelihoods and the patterns of occupational mobility--in the context of policies and institutions. ix · How social networks among the urban poor help members cope with poverty and its various dimensions (these networks often include links with rural residents). · A disaggregated analysis of health outcomes by zones within some major cities and comparisons among different cities with different sizes, growth rates, and degrees of service provision, to map health-related poverty in more geographic and socio- economic detail than is now known. · Evaluation of how of specific interventions or packages of interventions, such as neighborhood infrastructure improvements and tenure security, affect the well- being of low-income residents. The role of social capital and how it affects or is affected by residents' participation in such interventions would be a further research question. Executive Summary x Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam PART ONE: THE CONTEXT OF URBAN POVERTY 1.1 The present work aims to begin filling gaps in the Bank's understanding of Purpose and urban poverty in East Asia by a desk review, surveying as much as possible of the Approach of this quantitative and qualitative work that is currently available and extracting urban data Review from the poverty surveys, to arrive at implications for policy research and external assistance. Given resource and time constraints, the study is focused on three countries-- Indonesia, the Philippines and Vietnam--because each has carried out a recent Bank- assisted exercise in poverty assessment, and has had an active dialogue with the Bank on urban strategy and operations. The review looks at the three countries and some selected evidence on others to illustrate urban poverty developments within East Asia and the Pacific, but does not generalize to the entire Region. 1.2 The present study is therefore a preliminary effort to extend our knowledge on the extent and nature of urban poverty in the three countries, the apparent contributing or associated factors, and the possible approaches with which country policies and donor assistance might tackle poverty problems. Where it can help to clarify the distinctive nature of urban poverty, the study compares and contrasts the findings on poverty in urban areas with those for rural areas. Such comparisons can illuminate similarities and differences in policy or operational assistance for different spatial contexts, and identify potential synergies in urban and rural efforts to reduce poverty. By design, therefore, the study raises more questions and hypotheses than conclusive answers, thereby suggesting directions for future work. 1.3 The remainder of this section briefly outlines the demographic and economic context of urban development and the overall pattern of urban and total poverty in East Asia and the Pacific (readers familiar with the Regional background may want to skip this section). Section II begins with a conceptual framework to view urban poverty, and some methodological issues. It then reviews the empirical evidence from surveys in the three focus countries, supplemented where possible by qualitative studies and illustrations from other countries in the Region, to gain a preliminary picture of urban poverty in its several dimensions. Section III recapitulates the main elements of urban poverty that have emerged from this evidence and outlines some priorities for policy and program responses and for future research. 1.4 The urban population of the East Asia and Pacific will almost double between Trends in Urban 2000 and 2030, from 665 million to 1.2 billion, according to the UN (Table I.1). The Demographics urban increase alone will account for more than the total net increase in national populations, because of the absolute decline in the population counted as rural. The Urban 1.5 A comparison of the East Asia and Pacific Regions' rate of urbanization with Transition that of the other developing countries is presented in Table I.2. 1.6 Table I.2 shows that the East Asia and Pacific Region has urbanized more rapidly over the past 25 years than have the developing countries overall; it has also far exceeded 1 Table I.1. Total Urban and Rural Populations, the pace of the currently developed East Asia and Pacific Region* countries when they passed through a similar stage of their urbanization in EAP Region 2000 (million) 2030 (million) Increase (million) %increase the 1900-25 period. The developing Total Population 1877.0 2301.9 424.9 22.6 Urban Population 664.8 1230.1 565.3 85.0 countries have urbanized with much Rural Population 1212.5 1071.8 (-140.7) (-11.6) higher overall population growth than Urban/Total 35.4% 53.4% the industrial countries had. This has * The East Asia and Pacific (EAP) Region combines Southeastern Asia and been particularly true in EAP, which Eastern Asia, less Japan. It excludes Melanesia, Micronesia and Polynesia. has had 3.75 percent annual growth Source: UN, World Urbanization Prospects (WUP), 1999 Revision. of the urban population over the past 25 years, a pace exceeded only by Sub- Saharan Africa. That rate is projected to decline sharply over the next period along with falling population growth overall, but the urban increment will account for a historically unprecedented 130 percent of total population growth over the next generation (more than making up for rural population declines). Table I.2. Historic and Recent Trends in Urbanization: East Asia and Other Countries Compared Developed Countries All Developing Countries East Asia and Pacific Region Year 1900-1925 1975-2000 2000-2025 1975-2000 2000-2025 Share urban (%), 26-40 27-40 40-53 20-35 35-51 beginning-end of period Urban pop growth 90 140 82 151 75 over 25-year period (%) Growth Rate (% p.a.) Urban 2.57 3.56 2.43 3.75 2.25 Rural 0.05 1.12 0.20 0.60 -0.32 Total 0.87 1.92 1.23 1.45 0.74 Contribution of urban 96 62 92 71 130 to total pop growth (%) Country data developed from Brockerhoff and Brennon, 1998; Other data UN, WUP 1999 Revision. All averages weighted by population. 1.7 From the viewpoint of this study, it is noteworthy that Vietnam, Indonesia and the Philippines are at distinctly different points in their urban transition (defined here as urbanization, the shift in population share from rural to urban areas). Table I.3 presents the countries of the East Asia Region (excluding the small Pacific islands) grouped by level of urbanization. 1.8 Table I.3 shows the Philippines at upper-middle rank, Indonesia in the middle rank, and Vietnam in the low rank of urbanization. For the most part, these urbanization rankings correlate with levels of economic development according to the well-known log-linear relationship, although there are clear anomalies in the Region: Mongolia is highly urbanized for its per capita income and economic structure, and Thailand appears very much less so.3 Indonesia is still at the steep slope of its urbanization curve, with relatively high annual increase in the rate of change of urbanization, while Vietnam is just heading into this phase, and the Philippines' pace of urbanization is starting to taper off (Figure I.1). All three countries have urban growth (i.e. annual increase of the urban population) of around 3 percent per annum, with Vietnam's poised to accelerate in the coming decade (Figure I.2). 3) Mongolia is similar in this respect to many other countries in transition from Soviet-style socialism. Thailand followed deliberate decentralization policies through its national plans for many years. Vietnam's early urbanization trend was influenced by war up to the early 1970s, then was curbed sharply by policies of rural resettlement into "new economic zones" through the early 1980s and by urban residency controls. PART ONE The Context of Urban Poverty 2 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam .. tion, oupsr 5.8 1.3 4.0 8.0 4.0 9.2 6.6 6.5 8.1 4.4 6.9 3.3 6.5 3.7 4.7 Services 1990-99 Subg Growth gf .. .. .. 7.1 2.9 9.7 7.6 6.7 6.7 8.2 6.3 6.3 2.7 Urbaniza verageA Output** M 13.9 12.6 10.2 of and 1990-99 s y % of Annual .. tor 2.1 3.1 1.4 0.2 2.3 4.3 4.9 4.6 4.9 2.5 2.1 4.4 3.3 2.0 2.5 Agric Countr 1990-99 Indica yb al 15 39 52 43 37 33 13 25 04 50 53 42 41 .. 54 44 Services 1999 uctur Str Added, GDP** gf .. 7 6 8 .. of M 1999 23 21 32 25 38 17 81 32 33 25 19 evtia alueV % 5 32 18 11 19 18 06 53 52 10 51 03 14 .. form. Agric 1999 10 26 digital in set Compar Rate a** Ann. .. data 1.3. -7.5 1.9 -2.1 -8.0 6.4 1.4 4.3 -8.6 -2.3 0.0 -2.6 0.9 -1.3 1.8 National capit lebaT vg.A 1997-98 -18.0 Growth .1999/1), Income* per Gross $ 1999 093 006 087 .. 092 073 062 018 024 8,490 1,050 3,390 2,010 1,010 26,440 1,980 Secretariat, City Nations (POP/DB/WUP/Rev Rate: %sa Pop. 26 .. 25 10 13 3 33 .. 03 56 51 .. 10 17 16 18 Primary Population Largest National 2000 United Areas ni Urban the of Rural fairsfA and 2001. Growth Rate % Urbanfo Pop. 2.67 2.28 3.14 2.83 3.57 2.34 2.86 4.87 2.86 2.19 4.16 4.01 1.90 3.17 0.50 2.70 4.49 Urban 2000-05 Social 1: and and Part 2000 2030 47.9 2.9 84.1 23.7 179.9 752.1 28.0 4.5 38.0 29.0 5.5 2.6 933.0 397.4 527.2 1,009.8 3,879.6 Urban# Pop., (WDI) millions Economic Revision. 2000 38.4 1.7 44.5 12.8 86.8 410.0 12.6 1.3 15.7 13.3 1.8 0.8 of 571.7 192.7 903.0 167.4 1,942.1 1999 Indicators The 2030 90.5 76.0 73.8 72.7 63.5 50.3 46.6 42.6 33.7 39.1 31.9 33.0 54.6 55.9 83.5 56.2 44.5 Department Urban the % of Prospects: Development 2000 81.9 63.5 58.6 57.4 40.9 32.1 27.7 23.5 19.7 21.6 15.9 17.4 38.5 37.2 76.0 39.9 26.0 Division orldW Urbanization Bank, Korea Guinea Asia Level of veragesA Region Medium New Asia veragesA Population orldW orldW EAP Developed Developed Developed Urbanization Current of High Republic Medium Mongolia Philippines Malaysia Lower Indonesia China Myanmar Low Laos ietnamV Thailand Cambodia Papua Regional Eastern Southeastern otalT Group More Less Least Source: 3 Figure I.1. Trends in rates of urbanization: Figure I.2. Comparison of urban growth Indonesia, Philippines, and Vietnam in the Philippines, Indonesia, Vietnam Trend in Urban Population Shares Trend in Growth Rate of Urban Population 6 80 Philippines 60 Indonesia (%) 4 (%) 40 Vietnam Vietnam Percentage Percentage 2 Indonesia 20 Philippines 0 1950 1970 1990 2010 2030 1950-55 1965-70 1980-55 1995-00 2010-15 Urban population as a percentage of total population Average annual rate of change of the urban population Source: Population Division of the Department of Economic and SocialAffairs Source: Population Division of the Department of Economic and SocialAffairs of the United Nations Secretariat, World Urbanization Prospects: of the United Nations Secretariat, World Urbanization Prospects: The The 1999 Revision. Part 1: Urban and Rural Areas (POP/DB/WUP/ 1999 Revision. Part 1: Urban and Rural Areas (POP/DB/WUP/ Rev.1999/1), data set in digital form. World Bank, World Rev.1999/1), data set in digital form. Development Indicators (WDI) 2000 and 2001. 1.9 Characterizations of "urban" populations of course depend on the definitions used and thus can differ across countries and even in the same country over time.4 For example, China reset its urban threshold, causing a discontinuous jump in registered urban growth for 1985-90.5 Vietnam counts an area as urban if a settlement has at least 4,000 residents and at least 60 percent of the population is engaged in nonagricultural activities--a rather conservative definition that gives added weight to rural areas.6 In many countries, especially in East Asia, a major factor obscuring the measurement of urbanization is the failure to count peri-urban areas as urban (see discussion further below). UN data record Thailand's population as only 22 percent urban in 2000, while official government statistics report that proportion as 30 percent; the Thai planning ministry (NESDB) acknowledges that a more accurate figure, taking account of peri-urban settlement, would be about 40 percent.7 1.10 Households often depend on very diverse economic livelihoods, combining agriculture, manufacturing, commerce and other services, either in their residential location (urban or rural), through seasonal or other temporal migration, and/or through an extended family "portfolio" of economic activities in which different family members engage across different locations. Hence, to fully understand the economic circumstances and prospects of the poor, whether counted as urban or as rural, it is necessary to have a disaggregated picture of their livelihoods and their interactions outside their main geographic areas. The Sizes and 1.11 As countries go through the urban transition inevitable with development, the Growth of EAP rate of growth of the urban population has particular relevance to urban poverty. Urban Areas Urban growth results from a combination of natural increase in the urban population, net migration from other areas of the country, and reclassification of rural areas as 4) Most countries follow a UN or other standard statistical convention to define residents as urban (or rural) if more (or less) than a threshold level live in a single agglomeration. Besides population concentration, urban definition may take account of criteria such as sectoral shares of employment, "contiguous built-up area", and administrative designation. 5) In China, a place needs about 50,000 residents to qualify as urban, compared to 10,000 in the Canadian or U.S. definition. The Chinese urbanization level would be well over 50 percent if the latter threshold were used. 6) Campbell (2001), p. 19. 7) Communication with Dr. Douglas Webster (Stanford University), July 2001. PART ONE The Context of Urban Poverty 4 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam urban. Countries undergoing rapid urban growth, whether due to natural increase or to in-migration, confront rising demands for housing, land, and urban services. The urban residents least able to compete for such constrained supplies are the poor. 1.12 The absolute size of EAP urban areas is dramatic. The population of cities in the Region with over one million residents will increase by half, from 330 million to almost 500 million, between 2000 and 2015; similarly, that of "mega cities" (over 10 million) will increase from about 80 to 120 million. Whereas in 1950 only three of the world's 30 largest cities were in the developing countries of East Asia, and they were all in China (averaging 3.9 million residents each), by 2015 there will be seven there: Bangkok, Jakarta, Metro Manila, and Seoul along with the original Beijing, Shanghai and Tianjin (averaging 17.1 million residents each) (UN, 2000). 1.13 But size of cities is not in itself a development issue--big is not per se bad, and there is evidence that large urban areas, when well managed, are more spatially efficient than smaller ones (Prud'homme 1994). The key to residents' welfare lies with the city's capacity, both financial and managerial, to ensure essential services and to prevent or counteract crippling land costs and negative externalities: such as traffic congestion, pollution and crime. Such problems are often associated with large cities and they burden the poor disproportionately. In most developing as well as in developed countries, about half of the urban population resides in cities of less than half a million residents, and another quarter in cities of the 1-5 million range (Figure I.3). Growth rates are quite dispersed across the size ranges but tend to be highest in cities of over 1 million. However, the city size distributions, for both Eastern Asia and Southeastern Asia, are quite distinct from these international tendencies, although in different ways. The Eastern Asian countries show population more spread across the size categories, but with cities of 1-5 million slightly dominant. Southeastern Asia, by contrast, has population heavily clustered in the below-500,000 FigureFigure I.3. encesences insize patternstternseen the East Asian I.3. Differ Differ in City City size pa betw between the East category; but it also has subregions, as comparedastocompardeveloping countries Asian subregions, other ed to other developing countries relatively more in the 10- million-plus range than is Growth & Shares of Urban Populations in City-size Categories, Four Country Groupings the case in other countries. 2000 The Southeastern Asia 2015 subregion thus faces a special challenge: managing very high growth that is continuing in the largest cities. 70 60 Southeastern 1.14 Research on a large 50 Asia sample of cities of different 40 Population Eastern Asia size categories across all 30 Urban Least Developed developing regions has of 20 Countries % revealed that the combined 10 Less Developed Countries effects of city size and high 0 Fewer than 500,000 - 1 - 5 5 - 10 10 500,000 1 million million million million + growth rates raise infant City Size Categories (Number of Residents) mortality rates, which are an indicator of residents' basic Source: Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat, World Urbanization Prospects: The 1999 Revision. Part 2: Urban Agglomerations (POP/DB/WUP/Rev.1999/2/F16), data set welfare. At city growth rates in digital form. According to the UN WUP, Eastern Asia comprises China, Hong Kong, DPR Korea, Japan, Macau, slightly over 3 percent per Mongolia and the Republic of Korea. SoutheasternAsia comprises Brunei Darussalam, Cambodia, annum, effects (odds ratios) East Timor, Indonesia, Laos, Malaysia, Myanmar, Philippines, Singapore, Thailand and Vietnam. on infant mortality are no 5 worse for agglomerations above 1.5 million than for cities in the 750,000-1.5 million size range. However, the negative effects increase considerably with higher growth rates for each city size range.8 Where such urban growth is occurring, policies must manage its challenges by improving local government functions so that they protect and promote the welfare of the residents, especially the poor. 1.15 Another important structural urban characteristic is the concentration of the urban population in one dominant ("primate") city, or a very few cities. The EAP Region has an average rate of urban primacy lower than that in other developing countries, but also a very wide range, from China at 3 percent to Thailand at 56 percent (Table I.3). Recent international research has shown that urban concentration is related to economic efficiency (Henderson, 1999). The evidence suggests a "best" degree of national urban concentration that increases sharply as income rises up to a threshold for per capita income (about $5000 in purchasing power parity), then declines modestly. The study also shows that the "best" degree of urban concentration declines with country scale. Of the East Asian countries studied, Korea and Thailand appear in the Henderson analysis to have "excessive" urban concentration (at least according to their official urban statistics), Malaysia's is below the estimated optimum for its income level, and China's is about right, given its size. It is also likely from this analysis that the Philippines is nearing an excessive concentration in Metro Manila. The study author concludes that the economic costs of concentration rates that exceed or fall far short of the estimated best level are considerable, in terms of forgone economic growth. The factors affecting primacy are complex: history, openness to trade, degree of fiscal (de)centralization, and extent of interregional transport infrastructure. The research finds that greater road density significantly reduces urban concentration, an effect that rises with national income. What is of interest for the present review is the confirmation that although the process of urbanization (agglomeration of economic activities, population and markets) promotes economic growth, imbalances in a country's urban development have high costs. Ensuring that cities and towns throughout the country are well managed and are integrated in subregions should be a deliberate element of strategies to accomplish poverty reduction as well as national growth. Migration and 1.16 The significance of internal migration for urban growth is difficult to assess Urban Growth accurately because in census data, migration is not distinguishable from the Rates reclassification of formerly rural areas as urban. However, it can safely be said that for countries at relatively low levels of urbanization, rural-to-urban migration is the dominant factor in cities' growth, but that at higher levels of urbanization and of income (up to a threshold of about US$4,000 (1985 prices)), internal natural increase in the cities is the main source of growth (Lucas, 1999). For example, rural-to-urban migration is seen as explaining the bulk of the urban population growth in Vietnam (Lim et al., 2000), which is in an early stage of urban transition.9 In an analysis of 26 developing countries in the 1980s, net migration plus reclassification explained 40 percent of urban population growth; however, in a broader sample of 46 developing countries over 1960-1990, the average contribution of migration and reclassification appeared highest in East and Southeast Asia.10 8) Brockerhoff and Brennan (1998), p. 1-40. Conclusions apply to a sample including East Asian cities, although results for the latter are not separately identified. 9) A 1985 cross-country study found the contribution to urban population growth of net migration plus reclassification to be about 61 percent for Thailand, 55 percent for Indonesia, and 40 percent for the Philippines--consistent with their relative levels of urbanization. (Ogawa, 1985, cited in Shareen Joshi, 2001.) 10) Ibid. In China, the number of towns more than quadrupled during 1982-90 because of reclassification. In Indonesia, net rural-urban migration is estimated to account for 25-30 percent of urban population growth, with 30-35 percent attributed to reclassification of settlements and the remaining 40-45 percent due to natural increase (Wegelin, 2001, p. 4). PART ONE The Context of Urban Poverty 6 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam 1.17 Internal migration tends to have diverse patterns (including rural-rural and urban- urban) as development proceeds. Even some urban-rural migration is observed, at least temporarily, in times of macroeconomic crisis such as East Asia experienced in 1998-99. A distinctive phenomenon in Southeast Asia in recent years has been migration not to the city cores, but rather to peri-urban areas (zones beyond and not necessarily adjacent to city boundaries), as they have attracted much new investment. As already noted, the increased population of a peri-urban area is not necessarily counted as urban in the censuses. Furthermore, the fact that the residents fall outside local government jurisdictions means that they lack access to social services and are at greater risk of poverty even in the midst of a relatively strong local economy. 1.18 The transition countries in the region, notably China and Vietnam, have pursued policies in the past of strictly controlling rights to reside in the major cities. In Vietnam, as discussed below, such measures restricting access to employment and services have exacerbated hardships for poor migrants, who remain illegal "unregistered" residents for many years. In China, the household registration system, created in the early 1950s to limit residency and employment in cities, is now being relaxed in some provinces, and some flexibility in enforcement appears in parts of Vietnam. 1.19 Considerable flows of international migration, as well, go to the peri-urban and urban areas of East Asia. Burmese workers migrate to Thailand, and Indonesian and Philippino workers to Malaysia and Singapore. These migrants, an increasing share of whom are women,11 often suffer the lowest legal and social status, and the worst living conditions (Webster, 2001). 1.20 As has been documented in many countries, urban areas, especially large cities, The Urban contribute more than their population share to the national economy. For example, Economic Context the three major cities of Vietnam (Ho Chi Minh City, Hanoi and Haiphong), with a combined official population of 12 percent of the national total in 1990, produced 19 Changing percent of the gross domestic product; by 1995, with their population share largely Economic Activities unchanged, these cities generated nearly 30 percent of the national output. With their surrounding provinces, the Mekong and Red River Deltas included, the combined city-regions produce more than 70 percent of the national output with half the national population.12 1.21 Taking the broader regions surrounding urban centers into account is important, because though economic development proceeds through shifts in production and employment from largely rural agriculture to largely urban industry and services, those shifts go on, as well, in hinterland areas before they are officially classified as urban. To find cheaper land, traditional manufacturing shifts outwards from the large cities or to smaller cities. However, the more information- and technology-intensive industry and services remain in the central cities and metro areas, or new peri-urban areas, because their needs for skilled labor, infrastructure and amenities are more important considerations than land costs. For many unskilled workers and for most of the poor, 11) Females such as domestic servants in Malaysia (mainly from Indonesia) now outnumber males among official overseas migrant workers, although males still dominate the larger population of illegal and undocumented workers (Hugo, 2000, cited in Joshi, May 2001). 12) Campbell, op. cit., p. 21. Recently the Southeast Region alongside Ho Chi Minh City has demonstrated dynamic enterprise development and structural economic transformation more dramatic than that of the Mekong Delta; compared to the Mekong Delta, the Southeast accounts for a much larger share of private sector employment (especially manufacturing). (Comments by Carolyn Turk and Rob Swinkels, World Bank Hanoi office) 7 the small scale and "informal sector", with informal infrastructure and other services, construction, trade and small scale manufacturing, and urban agriculture, is the main source of income. That sector exploits the multiplier effects of the registered "formal" economy. 1.22 Similarly diverse economic activities are important in rural areas. For example, among villages of Thai Binh province in Vietnam (Red River delta), 52-64 percent of income is earned in services and small businesses.13 Such "nonfarm employment" generally benefits from proximity to urban areas (as does agricultural production) because of access to markets, information, and infrastructure.14 In addition, many studies show that migration of workers into both formal and informal urban activities is more common from regions relatively near the migrant's destination. In short, migration behavior weakens with distance,15 which further suggests that an urban economy's domestic influence should be considered beyond the official city boundaries. 1.23 What are the most vibrantly growing economic activities in the focus countries, and how do these relate to the urban demographics? In Vietnam, growth in output and employment was much stronger in industry, and especially in services, than in agriculture during 1993-98, as would be expected in its early stage of structural transformation. Ironically, employment growth in "nonfarm" activities was higher in rural areas of Vietnam than in urban ones.16 The incomplete state of reforms (doi moi) may have meant that private sector commerce and manufacturing in urban areas remained relatively more constrained.17 Vietnam's rate of growth for wage employment (considered in the Vietnam Poverty Assessment as synonymous with "formal sector") was only half of what Indonesia achieved in the first half of the 1990s after it embarked on similar reforms.18 Of course, wage jobs and reported unemployment affect a relatively small share of the labor force even in urban areas of Vietnam. There is a strong duality in the labor force (i.e. little employment moves from the informal to the formal economy), possibly more than is found in the Region's more fully market-oriented countries, because in Vietnam formal private enterprises are relatively underdeveloped. Job growth has been most dramatic in the labor-intensive light manufacturing, largely export-oriented, in the main cities especially for young female workers, many of whom are migrants.19 An example is shoe production in and around Haiphong and Ho Chi Minh City. 1.24 In the Philippines there has been little additional shift in employment from agriculture to industry and services since the mid-1980s. Although the labor productivity (output per worker) of industry is about five times that of agriculture and about twice that for services, labor productivity has deteriorated in industry and stagnated in the other sectors since 1984, with a decline in capital intensity.20 Therefore, though there is considerable scope for such sectoral shifts in employment to gain more income, it may not happen without better economic management. 13) Survey of 45,000 families from 1999 agricultural publication, cited in Ibid., p. 19. 14) See Chapter 5 of 2003 World Development Report: Sustainable Development in a Dynamic Economy. (The World Bank and Oxford University Press, 2002). 15) Lucas, op. cit., updated at presentation at World Bank, March 2000. 16) Vietnam: Attacking Poverty, tables 3.1 and 3.2. 17) However, the differential between urban and rural incomes continues to increase because the urban-to-rural productivity differential is high and growing (from a ratio of 2.4 in 1986 to 4.2 in 1998), due to the rising ratio in nonagricultural-to-agricultural productivity (from 4.4 to 7.3), and the ratio of informal sector productivity to primary sector productivity (from 4.0 to 4.5) over the same period. Jean-Marie Cour, 2001. 18) NGO Poverty Working Group, 2000, p. 47. (Hereafter referred to as "Vietnam Poverty Assessment"). 19) Campbell, op. cit. 20) Philippines Poverty Assessment, Vol. II, Chapter 2, Fig. 2.13. PART ONE The Context of Urban Poverty 8 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam 1.25 In East Asia, industrial relocation and new investment are most apparent in the Increased "extended urban regions" (peri-urban areas) outside major cities. This spatial trend will Concentration in account for 40 percent of urban population growth in the Region over the next 20-25 Peri-Urban Areas years--specifically, 53 percent of that in the extended Bangkok region, 70 percent in Jakarta's extended urban region, and about 40-60 percent in major Chinese cities.21 Peri- urban areas can be defined by their underlying process characteristics: that is, they demonstrate an ongoing shift from an agricultural to a manufacturing-dominated economy, a corresponding change in employment structure, rapid population growth and urbanization, and changing spatial development with rising land costs. These peri-urban areas are becoming home to most large manufacturing investment and to foreign direct investment, locating in industrial estates with large perimeter structures and infrastructure networks along with access to a major city for higher level services. Their workers are not commuters but residents, often migrants from other urban areas and including both highly qualified and less skilled workers as well as migrants from rural areas, all drawn by both the core formal sector jobs and by the spin-off demand for other services (see Box I.1). 1.26 A result of this pattern of peri-urban development is that in the cities proper, a wide range of formal sector jobs across skill levels appear in services, but not in manufacturing, since the manufacturing is increasingly moving outside the city limits. Thus, "high end" modern services and very low-end, informal production and services are remaining as the officially recognized urban economy. An implication is that the urban poor may have less occupational mobility within the cities than has been available to urban residents in other countries at other times (e.g. in the U.S or Europe during the 20th century). To avoid spatial segmentation of population and of jobs, therefore, policies to ensure a well-integrated internal labor market, at the minimum, are essential. Workers should be able to migrate within the country without administrative restrictions and should have physical mobility (which depends further on land tenure, housing, and urban and inter-urban transport). The peri-urbanization phenomenon also calls for realistic urban management policies that integrate the outer zones with the cities and provide adequate urban services for them. 1.27 In East Asia, the urban economies and their residents have been particularly Susceptibility to vulnerable to macroeconomic shocks such as the financial crisis of 1997-98. That impact Macroeconomic hit urban areas through price increases for imported goods and food products (the latter Shocks due also to coinciding effects of El Nino); job losses initially in the high-end services such as finance and construction; and consequent reductions in demand for other urban outputs and services such as transportation, hotels and restaurants, entertainment, domestic help, etc. Thus, losses in real earnings affected the poorer workers more as second-order effects, while higher income earners experienced the initial job cuts more directly. 1.28 In Indonesia, the financial crisis was estimated to affect the urban economy and especially the largest cities more than it did the rural sector, with urban-based GDP declining in 1998 by 18 percent, versus 14 percent for the entire economy.22 During 1997-1998, urban households reduced real spending on food by 28 percent, as compared to 8 percent reduction for rural households.23 While all regions of Indonesia experienced an increase in the incidence of poverty between February 1996 and February 1999, the relative increase was much higher in urban areas than in rural areas: the urban poverty rate rose by 126 percent, almost double the increase in rural areas.24 Similarly, the urban 21) Webster, op cit., p. 1. 22) Wegelin, op. cit, p. 8. 23) Ibid., p. 108. 24) Pradhan et al., 2000. 9 Box I.1 Why East Asia shows little relationship between urbanization and industrialization Urbanization and Industrialization by Country Source: Population Division of 60 the Department of Economic and ALG 50 CHN Social Affairs of the United PNG MLY KOR. Rep. Nations Secretariat, World 40 THL IND added Urbanization Prospects: The 1999 VNM MRC MGL JPN SGP PNG GER alueV, 1999), 30 BFO Revision. Part 1: Urban and Rural IDA CDI PHL TUN BZL LBN LAO PNG UK GHN Areas (POP/DB/WUP/Rev.1999/ GDP 20 NTH of CAM HKG 1), data set in digital form. World Industry (% 10 MYR Bank, World Development 0 Indicators (WDI) 2001. 0 10 20 30 40 50 60 70 90 East Asia Pacific Countries % Urban Population Non-EAP Countries (2000) There is a well-known linear relationship between urbanization levels and per capita income, which represents the robust correlation between economic development and increasing shares of urban population both across countries and over time. Since the process of economic development involves a shift of output and employment from agriculture towards industry and then later services, there is a less straightforward relationship between urbanization and industrialization. Rich countries of Europe, and even Japan, have shed industry, so their industrialization levels are declining even though they are already highly urbanized. But it would be expected that countries in the low-to-middle range of urbanization have correspondingly rising levels of industrialization. In the East Asian Region, however, industrialization appears to rise very sharply at low levels of urbanization as measured in official statistics. As the figure shows, China, Thailand, Indonesia and Malaysia are now the foundries of the world, the Manchesters of the 21st century. In every one of these countries (though perhaps less pronounced in Indonesia), industry is not locating in cities but in peri- urban areas. Yet, in many cases, e.g., Thailand and China, these places are not being defined officially as urban. The other factor skewing the urbanization-industrialization relationship across the East Asian region is the widespread practice of urban de-industrialization. For example, virtually every Chinese city (right down to the cities of 800,000 and smaller such as Kunming) has a policy, which has been very successfully implemented in almost all cases, to deindustrialize the city (even without the policy, as seen in Southeast Asia and other market economies, the market would have done the same thing in China over time). Bangkok and Kuala Lumpur, and dozens of other Southeast Asian cities have similar policies, which have been implemented successfully, that are reinforcing market forces. The result is that officially non-urban areas such as Thailand's Eastern Seaboard (the industrial heartland of Southeast Asia) and Chinese peri-urban zones, especially in the Lower Yangtze and Lower Pearl River regions, have regional economies with manufacturing GRP shares of 60-70 percent (the highest in the world), while cities such as the Bangkok Metropolitan Area have manufacturing shares of under 30 percent and falling. Meanwhile, the share of services in cities (measured in terms of GRP) such as Bangkok is over 50 percent and rising. In summary, urbanization in East Asia is increasingly highly correlated with service activities (the more developed the country, the more higher-level services predominate), and increasingly inversely correlated with industrial activities which market forces and national and local policies are moving out of cities. Of course, the above has enormous implications in terms of poverty. Many of the jobs accessible to rural-urban migrants and those with modest education (high school or less) are found outside, but within 200 kms of, the city limits. The cities themselves are often less effective mechanisms for providing such individuals a foothold on the economic ladder as they offer mainly either very high end jobs or very low (often informal sector) jobs (the city proper of Shanghai epitomizes this duality). The cities of the Region are increasingly lacking the equivalent of the solid blue collar jobs that helped so many urban and migrant households in North America to get ahead. Source: Dr. Douglas Webster, Stanford University and the Asia-Pacific Research Center PART ONE The Context of Urban Poverty 10 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam indices of poverty headcount, poverty gap, and poverty severity in Indonesia increased by 152 percent, 184 percent, and 202 percent, respectively, between 1996-99, well over twice the rural increases.25 1.29 In the Philippines, the labor market shock of job or earnings loss was found to have had a more negative impact on relatively commercially-developed communities (in effect, those with more urban characteristics). The Philippines poverty assessment (PA) did not analyze survey data on the effects of the crisis separately for urban and for rural households, but concluded that "the labor market shock affected the relatively better-off wage earners more severely, while the impact of the drought from El Nino was heavier on the relatively poorer agriculture-based households".26 However, since many urban poor are not earning wages but are subject to second-order reductions in the demand for their labor, and since they are vulnerable to inflation due to both the drought and macro/financial instability, it is highly likely that the urban poor were harshly affected. 1.30 The macroeconomic-financial crisis in 1997/98 sharply worsened income poverty in other countries of the Region as well, notably in Korea and Thailand. In Indonesia, poverty has since declined, although not recovering entirely to pre-crisis levels by 2000.27 The Region's economic crisis demonstrated that urban areas, though generators of economic activity, are also subject to ripple-through effects of cyclical and other macroeconomic shocks. The urban economies as a whole can sometimes spring back fairly rapidly from shocks but the setback may be more serious and lasting for residents living at the margin of poverty. 1.31 How the trends in urbanization and economic activity in the EAP countries will Implications of affect the magnitudes and proportions of urban poverty is not simple to predict. A Demographic higher growth in the share of the total population that is urban would be expected to and Economic associate over time with reduction in a country's total poverty, since urbanization is Changes for highly correlated with increasing national income and other development indicators. Urban Poverty That correlation has several explanations. · Labor productivity is in general higher for secondary and tertiary activities, which Associations predominate in the urbanized economy, than for primary production. among Urban Growth, Poverty, · The agglomeration economies associated with the concentration of population and and Total Poverty economic activity in urban areas permit more efficient use of labor, land and capital-- i.e. with higher returns. · The resulting urban economic growth generates revenues that can be used for equalizing transfers. 1.32 Increasing urbanization can thus increase incomes for the urban residents; for migrants from other regions to the cities; and for the rural areas themselves, by relieving pressure on rural land and providing savings for public and private transfers. However, the pace of income growth in urban and rural areas does not necessarily remove the income inequalities between them. Indeed, the urban-rural gap may widen in the medium term, especially for rural areas that are intrinsically under-resourced in natural or human capital, or if governments do not carry out effective transfer policies. 25) Suryahadi, Sumarto, Suharso and Pritchett (2000), Table 8. 26) Philippines PA, Vol. II, Chapter 5, p. 99. 27) WDR 2000/01, p. 26. 11 1.33 Although, over time, increasing urbanization should reduce both total and urban poverty, the processes of the transition--an influx of migrants and other factors creating new demands on cities for services, jobs, housing, infrastructure, etc.--can increase hardships for some residents in the short- to medium-term. Urban growth itself creates a challenge to manage what could be considered "frictional" poverty--that is, it poses the transitional task of settling in new arrivals or integrating residents at the expanding borders of the urban area. That category of poverty can be corrected as fast as those demands can be met, but that depends on the size of the backlog and the pace of new growth. Since the numbers are so large of both backlog and new residents, cities and countries need to find new strategies, adapting flexible approaches such as allocating service delivery to public-private partnerships; liberalizing conditions for private enterprise and business development; facilitating the private sector's development of housing, land and infrastructure; and encouraging private enterprise in other ways as well, as the best way to deal with these growing demands. 1.34 Even with good conventional urban management, however, cities may face a more deep-seated phenomenon of "structural" poverty: a manifestation of social, political, and institutional disparities that are exacerbated and made newly apparent under the pressures of high urban growth. Such problems, evident even in rich cities, emerge where there are deep divisions between social groups, such as that between "legal" and "nonlegal" residents; where certain settlements within the city are spatially segregated, leaving those residents to deal with persistent disadvantages including the risk of eviction; where many people lack access to normal political expression, and voice and legal redress are out of reach or at best weak; and where vulnerable groups are insufficiently protected by social networks and other institutions. These structural determinants of poverty require fundamental reforms in governance, both locally and nationally, that go beyond simply adding financial resources. 1.35 It follows that both the ways in which poverty is generated in urban areas and the ways of alleviating it depend on the nature of the poverty and how institutions respond to it. Definitions of urban poverty and their detailed application to the focus countries are discussed below. Figure I.4. Increase in the urban share of the 1.36 In terms of the conventional income or total poverty in East Asian countries, consumption (expenditure) measures of poverty, and 1990-2000 and projections to 2030 with due caveats about estimations and projections, it is fair to say that in developing countries, including Urban Share of Poverty (%) 0.6 those of the EAP Region, the share of urban within Indonesia the total poor population will increase significantly 0.5 with their rising urbanization. Ravaillion (2000) shows Philippines that, under reasonable theoretical assumptions and 0.4 Cambodia after testing with data on 39 developing countries, 0.3 the urban share of the total number of poor is seen to Lao PDR be a strictly increasing and convex function of the 0.2 China urban share of the total population. That is, as a Vietnam 0.1 country urbanizes, urban poverty rises relative to the national mean, and the poor urbanize faster than the 0 1990 1995 2000 2005 2010 2015 2020 2025 2030 population as a whole. The cross-country data suggest that the urban share of the total poor population would Source:Population projections from UN, WUP, 1999 Revision. Urban and rural poverty reach 50 percent by 2035, when the urban share of headcount rates are for the latest year available from World Bank, WDI 2000. Projection assumes that the ratio of urban to rural headcount rates remains the population reaches 60 percent. This simple unchanged from the base year. analysis implies that as countries urbanize (at least up to some threshold), the share of the urban poor will increase both within the urban population and within the population at large. PART ONE The Context of Urban Poverty 12 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam 1.37 Hentschel and Bump (1999), starting from UN urbanization projections and current estimates of the incidence of rural and urban poverty for each Region, project the share of total poor living in urban areas by 2025 according to three possible assumptions: that the ratio of urban to rural poverty headcounts remains the same as in 1998 (base case), or that the ratio increases (deemed least likely), or that it decreases. The range thus obtained for the entire East Asia Region indicates that the urban share of total poverty would reach 35 to 50 percent in 2025, with a midpoint of 40 percent, in contrast to about 25 percent in 1998. A similar projection for six East Asian countries, taking just the base case (no change in the current ratio of urban and rural poverty headcounts, shows that by 2030 urban areas would account for more than half of the total poor in Indonesia and the Philippines, for 35 percent in Cambodia, but still for less than a fifth in China and Vietnam (Figure I.4). Thus there is ample evidence for the present analysis that rising income poverty in urban areas is a secular trend that merits serious attention from policy makers and other stakeholders, including the Bank. Moreover, beyond the income measurement of poverty, the vulnerability of much of the urban population, who live near the poverty line, in both material and nonmaterial terms, is a further cause for concern. That is discussed in Part Two. 1.38 Despite the common belief that migrants to cities remain at the margin of the The Relationship urban labor force for a long time, that generalization is not valid for EAP, or elsewhere. between Observations from many developing countries have revealed that migrants, whether Migration and from other urban areas as is often the case, or from rural areas, initially earn less than Poverty comparable natives of the city and may make their start in the informal sector, but that the earnings gap tends to close within a few years and can even reverse. Migrants are typically well informed through informal networks about the labor market at their destination, although the networks by their nature are not available to all potential candidates for migration; limited information and lack of facilitating social connections undoubtedly constrain some would-be migrants.28 1.39 Individuals move to urban areas both for economic hopes of better living conditions and earnings and for other reasons (marriage, schooling, to rejoin family). In Indonesia, on average across all regions only 40 percent of migrants from rural to urban areas report seeking employment as their main reason.29 A survey of migrants to Ho Chi Minh City in Vietnam found that slightly over half of males and only a third of females reported moving for economic reasons; those respondents tended to come from the poorer provinces. The migrants who reported non-economic motivations moved from the nearby Mekong Delta and Southeast regions and had strong family networks. Other research on migrants in Vietnam, however, has found that employment and economic factors influence interprovincial migration at least indirectly, since the volume of net migration is larger to the more urbanized and industrialized areas.30 1.40 How government policy and the receiving population treat the migrants is critical to their welfare and their ability to integrate with city life. The Vietnam Poverty Assessment finds that "while most of the migrants to urban areas fare well and add to the prosperity the plight of a particular group of poor migrants was highlighted. These are migrants to urban areas who have not secured permanent registration. This group of the urban poor face difficulties in accessing public services and may also be socially marginalized" (p. vii). Although controls have recently been relaxed, it still can take a migrant a decade to move from "category 4" (illegal, without residency papers) to "category 2 or 1" (with 28) Lucas, op. cit.; de Haan (2000). 29) Wegelin, Table A.7. from SUPAS census, 1995. 30) Dang et al., 1997 and Anh et al, 1996, cited in Phan, 2001. 13 full urban citizenship).31 Even if they acquire a decent livelihood, migrants can have very poor living standards , to the extent that they settle in peripheral slum zones with no basic services or secure housing tenure, whether because labor-intensive manufacturing is relocating there to flee high land costs in the city or because the central areas lack affordable housing. EAP Trends in the 1.41 In the 1987-98 period, the East Asia and Pacific Region had by far the most Incidence of dramatic reduction in the numbers and percentages of poor people, as measured by Poverty, and the international income poverty threshold of $1/day, among all regions of the Urban Poverty, in developing world. Only the Middle-East and the North Africa Region also showed Particular declines. The share of the population living on less than $1/day fell from 26.6 percent in 1987 to 15.3 percent in 1998 (excluding China, from 23.9 to 11.3 percent)--a drop of fully one-third, or 139 million persons (of Table I.4. Standard Measures of Income whom 90 million were in China). Even relative (Expenditure) Poverty in Rural and Urban Areas of the Philippines and Vietnam, 1985, 1991, 1997 income poverty (share of population living on less than one-third of average national THE PHILIPPINES 1985 1991 1997 consumption) fell in the EAP Region over the Total Population period, from 33 percent to 19.6 percent Poverty Incidence (%) 40.9 34.3 25.1 (excluding China, from 45 to 25 percent).32 Depth (%) 13.2 10.6 6.4 Inequality (Gini) 0.412 0.428 0.427 1.42 For most of the Region countries, poverty Rural reduction accompanied economic growth, Poverty Incidence (%) 53.1 48.6 36.4 though with different degrees of elasticity in Depth (%) 17.8 15.6 9.8 responsiveness of poverty to changes in income. Inequality (Gini) .352 .359 .352 From about the mid-1980s to mid-1990s, China Urban led the Region in economic growth but with Poverty Incidence (%) 21.7 20.1 11.9 relatively little reduction in poverty, because Depth (%) 5.9 5.7 2.6 inequality grew. Relative to other countries in Inequality (Gini) 0.410 0.421 0.425 the region, poverty reduction in the Philippines VIETNAM 1993 1998 was also somewhat low for its economic growth. In Vietnam poverty could have been reduced Total Population more if the benefits of economic growth had Poverty Incidence (%) 58.0 37.0 Depth (%) 18.50 9.50 been spread more evenly between urban and Inequality (Theil Index) 0.177 0.201 rural areas. Thailand demonstrated particularly Gini 0.33 0.35 high elasticity in poverty reduction, followed Rural by Indonesia and Malaysia.33 Such differences Poverty Incidence (%) 66.0 45.0 reflect the degree to which each country's Depth (%) 21.50 11.60 growth was "pro-poor." However, changes in Inequality (Theil Index) 0.128 0.126 inequality are not strongly correlated with Urban overall growth in EAP or elsewhere. Poverty Incidence (%) 25.0 9.0 Depth (%) 6.40 1.70 1.43 In both the Philippines and Vietnam, Inequality (Theil Index) 0.187 0.197 income, or expenditures, poverty34 declined Source: The Philippines Poverty Assessment, Tables 4 and 6, Volume I, Main over the 1990s for both urban and rural Report; Vietnam Attacking Poverty, Figures 1, 2 and 1.3, pages 12-13 populations, whether measured as basic Note: "Depth" of poverty refers to distance below the respective poverty line. For headcount or as depth of poverty (Table I.4).35 both Gini and Theil indices, increasing value indicates greater inequality 31) J-M. Cour, 2001. 32) WDR 2000/2001, Tables 1.1 and 1.2. 33) WDR 2000/01, Figure 3.4. Largely similar comparisons emerge from the Philippines PA, Volume 1, Figure 1, and from Vietnam PA, Table 6.3. 34) Termed the "current consumption expenditures deficit," the most common quantified definition of poverty. Pradhan, op. cit. Sometimes referred to simply as "income poverty" or "expenditure or comsumption poverty" these terms are used interchangeably here. 35) These poverty rates cited in the Bank's PA for the Philippines are substantially below the official estimates (e.g., 36.8 percent in 1997, compared to 25.1 percent in the Table above), due to higher official poverty lines. (Philippines PA, Chapter 1). PART ONE The Context of Urban Poverty 14 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam 1.44 Urban poverty continues to be well below Table I.5. Indonesia: Comparisons of Rural that of rural. However, measurements of urban and Urban Poverty Incidence Derived with poverty are subject to important caveats, as Different Measurement Methods discussed further in Part Two. The Vietnam Poverty Incidence, February 1999 sampling frame is believed to have excluded Indonesia Iterative Method BPS Method those migrants without permanent residency % % permits, and hence left out those who are likely Total Population 27.1 23.6 the poorest. It is estimated that correcting for Rural 34.1 25.9 this oversight could raise the urban poverty Urban 16.3 20.0 headcount from 9 to possibly 15 percent.36 For Urban Head Count as % Rural 47.8 77.2 Indonesia, differences in the measurements used Source: Pradhan, et al. (2000), Table 2. See paper for explanation of methods. The according to assumptions about consumption Indonesia Poverty Assessment (World Bank, September 2000 draft) uses the baskets and poverty lines have an enormous "iterative" method estimates. BPS method is that of Badan Pusat Statistik, Jakarta. impact on the ratios of the poverty headcount between rural and urban populations (Table I.5). 1.45 These statistical and survey issues underscore the point that differences in measured income or expenditure poverty between rural and urban areas "are possibly as much an artifact of method and assumptions as they are a finding of `fact'--the poverty line is higher because it is assumed to be higher",37 and should not be interpreted too dogmatically. Then, too, the low and improving "depth" indicator of poverty, which identifies the distance below the poverty line, though favorable, means that the population is close to the poverty line, so changes in estimations about this line--as well as actual adverse events harming incomes--can have large (whether statistical or real) consequences for the numbers of individuals categorized as poor. 1.46 For China, for example, many analysts question using the $1/day international standard of absolute poverty as irrelevant to urban poverty because of the much higher living costs in cities. By that measure, only 1.7 percent of the urban population (5 million people) was poor in 1998. Raising the threshold to $1.50/day yields an urban poverty headcount rate of 10.1 percent, more than a five-fold increase (31 million people); and $2.00/day implies a 24.4 percent headcount rate, or 75 million poor people.38 1.47 Inequality is shown to be higher in urban than in rural areas for all three countries. For example, Gini coefficients were 0.266 for the rural population and 0.328 for the urban in Indonesia in 1999 (see Table I.6), and 1.4 for the Philippines and Vietnam. 1.48 In Indonesia, inequality declined by about the same degree for urban and for rural households between the pre-crisis (1996) and post-crisis (1999) periods, but still left urban Ginis consistently higher than rural.39 In the Philippines and Vietnam, urban inequality as shown in Table I.4 is worsening, a pattern found in many other developing countries as well. Analysis in Vietnam indicates that the national increase in inequality is due Table I.6. Cumulative distribution function of per capita expenditure, by type of settlement (Indonesia) primarily to a growing rural- urban gap in incomes, Urban ­ by Size of Settlement Gini Rural <250K 250-499K 500-999K 1M-2M >2M All Urban outweighing diverging trends in inequality for the Total Population 0.2662 0.3014 0.3257 0.3448 0.3168 0.3367 0.3282 rural and the urban settings. Lowest quintile 0.1366 0.1304 0.1294 0.1576 0.1622 0.1688 0.1453 Source: SUSENAS 1999 household survey. 36) Vietnam PA, Box 1.4. The annual urban household surveys in China also exclude the "floating population" (undocumented urban residents) who are counted instead among the rural population. 37) Pradhan et al., op. cit., p. 16. 38) The ADB has used both $2 and $3/day as poverty lines for measuring urban poverty in China according to Fan et al. (2001). 39) Suryahadi et al, March 2000, Table 9. 15 Implications of 1.49 An important question for this review is what factors are associated with, or Urban contribute to, increases or decreases in overall poverty, and especially in urban poverty. Development for The poverty assessments for the three focus countries analyze the regional and sectoral Reducing Poverty breakdown of poverty trends, providing insights into the urban and rural phenomena; however, none of those reports explores urban poverty directly, nor the links between rural and urban livelihoods. Causality of change and implications for poverty strategies are not easy to identify. 1.50 Certain observations are worth making. In all three countries, the agricultural population has the highest incidence of poverty and the highest share of the total poor. Only Vietnam reduced poverty relatively more in the agricultural than in other sectors. That achievement reflected higher agricultural productivity, which relieves poverty significantly because a large share of the population depends on agriculture, and also weak employment growth in the industrial sector.40 In the Philippines, from 1985 to 1997, poverty reduction was much faster in the nonagricultural activities of industry and services than in agriculture, and the occupational distribution shifted towards those sectors where poverty was rapidly declining. The shift was not large enough, however, to change the occupational composition of the poor. Both the Philippines and the Vietnam poverty assessments conclude that given the continuing preponderance of the poor in agriculture, better growth there is crucial to relieve poverty, along with a structural shift of population from agriculture to activities with higher productivity. 1.51 The key questions are then: In what locations do the high productivity activities (including high-value agriculture) take place? What conditions best support them? The regional breakdowns of poverty and growth in the focus countries show, first, how sharp and durable the spatial variations are. The Philippines PA, for example, notes that "there is no strong evidence for convergence of poverty levels" among Philippine provinces between 1988 and 1997.41 In all three countries, the regions with highest poverty incidence tend to remain those that are remote from population centers, and especially those that have relatively poor natural resources but depend on agricultural activities.42 The remoteness of the less well-off regions means less access to the economies of agglomeration, which arise from producers' proximity to markets for goods and labor, to infrastructure reducing production costs, and to networks for exchange of information and technology. The fact that the more urbanized regions such as the Red River Delta and the South East in Vietnam have lower poverty incidence probably reflects the benefits of agglomeration economies in terms of higher and more rapidly increasing productivity, creating conditions that reduce poverty. 1.52 Although causality with respect to the functioning of these agglomeration economies and poverty reduction within their areas of reach cannot be empirically demonstrated at least by the present study, there is ample evidence of association. Both nonfarm employment, which as noted in the Vietnam PA is critical to reduce poverty, and high value crop and livestock agriculture require a buoyant source of demand. The PA reports note, for example, that high-value agriculture is more successful closer to urban sources of demand. Market impetus is important to poverty reduction; the liberalization of internal trade is cited in the Indonesia PA as contributing to pro- poor growth there. The Vietnam Interim Poverty Reduction Strategy Paper (I-PRSP) points out that rural living standards have improved most in the peripheries of cities and towns. Similarly, a poverty mapping analysis has shown that rural poverty is less 40) Vietnam PA, Table 3.7 on rural income sources. There has been much less growth of labor-intensive industry in Vietnam than in other more liberalized industrializing countries of the EAP region; the lack of industrial employment growth is attributed to the still small and immature formal private sector (p. 60). 41) Philippines PA, Vol. I, Main Report, para. 25. 42) See Chapter 4 in 2003 WDR. PART ONE The Context of Urban Poverty 16 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam prevalent or less severe in the provinces closest to the major urban centers. Even in the poorest provinces, the urban areas have poverty headcounts lower than those in neighboring rural zones; and the more prosperous provinces, e.g., in the South East, now have a relatively small rural-urban poverty gap, indicating economic spillovers. Education is also most strongly related to income in the two urban regions with major cities (Red River Delta and the South East), suggesting that the returns from education rise with the opportunities for nonfarm employment.43 1.53 Such differences are often assumed to show that urban areas simply enjoy favorable policy and political environments, endowing them with more than their share of tax revenues and public investment, high wage protection, favorable prices, and a stronger overall safety net. While countries have maintained some such policies at times (e.g., the "iron rice bowl" policies of China up to the 1980s), it is surely a leap of faith to assume that the better urban performance in so many circumstances, even in the more urban parts of rural areas, is explained by a deliberate policy carried out effectively. There are elements of the generic urban context that give some growth- promoting and poverty-reducing measures lower costs and higher returns than they would have in the generic rural context. Besides being true of public investment in major infrastructure with its high fixed costs and increasing returns to scale, it is also true of much private investment, for which the physical concentration of workers, consumers and suppliers as well as other producers in spatial proximity provides a notable economic advantage.44 1.54 Nothing about the theory of agglomeration economies however, guarantees that population concentration alone will spur economic growth or reduce poverty. What matters is how well producers and workers can respond to the opportunities in urban marketplaces. Therefore, when unnecessary regulatory constraints impede entrepreneurial action; when excessive or unpredictable taxation raises risks to investment; or when weak governance and inefficient urban management fail to deliver the services demanded, urban growth in itself will not fulfill the promise of economic improvement for any but a few people. In such circumstances, increasing urbanization does less than it could for the country's development. People still come hopefully to the cities, but encounter undue obstacles to improving their incomes and living conditions. Private investment then focuses on the surest quick return and on low-risk activities. Moreover, public resources are then unlikely to be allocated to the areas with the most potential for broad-based growth. 1.55 How well urban growth will be managed in the EAP countries will matter greatly not only for levels of urban poverty, but even more for how much the urban areas contribute to the growth of the overall economy, and hence to the prospects for reducing rural poverty. Correspondingly, strategies to reduce national and rural poverty should draw on the dynamics of economic agglomeration. The Vietnam and the Philippines poverty assessments acknowledge that quite explicitly. The latest national plan for China incorporates the urban transition as a strategy to combat poverty in China's western regions, which have missed out on economic growth. The Vietnam PA presents growth of rural nonfarm employment as well as of urban employment as two elements of an integrated 43) Minot and Baulch, p. 9. 44) An example is industrial promotion policies. Experience in Malaysia, Taiwan, Korea and the Philippines demonstrates that industrial estates (which are often government funded) are most successful when located in major ports, regional centers or near a natural resource base. Estates intended to facilitate rural development in isolated locations lacking market potential have generally not been sustainable. (Jones (1988), cited in Joshi, June 2001). The preferences of investors to cluster in close physical proximity, for both economic reasons and to ease their access to government and to services, is also discussed in Webster, op. cit. 17 approach to regional development. The priority for developing nonfarm production to diversify and raise rural incomes will require more public investment in rural areas. Such investment should be planned in a broad spatial vision, considering where non- farm production can best be connected to upstream and downstream markets. Policies and investments that enable the rural population to shift to higher return activities and/ or to move to locations with better opportunities, and those that increase the effectiveness of cities and towns in providing jobs and services, should all be integral to the national poverty strategies in countries of the Region. PART ONE The Context of Urban Poverty 18 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam PART TWO: AN ANALYSIS OF QUANTITATIVE AND QUALITATIVE INFORMATION 2.1 It is now widely accepted internationally that an understanding of poverty must A Conceptual look beyond traditional definitions based on income, expenditure or consumption. Framework for WDR 2000/01 expressed the consensus that poverty has all the following dimensions: Urban Poverty lack of opportunity (including access to markets and assets, and to the conditions that affect returns on assets); low capabilities, constrained by limited health, nutrition and education; a state of insecurity arising from habitual exposure to risk and income shocks, whether as an individual/household or as part of a larger group; and disempowerment because of social or political barriers. These dimensions are equally relevant to urban and to rural poverty, but the manifestations differ with the spatial context. In the urban setting, markets and productive assets (such as infrastructure) are often more available than in rural areas, but institutional failures can nevertheless hinder effective access by the urban poor. For example, in a large city, lack of public transport can prevent them from traveling from the only residential areas they can afford to job locations; services such as piped water supply, though nominally provided, may be of such low quality and reliability that they have to seek their own supply at higher cost. And though urban residents live closer to government agencies, that does not ensure the urban poor a political voice. In fact, the proximity can subject them to frequent instances of police harassment and the costs of corruption. 2.2 Three particular aspects of urban life directly affect how poverty is manifested:45 · "Commoditization"--cash income is required for food, housing, transportation and other essential goods/services. · Environmental hazard due to negative externalities--the relative density of urban settlement means that behaviors and environmental sanitation have immediate and widespread effects on human health and the amenities of public space. Asian cities have the highest density (inhabitants per hectare) in the world. · Social fragmentation--by their nature, urban areas consist of a mix and churning of population that is particularly intense, even when natural growth or internal migration is relatively slow. The social implications of this characteristic are complex. That is, the constant transformations of the urban population can break down traditional social patterns and networks, but also create and reinforce different ones. Nonetheless, it is likely that the sources of social support in cities are subject to stronger forces of change and stress than are those in rural areas, which partly explains different urban rates of crime, substance abuse, and other social pathologies. 2.3 To better incorporate these contextual characteristics and the dynamic nature of poverty, some urban poverty research has used an "asset vulnerability" framework (Moser, 1998). This concept captures the view of poverty as a state of heightened risk--people 45) Moser et al., 1996 ; Moser, 1998. Commoditization is Moser's term, but " monetization " may be a more accurate way of conveying her point that urban livelihoods depend on cash income and expenditure for essential goods and services (e.g., self-provision of food, water, housing, transport, and so on is less an option than for the rural population). In some countries, this may be only a distinction of degree as compared to rural poverty. 19 move in and out of poverty according to their exposure to hazards and shocks. Their ability to cope and adapt depends on using multiple assets: labor; human capital (qualified by health and education status); productive assets including housing, which in the urban context is often the location of informal employment as well as a source of rental income; household relations, enabling mutual reliance for pooling income and shared consumption; and social capital, demonstrated as reciprocity and trust between households. Other analysts may describe or group the relevant assets somewhat differently, but those are the main components. In the rural context, natural capital would obviously be included. De Soto (2000) emphasizes the importance of informal assets held by the poor, and the benefits to them once these assets can be collateralized and used to generate liquidity. 2.4 A combination of qualitative and quantitative methods of inquiry can assess the poor's effective stock of assets and how they manage them, for example, to cope with macroeconomic crisis. Such analyses are cited below. Their findings show that, faced with economic stresses, the urban poor cut back expenditures, frequently by reducing or substituting their food (often 60 percent or more of total household expenditures46). They also mobilize all possible assets--increasing work by family members including children, using the home as a source of income, and drawing more on extended family support and on community networks (e.g., for informal credit). Qualitative studies of poor households show how their liabilities as well as their assets affect their capacities to cope. Important liabilities include not only debt, but social obligations such as sending money to rural relatives.47 2.5 Focusing on the assets of the poor and their vulnerability to risk highlights aspects peculiar to the urban context and relevant distinctions from the rural context. For example, the rural population faces a high covariance of risks from the climate and from livestock and crop diseases that can undermine an entire farming community. The urban population, too, suffers covariance at the household and community levels, but with different characteristics. Their need for cash for all essentials means that the loss of income threatens their consumption of food, as well as use of education, healthcare and transport. The resulting struggle with numerous dimensions of welfare may lead them into debt. The threat of eviction from illegal housing or "slum" settlements and forced resettlement is a shared risk in low-income communities--risk of an event devastating to their incomes, sense of security and social networks. 2.6 While the "asset vulnerability" perspective lends itself well to qualitative studies of poverty, vulnerability in income terms can be measured quantitatively as "the standard deviation of intertemporal changes in expenditures", thus capturing the notion of variability or transient states of poverty. Applying this measure to expenditures data from Indonesian surveys shows that while headcounts of poverty and of income vulnerability are higher for rural than for urban respondents, the ratio of the "income vulnerable" to the "poor" is substantially higher for the urban population (3.67 vs. 2.05), and similarly higher for individuals occupied in trade and services than for those in agriculture (averaging 3.33 vs. 1.95). The finding indicates that even where the incidence of poverty is not a major problem for an urban population at one point in time, that condition is highly unstable because the expenditures of urban informal sector workers such as petty traders and scavengers are so variable.48 46) The Indonesian living standards survey reveals that among the three lowest expenditure deciles in urban areas, at minimum 60 percent of household expenditure is devoted to food; the shares rise by decreasing size of settlement to 75 percent for the poorest urban decile in towns, which is close to the share for the rural poor. The high shares the rural population devote to food (even 66 percent by the richest decile) may indicate that there are fewer claims on household expenditure in rural areas, while the urban population has more requirements (and options) to spend on housing, transport, education, health care, and other goods/services. See Annex Table IN1. 47) Craig et al., op. cit. 48) Pritchett et al., 2000. Figures cited are from the mini-Susenas; the same pattern was found in a 100-village survey that included small urban areas. PART TWO An Analysis of Quantitative 20 and Qualitative Information Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam 2.7 The inequality, vulnerability and multiple dimensions of poverty enter into individuals' own assessments of whether they are poor. The Philippines has conducted surveys of self-rated poverty that place poverty incidence at more than twice the official estimates for 1997-99 reported in Table I.4. There is also a very high urban-rural differential in these self-ratings, 2 to 5 times higher than the cost of living differential implied by the official or basic-needs poverty lines. Urban respondents judged poverty incidence to be 55 percent, as against 70 percent reported by the rural respondents (compare to the measured headcount noted above at 12 percent urban versus 37 percent rural).49 These self-ratings may reflect higher expectations by urban residents conscious of stark inequalities in welfare. They also suggest, however, that the official poverty measurements fail to recognize the full measure of absolute hardships of urban poverty. 2.8 The next section reviews some of the quantitative and qualitative findings on urban poverty in the focus countries and on the sorts of assets that matter in avoiding or coping with poverty. Since the empirical analysis of the survey databases breaks down respondents by consumption-poverty status shown as per capita expenditure deciles, the discussion does not include a separate quantitative evaluation of the survey samples according to other non-income dimensions of poverty such as education, health or housing status. 2.9 The present report analyzed the poverty survey databases used for the Bank's Methodological poverty assessments in Indonesia, the Philippines and Vietnam, to extract information Issues in about the urban population and the urban poor in particular. For the Philippines, the Quantitative 1997 Family Income and Expenditure Survey (FIES) dataset contains information about Assessments of 38,000 households in 81 provinces and 16 regions; 47 percent of the total sample are Urban Poverty urban households. For Vietnam, the 1998 Household Living Standards Survey (VLSS) was used, containing 6,000 households of which 1,730 (29 percent) are urban. The Indonesia poverty assessment is based on the SUSENAS 1999 dataset of 205,700 households (approximately 800,000 individuals), 31 percent of which are urban. Each survey permits both relative comparison of income and non-income characteristics across per capita expenditure deciles, and absolute measures of the poor versus non- poor population based on separate poverty lines established for the surveys by region and by rural and urban zones. 2.10 Several caveats about these surveys also apply to most other nationwide living Limitations of standards surveys (LSS) of similar type. The caveats imply that the LSS cannot be used National to cover all the dimensions and population breakdowns that would be needed to Household (Living accurately assess urban poverty, and that additional special-purpose surveys may be Standards) required. Surveys 2.11 Spatial attribution. The surveys distinguish between rural and urban settlement according to the countries' own conventions, with the usual problems inherent in defining this boundary. The main practical problem is that households are therefore classified as wholly urban or wholly rural, without accounting for multi-spatial households or mutual dependencies among them. 2.12 Cost of living and consumption differences. An underlying issue for surveys of this type is how well the price adjustments and consumption baskets differentiate between 49) Philippines PA, Chapter 2. 21 the actual cost of living and costs of avoiding poverty among the different settlements, and especially between large cities and smaller urban or rural areas. For example, the urban poverty line is based on cost-of-living estimates in the more urbanized regions, which probably underestimates the costs in major cities. 2.13 Aggregation. Although the databases for the three countries contain a good proportional coverage of urban households, the absolute sizes of the surveys are either too small, or the data are otherwise unable to be broken down, to produce representative samples for different sizes of urban areas, or even in most cases for specific cities. The Philippine survey provides a representative look at Metro Manila, but the data do not permit comparisons with other distinct urban areas. A poverty mapping exercise done with the VLSS and the subsequent 1999 Population and Housing Census provided considerable regional differentiation, but only the combined Hanoi/Ho Chi Minh City on the one hand, and other urban areas as a group can be broken out.50 For the present study, the Indonesia LSS data were subjected to an approximate disaggregation by size of locality.51 Even this very limited analysis reveals that Metro Manila has worse living conditions for the poor than do other regions, and that larger cities in Indonesia have higher levels of inequality (Gini coefficient for lowest quintile rising from 0.130 to 0.169 from small to large settlement size categories, with Jakarta being the most unequal--see Table I.6). 2.14 Locational detail within urban areas also cannot be obtained from any of these surveys as typically designed. That is a serious omission, since analysis elsewhere indicates that living standards can vary greatly across neighborhoods of the same city.52 A participatory poverty assessment carried out in Ho Chi Minh City as part of the national Poverty Assessment confirmed that this largest city in the wealthiest region has pockets of severe poverty. 2.15 These and similar income- or- expenditure-based poverty surveys that are derived from rural research can have two other shortcomings from the urban perspective, those of the sampling design and the survey instrument (questionnaire) itself (see Annex 1). Both the sampling and the survey problems tend to understate the extent of urban poverty and not capture its complexities well. Although design techniques exist to correct for these problems, they are not always applied. 2.16 Outmoded sampling design. Poverty estimates are usually calculated from surveys of large nationally representative samples whose sampling frames are based on decennial censuses. In other words, the samples assume that populations are stable over a ten-year period. In rural areas this is normally not a problem, since rural populations tend to be quite stable. Samples drawn from censuses collected every ten years are unbiased and representative for rural populations. 2.17 The urban situation is quite different, however. Urban populations, particularly the urban poor, have high levels of mobility, poorly defined property rights, and fluid family and social structures. The poorest of the poor tend to be squatters living in temporary settlements that are subject to demolition, with ill-defined addresses and a high probability of changes in residence, and even homelessness. Thus, the poorest 50) Minot and Baulch, op cit., p. 7. The census did not permit distinguishing between "other cities" and towns. A Chow test suggests that Hanoi and Ho Chi Minh City differ significantly from other urban areas. 51) The breakdown by size of locality in Indonesia was calculated using household weights from the survey. For each district were calculated the actual number of the households in the district and then, using the urban/rural variable, the size of the urban locality. The problem with this approach is that it can categorize some households incorrectly if there are many small towns in the region; in that case, such households would be considered as if they were from a larger city. The category for largest city (2 million plus population) refers unequivocally to Jakarta. 52) See references in the section on Health, e.g., in Sao Paulo (Brazil), poverty incidence in some parts of the city is more than 40 times higher than in the richest neighborhood (ranging from 1.7 to 77 percent). In Cali (Colombia), a quarter of the poor are concentrated in one of the 20 communes of the city (Danny Leipziger and Marianne Fay, Presentation to LCR Corporate Day, April 23, 20001). PART TWO An Analysis of Quantitative 2 2 and Qualitative Information Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam and especially new migrants are particularly mobile. The problem is compounded in those countries where residents must be "registered". New migrants there are automatically excluded from the urban census and from the samples based upon them, unless special provision is made to capture them. Furthermore, urban homes often have several families living in them, both because migrants may be sheltered by existing residents and because young households "nest" within extended families for economic reasons. Census-based sampling frames are consequently out of date almost the moment that they are tabulated, and surveys based on them, unless adequately corrected and updated to reflect current residential patterns, are very likely to miss the poorest among the urban poor and to underestimate the extent of urban poverty. 2.18 Inappropriate survey design. A further shortcoming of many living standards surveys concerns the questionnaire instruments, which are typically geared to agricultural activities and village lifestyles, but less well designed to capture the particularities of urban life. The risks and insecurities that the urban poor face, especially regarding property rights for housing and employment, have been noted here. The nature of urban occupations, especially the functioning of the informal economy, often is not captured by national surveys, whose questions on income and occupation tend to be based on agricultural and salaried categories. Urban social relationships can be characterized by unconventional connections peculiar to urban communities, where individuals from diverse ethnic and economic backgrounds live together. Among the poor, networks are often directed at finding income opportunities and are part of social protection strategies. At the same time, the urban population and especially migrants are heavily interconnected with rural families and communities. Survey instruments should be both more sensitive to the complexities of the urban experience and more capable of capturing the rural-urban inter-linkages at the household level. That can be done by including modules on non-farm businesses in the surveys, and by using both quantitative with qualitative methods of inquiry. Combining different techniques such as participatory poverty assessments, focus groups, etc., is especially useful for diagnostic purposes and to test the relevance of a formal survey. 2.19 Having duly noted these qualifications, a preliminary picture from the survey data on the urban poor in the three focus countries is summarized below, supplemented by more textured information from other studies drawing on focus groups and case study interviews. City-specific data from other sources (e.g., UNCHS (Habitat) Urban Indicators Program) are included where available, and some relevant indicators are summarized in Annex 1, Tables A1-A5. The findings are presented according to the main asset categories: labor resources and quality of human capital, housing and infrastructure (physical capital), security, social capital and empowerment. 2.20 The country-specific tables referred to below can be found in the Annex 2 of this Report. 2.21 There is not a strong difference in the age composition of rural and urban Demographic households in the three countries. In both types of localities, larger families tend to be Profile of the poorer and clustered in the lowest expenditure decile. Nor is there a significant spatial Urban Poor difference in these countries in the size structure of families (Figure II.1).53 53) Note that this finding of more poverty among larger families did not correct for different assumptions about possible economies of scale in household consumption. 23 Age, gender, 2.22 For Indonesia, household headship by a single parent is more common among household size the urban poor than it is in rural areas. The elderly are not especially poor in either and composition type of location, nor are female-headed households (Table IN2). 2.23 In the Philippines, female headship is more an urban than a rural phenomenon (18 percent of urban households vs. 12 percent of rural), but female-headed households and the elderly are over-represented among the higher Figure II.1. Indonesia ­ Urban Household Demographics: Female-headed Households, deciles in both areas (Figure II.2 and Tables PH1,2). Elderly Households and Large Families, by per Capita Expenditure Deciles 2.24 On average, urban households in Vietnam are smaller and have fewer children than rural households 25 do, across all deciles. Female-headed households, as well as the elderly, are also more common in urban 20 areas and are not particularly poor (Table VN1), 15 although they do represent a larger share of the urban Percent 10 poor than of the rural poor(Table VN2). Only in 5 Vietnam did the survey include ethnicity. Ethnic groups other than native Vietnamese (Kinh) are almost 0 Poorest 1 2 3 4 5 6 7 8 9 10 Richest nonexistent in the urban population, especially the Per capita expenditure urban poor. Either the other groups do not migrate, or they have not been enumerated in the survey (see Female headed household Elderly Family with 3+ children Table VN3).54 Source: Table IN2 Migration Status 2.25 Migration status is usually captured in living standards surveys through questions on place of origin or length of residency. In the Vietnam survey (the only one asking about that topic), 63.7 percent of all urban respondents said they were born in their current locality, as did 74 percent of the urban poor (Table VN4). Those figures confirm that the long-standing residents surveyed are not Figure II.2. Philippines ­ Urban household necessarily much better off than the migrants (or at demographics: Female Headship and Elderly least, than those with legal status who were Headship, by per Capita Expenditure Deciles enumerated). Many may be worse off, although a true comparison is not possible because unregistered 30 migrants are not covered. This tentative interpretation 25 is consistent with many other migration studies 20 showing that migrants do eventually do as well as or 15 Percent better than the native residents. The unregistered 10 migrants, however, are not officially entitled to social 5 services or formal employment, cannot register the births 0 Poorest 1 2 3 4 5 6 7 8 9 10 Richest of their children or enter them into regular public school, Per capita expenditure and have virtually no legal identity55 (Lim et al, 2000) . Such unregistered migrants are estimated to number 1.5- Female head of household Age of head of household 65+ 2.5 million in Vietnam cities, and between 10-50 percent Source: Table PH1 of them are poor.56 54) The Hoa (Chinese) are the only significant urban ethnic minority in Vietnam. 55) However, there are several categories of registration which confer different levels of access to services; access can also vary by ward and by availability of ward funds. Children of unregistered migrants may attend evening literacy classes if not the regular schools, and relaxed enforcement of restrictions is seen in some cases. 56) Vietnam PA, Box 1.4. PART TWO An Analysis of Quantitative 24 and Qualitative Information Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam 2.26 Access to schooling and educational attainment are usually reported to be higher Labor and Human for urban than for rural populations because in denser settlements, it is easier to provide Capital Assets education, since per capita fixed costs are lower. That holds true for the three country databases. Differentials in education status can also reflect the influx to the urban Education Status areas of the better educated among the rural population. In Indonesia, urban households are on average better educated than rural ones, although urban areas are characterized by sharper differences in educational level between the rich and the poor. The proportion of individuals who did not complete primary school education is almost four times higher among the urban poor than among the urban rich, compared to only a two-fold difference in rural areas (Table IN3). Rates of school participation and literacy are highest in Jakarta and lowest in rural areas, but do not decline uniformly with declining settlement size. Illiteracy is more concentrated among the lowest decile in several of the city size categories than it is in rural areas (Tables IN4,5). 2.27 In Vietnam, the urban poor do have much higher literacy rates than their rural counterparts: 75 percent of individuals in the lowest expenditure decile in urban areas can read, versus 49 percent of their rural counterparts. Although the urban poor are better educated than the rural poor, the differences in education levels among urban income deciles are more dramatic than the rural/urban differences among the poor (Tables VN1,5). In the country overall, the education of the household head is related to per capita expenditure, but not closely. Households whose head has completed primary or lower secondary education do not have much higher expenditures than those whose head has not done so. However, higher levels of education are associated with significantly higher earnings. These findings are true both in the major cities (Hanoi and Ho Chi Minh) and in other urban areas. Families with unskilled heads are clearly the worst off. The education of the spouse is a better predictor of household expenditure in urban than in rural areas, suggesting that perhaps members have more attractive earning options in the cities.57 2.28 The Philippines LSS does not provide education or literacy data. School- related expenditures can be considerable for the urban poor and are often sacrificed as a coping strategy (e.g., 12 percent of the lowest decile in the Philippines reported removing children from school, at least temporarily, during the recent crisis compared to 6 percent for the total population58). Many households, though, strive to protect their children's schooling for longer-term advancement. The economic returns to education would be expected to be higher in urban areas, where the labor market is larger and better able to absorb and reward workers for each level of attainment, especially at the higher levels. However, actual employment experience does not always match the theory because of differences in individuals' own expectations, and because of macro- or microeconomic factors, even in the large urban areas, that impede labor utilization (this review does not cover labor market issues, which can be quite important in explaining poverty). 2.29 Although indicators of access to health facilities are normally better for urban Health Status than for rural populations, as would be expected given the greater ease of servicing denser, more accessible settlements, the actual health outcomes may not show the same differential, because many factors intervene including environmental health and sanitation (discussed further below), and behaviors. Averages for the reported availability of facilities or services for urban or rural aggregates can mask issues of effective access (e.g., lack of transport) and quality, especially for the poor. Studies that have broken down the data on both health care and health outcomes within urban areas (e.g., in slum 57) Minot and Baulch, p. 9. 58) Philippines PA, Table 5.8. 25 Table II.1 Health Indicators by Urban-Rural Residence neighborhoods) have often found Indonesia 1997 results that are no Indicators URBAN Income Quintiles RURAL Income Quintiles more favorable, and Poorest Second Middle Fourth Richest Poorest Second Middle Fourth Richest sometimes much Infant Mortality Rate (46.5) 75.3 3.3 34.9 25.6 78.9 55.5 54.0 42.2 18.5 worse, than those in Under-5 Mortality Rate (102.3) 112.8 456.5 44.7 31.3 109.1 72.8 73.5 56.0 24.6 rural areas.59 Health problems such as Philippines 1998 low nutrition and Indicators URBAN Income Quintiles RURAL Income Quintiles mental illnesses Poorest Second Middle Fourth Richest Poorest Second Middle Fourth Richest related to economic Infant Mortality Rate 49.7 40.1 37.6 24.8 17.7 48.7 38.7 28.4 25.1 (35.5) and life stresses are Under-5 Mortality Rate 70.5 62.9 57.9 33.2 26.9 81.2 59.2 38.8 33.7 (39.8) growing in urban Infant Mortality Rate: Deaths under age 12 months per thousand births areas, among the Under-5 Mortality Rate: Deaths under 5 years per thousand births poor in particular.60 Source: Demographic and Health Surveys. 2.30 The national Figures in parenthesis indicate large sampling errors due to small number of cases. Demographic and Health Surveys (DHS)61 in Indonesia and the Philippines (Table II.1) reveal that the infant mortality rate and the under-5 mortality rate of the second poorest income quintile are significantly higher in urban than in rural areas of Indonesia; the same is true in the Philippines, even at the middle quintile. This disparity in mortality outcomes may reflect higher environmental health risks in cities. The DHS surveys also show surprisingly lower immunization coverage for measles and diptheria/polio/tetanus in urban areas than in rural for most quintiles in the Philippines. A similar situation is found for contraceptive use in Indonesia. The findings indicate that an urban advantage in access to services does not always exist and cannot be taken for granted. 2.31 The Philippines Poverty Assessment also found that while health indicators vary predictably by income status, the rural/urban differences are small, and both infant and child mortality are somewhat worse for urban than for rural populations at the lower quintiles. Access to medically trained personnel is better in urban areas, and immunization rates show a slight advantage for the poorest urban residents, but health outcomes are not commensurate.62 Fees for health care, as for education in the Philippines create a significant strain on urban households. Survey data on the poor in Naga City found that 68 percent of respondents described health and sanitation as a problem for them, greater than the percent citing housing (57 percent).63 Noncommunicable diseases (cardiovascular, cancers, asthma) and traumatic injury are reportedly on the rise among the urban poor, possibly from increased effects of pollution, as well as the pressures of urban survival.64 59) E.g., a spatially disaggregated study of morbidity and mortality across zones of Accra (Ghana) and Sao Paulo (Brazil) found that age adjusted death rates are up to three times higher in the most disadvantaged areas of each city compared to those with the best socio-environmental conditions. The death rate differential applies to both diseases of childhood and those of adulthood, and to multiple categories (infectious, respiratory and circulatory) of illness, as well as (in Sao Paulo) to external causes such as traffic accidents and mortality (Stephens et al.1997). 60) "Demographic and health surveys from eleven developing countries show that the ratio of stunting prevalence between poorer vs. wealthier quintiles is greater within urban than within rural areas, and that intra-urban differences between socio-economic groups are greater than urban/rural differences. Urban poor households have worse nutritional status than rural poor households do, contributing to greater ill-health related to nutrition. Malnourishment, hunger, dietary problems (and ill health) often coexist in urban populations." (UNCHS (Habitat), 2001a, p. 108, citing study from Popkin, 1999.) 61) The DHS urban sample sizes for Vietnam were too small to permit a similar analysis. 62) The Philippines PA, Figure 8. 63) Ateneo Social Science Research Center, 1998, cited in Racelis, 2000. 64) Racelis, 2000. PART TWO An Analysis of Quantitative 26 and Qualitative Information Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam 2.32 Health outcomes, especially for young children, may also reflect difficulties in affording food and undernutrition, which became most apparent during the financial crisis.65 The Philippines Social Weather Report Survey in June 1999 found that 20 percent of the extremely poor in urban areas reported hunger in the last three months, and 11 percent said they felt hunger "always". Across all urban areas except Luzon, 19-40 percent of respondents had gone hungry "often".66 In Vietnam as of 1994, over one million (9 percent) of the total urban population could not meet the basic requirement of 2100 calories daily. About one-fourth of the children who were found to be malnourished live in the major cities such as Ho Chi Minh City and Hanoi.67 2.33 The living standards survey data examined for this report do not permit within- city disaggregation of the urban samples. Health indicators are included only for Vietnam and Indonesia. The Indonesian data show virtually no pattern in 16 categories of health complaints across deciles or settlement sizes. For Vietnam, the urban population overall reported a lower incidence of sickness (38 percent) than did their rural counterparts (43 percent); however, the urban poor were sick more often than the rural poor were (46 versus 42 percent incidence) (Table VN6). The urban population engages in damaging behavior (smoking) at a higher rate than the rural population does, and the poor more than the nonpoor; in fact, the share of poor urban females who smoke (83 percent) rivals that of the men68 (Table VN7). The urban poor spend almost as much on health care as the nonpoor do, even though the poor would be expected to rely more on free public services. Both groups spend proportions of their income on health care than are double that of the rural population (Table VN4). Sexually transmitted diseases, including HIV-AIDS, are increasing rapidly in urban areas, in particular among sex workers in Ho Chi Minh City.69 The higher prevalence of HIV-AIDS in large urban areas as compared with smaller urban and rural areas is apparent in many countries.70 2.34 Employment status as reported in the surveys is not very indicative of the actual Employment situation of the urban population, for reasons discussed earlier--e.g., questions asking Status and Types "have job/no job" are not very relevant to the fluid informal economy. As would be of Work expected, unemployment is reported as higher in urban areas than it is in rural areas. However, employment status alone is neither a good predictor of income nor a useful criterion for targeting poverty in any of the three countries. 2.35 In the Philippines, reported unemployment of the household head is twice as high in urban as in rural areas, although this may imply that urban residents have higher expectations of what constitutes a "job", namely a formal-sector or regular activity. The lower deciles report unemployment much less frequently than do the better-off households, whether urban or rural--presumably because the poor cannot afford to remain inactive (Table PH3). 2.36 However, in Indonesia by contrast, the urban poor report consistently higher rates of "Not working" than the nonpoor do, and the same pattern appears for the question about officially defined unemployment, across all urban size categories. The share of women claiming to not work is almost twice that of men in all settlements, including rural ones--which is curious, considering that female-headed households are not particularly poor, as noted earlier (Table IN6). Since women do not more often 65) International research elsewhere has found that malnutrition in the poorest areas of cities often rivals that found in rural areas. Ruel et al. (1999), pp.1891-1905. 66) Social Weather Report Survey, June 2-16 1999, cited in Racelis, 2000. 67) Lim et al., citing a UNDP report of 1998. 68) Poverty experts in the Bank's Resident Mission in Vietnam dispute the validity of the survey finding of smoking prevalence among women. 69) Lim et al., 2000 70) UNCHS (Habitat), 2001a, p. 107. 27 claim to be officially unemployed, it is possible that women do not consider their types of work to be formally recognized as such (Table IN7). It is also possible that many of the relatively well-off, female-headed households consist of single women migrants, for example, those with jobs in the new manufacturing industries. In Vietnam, female unemployment rates for 1993-98 have remained about two-thirds those of men.71 2.37 The unemployment rate is reported to be slightly higher in Metro Manila than in other urban areas of the Philippines; 21 percent of household heads in Manila claim to be unemployed versus a 19 percent overall urban average (Table PH4). Similarly in Indonesia, the unemployment rate tends to increase, for both poor and nonpoor, with increasing settlement size. In the largest urban category (2 million-plus residents-- essentially Jakarta), 12.6 percent of the poor males and 10 percent of poor females claim to be unemployed (officially), compared to 8.5 and 6.7 percent, respectively, in the urban areas with less than 250,000 population (Table IN7). These results for what are by far the largest cities of these two countries are somewhat surprising, in that demand for all types of workers should be the strongest there; however, the influx of aspiring workers to these metropoles, and their expectations for formal sector employment, also may be greater. 2.38 To get a fuller and perhaps more accurate picture of the urban poor, more information is needed about underemployment (defined as "employment that doesn't pay enough to meet the basic needs, so workers must search for additional sources of income"72), and about work in the informal sector. In the newly developing peri-urban zones of the Philippines, Thailand and other East Asian countries, poverty is associated with the "the working poor" for whom jobs are abundant but do not provide enough earnings to cover the cost of living.73 In both Indonesia and the Philippines, estimates of underemployment from national labor force surveys showed an increase in the numbers and shares of workers, particularly in manufacturing, who were on the job less than full time, during and after the macroeconomic crisis--a trend that may have pushed some marginal earners to seek informal sector work, or even to fall below the poverty line.74 Among poor households, individuals working longer hours are a reflection of underemployment. In times of crisis households mobilize more family members, especially secondary school age children.75 In Vietnam, 33 percent of the urban employed worked over 50 hours/week in 1998, and 17 percent over 60 hours, in contrast to 20 and 9 percent, respectively, of the rural labor force.76 Thus simply comparing urban household earnings to rural can overestimate the hourly pay differential. 2.39 The informal economy, which is said to employ 60-80 percent of the working poor in most developing countries, represents a wide range of activities and degrees of productivity. Earnings from petty trade in urban areas, such as sidewalk sales of prepared foods, in-home piecework, domestic service, or rental of housing space, can fluctuate greatly and tend to be reduced by general economic declines. 2.40 The participants in informal sector employment lack the basic protections of formal jobs, but even worse, they are often vulnerable to official harassment and 71) Vietnam PA, Table 3.6. 72) Esguerra et al., 2000. 73) Webster, op cit., p. 50. 74) Islam et al.., 2000. 75) Thomas, 2001, cited in Shareen Joshi literature review #3. 76) Vietnam Poverty Assessment, Table 3.5. PART TWO An Analysis of Quantitative 28 and Qualitative Information Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam intrusive regulations that undermine legitimate work. In Vietnam, for example, a recent decree has made it illegal for street vendors to hold a fixed place on the pavements, under threat of confiscation of their wares, that adds to the instability of their business. Similarly, cyclo drivers have been banned from the main streets of Ho Chi Minh City and some of the main routes in Hanoi, which cuts into their market.77 In the Philippines, informal urban labor is less organized than poor labor in rural areas has traditionally been, and so it has not made effective claims for social protection. 2.41 Housing for the urban poor represents not merely shelter, but also a workplace Physical Assets: and a source of rental income. In a well functioning real estate market, households Housing, Land can choose where they wish to live by balancing values such as convenience of access and to jobs and services against other amenities such as living space, and trade-off transport Infrastructure costs against housing costs. But in the cities of many developing countries, including those of East Asia, poor households especially are faced with only housing choices Housing and Land that leave them with low welfare. They must often accept living in a location that Tenure provides access to their employment possibilities, but has little tenure security and unsafe environmental conditions, so that their health, their financial assets, and their earnings are put at risk. Furthermore, the costs of such housing can still be substantial, especially when the associated nonmonetary costs (necessity for purchasing or self-provision Figure II.3 Vietnam ­ Urban Housing: Types of of alternative infrastructure services, absence of Temporary structures used, by per capita expenditure deciles public transport, exposure to pollution, etc.) are taken into account. 50 2.42 The living standards survey data for 40 Vietnam generally confirm that the quality of 30 housing for the urban poor is low, and only Percent 20 modestly better than that of the rural poor. 10 Almost half of the urban poor live in temporary 0 housing, versus 10.6 percent of the urban non- Poorest 1 2 3 4 5 6 7 8 9 10 Richest Per capita expenditure poor and 38 percent of the rural poor (Table VN10). Housing for the urban poor comes in a Temporary house Walls of leaves/branches Roof of straw/leaves wide range of qualities, with half of the lowest Note: "Other" types not graphed decile reporting brick or stone walls, but 17 Source: Table VN8, VN11 and VN12 percent having only leaves or branches (Table VN11). Roofs are similarly varied--half have tile or iron, but 20 percent only straw or leaves (Figure II.3 and Table VN12). 2.43 Urban housing offers more space per person than rural housing does, which is not generally expected given the higher land costs in cities. But the urban poor have the least space: 8.85 m2/person, compared to 11.4 for the rural poor (Table VN13). Nonregistered migrants are not included in this sample, and the quality of their housing would probably lower the urban averages. Participatory poverty assessments in Ho Chi Minh City and Haiphong found that certain locations for housing, in particular sites in areas scheduled for clearance or off main alleys and so farther from commercial activity, are considered a sign of poverty status.78 77) Lim et al., 2000. 78) Vietnam PA p. 10; Luan et al., 1999. 29 Figure II.4 Indonesia ­ Urban housing status: 2.44 As with Vietnam, the survey instrument Private ownership, rent, lease and "other," by per for Indonesia is not very sensitive to the kinds capita expenditure decile. of housing status and characteristics that would be most meaningful in the urban context. In Indonesia the share of the population in all deciles owning their houses is predictably much greater, and renting/leasing much less, in rural than in urban areas (Figure II.4), Table IN9). 2.45 The share of the urban poor in "free" housing (presumably squatters) is 6.4 percent (5.4 percent of the nonpoor), although even some of the 77 percent of the urban poor who own their housing and some of the 14 percent who rent or lease may lack secure tenure (Table IN8). The share of the lowest expenditure decile Source: Table IN9 in "free" housing is highest in settlements of 500,000-1 million (9 percent). Rural housing has Figure II.5. Philippines and Manila ­ Urban Housing slightly more space than urban housing has, Quality, by per Capita Expenditure Deciles which holds true for the poor as well: 34.5 percent of the urban poor have less than 40 m2 floor area, compared to 29.4 percent of the rural poor (Table IN8). Surprisingly, the urban poor have larger houses than the urban nonpoor do; however, the differential in quality across urban income groups is considerable (e.g., 9 percent of the poorest decile have bamboo walls compared to 0.7 percent of the richest), and is more pronounced than it is for rural housing (Table IN9). 2.46 In the Philippines, about half of the lowest urban expenditure decile own their houses and lots, which is close to the urban Note: Other "light" building materials not graphed average of 60 percent (Table PH5). The poor Source: Table PH6, PH7, PH8 and PH9 report a wide diversity of formal and informal ("with charge"/"without charge") housing arrangements, although actual tenure status is difficult to determine from the survey instrument. The proportion of households that live in the lowest quality houses ("improvised") is five times as high among urban as among rural households. Almost 10 percent of the poorest urban decile live in such housing (21 percent in Manila), compared to 2.7 percent as the urban average (Figure II.5, Tables PH6,7). 2.47 The use of salvaged building materials for roof and walls is much more prevalent among the poorest urban decile than it is for any other groups, but is also quite prevalent among the lower-middle income groups in urban areas, indicating possibly a poorly functioning construction market (Tables PH8,9). The share of the poorest decile living in houses of salvaged materials is almost twice as high in Manila as in all urban areas. Ownership is not usual for all income groups; only 40 percent of the poorest decile in Manila report full ownership of house and lot (Tables PH10, 11). As acknowledged in the Philippines Poverty Assessment, housing is a problem for the population as a whole. The prevalence of illegally occupied housing is highest for those in the middle of the income distribution and occurs even among the wealthy. That situation reflects PART TWO An Analysis of Quantitative 30 and Qualitative Information Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam fundamental problems in the housing market, possibly including corruption and bad regulation, which burden the urban poor the most.79 2.48 The Philippine living standards survey reveals that electricity is almost twice as Electricity and available "in their building" in urban as in rural areas (90 vs. 52 percent), and even 71 Durable Consumer percent of the poorest urban decile have such access (Table PH12). Rural access varies Assets much more sharply with income distribution, doubtless reflecting the costs and difficulty of connecting more remote and often poorer rural areas or the costs of acquiring private generators, compared to the low unit costs of providing networked service in denser settlements. Electrification is universal for all income groups in Manila (Table PH14). Reliability, or frequency of outages, is not measured in the survey data. Partly because of differences in service, the urban poor report more ownership of electrical appliances than do the rural poor, but not very frequent ownership outside Manila of any appliances but radios (which are widely available to all income groups in both areas) and television (owned by a third of the urban poor) (Table PH13 and PH19). 2.49 Vietnam has a higher rate of electrification than the Philippines, both for the total rural population (71 percent) and for the lowest urban decile (94 percent), despite a much lower per capita electricity consumption (Table VN14). In Vietnam television ownership is one of the strongest predictors of per capita expenditure in both urban and rural areas, and radio ownership somewhat less so. Vietnam has a higher ownership of television (battery-operated Chinese-made) than of radios, even in rural areas, and higher than that in the Philippines.80 Electricity is not available to low-income residents to use for heat or cooking. They mostly use coal or wood, both of which can contribute to health risks from indoor (and outdoor) air pollution.81 2.50 Access to safe, reliable water and sanitation is particularly important in urban Water supply areas, where the density of settlement limits the availability of water from private wells, further raising the public health risks from unsanitary disposal of liquid and solid waste. The poor often reside in low-lying areas prone to flooding and so are particularly vulnerable to water-borne diseases. Thus, even while urban access to water is usually reported to be higher than for rural populations, there remain serious problems with its service reliability, quality, and affordability for many poor urban residents. Unfortunately, the available living standards surveys do not delve into these questions. 2.51 Data from the Indonesian living standards survey confirm that a higher share of urban households than of rural use tap water (46.6 percent versus 11 percent) and that rural households are more likely to use such unprotected water sources as springs, rivers, and, rain water (Table IN10). Half of all urban dwellers, and 42 percent of the urban poor purchase water, compared to only 15 percent of the rural population, reflecting the fact that in urban areas, water access requires a cash outlay. More urban households have "private" drinking water facilities (house or yard taps)--66 percent, as compared to 48 percent for rural--and even 55 percent of the urban poor do so. These gross comparisons of rural and urban areas are not surprising. However, what is less evident from traditional poverty analyses is that the inequalities in sources of 79) Philippines PA, Chap. 1. 80) In rural and urban areas, respectively, the 1999 Population and Housing Census found that 46% and 76% of households owned TVs, and 42% and 56% owned radios. (Minot and Baulch, p. 19). 81) Luan, Mai and Anh, op. cit., p. 66. Pollution from various sources, but in particular, from household and industrial energy consumption, is reflected in concentrations of total suspended particulates (micrograms per cubic meter) averaging from 200 to well over 500 in a large sample of Chinese cities, 271 in Jakarta, 223 in Bangkok and 200 in Manila, but below 100 in reporting cities of Korea and Malaysia and below 50 in many OECD cities. World Bank, WDI, 1995 data. 31 Figure II.6. Vietnam ­ Urban sources of drinking drinking water are sharper between water, by per capita expenditure deciles. urban poor and urban rich than between the rural poor and rich. For example, in rural areas, twice as many poor as rich drink from unprotected water sources; but, the urban poor are five times more likely than the urban rich to use unprotected wells and to rely on public (communal) drinking water facilities. In the largest urban area (Jakarta), the poor are even more likely than the rich to purchase water (Table IN11). As other studies have documented in cities in Indonesia and elsewhere in the developing world, the poor tend to Source: Table VN16 pay higher unit prices and also higher shares of their income for water than do the rich, and they also Figure II.7. Philippines ­ Urban access to "own consume much smaller quantities faucet" rises with per capita expenditure status and less safe quality water.82 Moreover, the nonpoor in Jakarta have, overall, less satisfactory sources of water supply than do their counterparts in smaller urban settlements, despite Jakarta's capital city status. The failures in the water sector have widespread impacts on all groups, but especially on the city's poor. 2.52 In Vietnam there are sharp disparities in water sources between rural and urban areas, with rural Note: "Other water sources not graphed residents having negligible use of Source: Table PH12 tapwater whereas, on average, 53 percent of urban residents use it. However, the disparities across urban income groups are deep here as well. Only 16 percent of the urban poor have an indoor or outdoor tap, as compared to 57 percent of the nonpoor, and 64 percent of the urban poor rely on informal sources (hand-dug well, rainwater, surface water, or other), as compared to 25 percent of the nonpoor (Table VN15). The inequalities are even sharper, of course, across the extremes of the urban expenditure distribution (Figure II.6 and Table VN16). 2.53 In the Philippines the inequalities in use of different water sources are not very great between urban and rural households overall, apart from "own faucet," which clearly is more an urban asset, used by 37 percent of urban residents versus 7 percent of rural (Table PH15). However, again, large gaps appear across the urban expenditure distribution: e.g., 66 percent of the rich have private taps, versus 11.5 percent of the poorest group (Table PH12, Figure II.7). 82) Hardoy, Mitlin and Satterthwaite, 2001 PART TWO An Analysis of Quantitative 32 and Qualitative Information Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam 2.54 Shared facilities and Figure II.8. Indonesia ­ Urban sanitation (final disposal): unprotected sources are used by 75 Septic tank use and unsafe methods, by per capita percent of the urban poor, but by only expenditure decile. 41 percent of the urban nonpoor (Table PH15). The most striking feature in Manila is that all residents rely on "peddlers", twice as much as do residents in other cities (10.3 percent versus 5.3 percent), and the poorest four times as much (20.6 percent of the bottom decile). Faucet water is modestly more available to the poor in Manila, but one would expect that this major city would have much more formal water provision even for the poor than other urban areas have (Table PH14). It is possible that even the poor find the quality and reliability Source: Table IN12 of Manila's formal water supplies inadequate, and turn instead to peddlers. 2.55 In Indonesia the form of toilet facility (private or shared) does not vary greatly Sanitation between rural and urban, or even across the expenditure deciles, except that public facilities in urban areas are used almost exclusively by the poorest deciles (Table IN12). Squatter toilets are clearly the type preferred by the better-off population in both urban and rural areas, and even by 60 percent of the urban poor. Dry latrines are an alternative for the poor (used by 16 percent of the poorest versus 2 percent of the rich in urban areas). The use of squatter and throne toilets by the poor is more common in the larger urban areas (Table IN13). Safe waste disposal using septic tanks is twice as available in urban as in rural areas (73 percent versus 32 percent). An alarming percentage (40 percent) of those in the bottom third of the urban expenditure distribution Box II.1 The costs of inadequate sanitation in Indonesian cities Indonesia has one of the lowest rates of urban sewerage coverage in Asia. This causes widespread contamination of surface and ground waters all across the country. As a result there have been repeated local epidemics of gastrointestinal infections, and high incidence of typhoid (Foley, Soedjarwo, and Pollard, 2001). Economic losses attributed to inadequate sewerage are conservatively estimated at US $4.7 billion per year, and 2.4% of 1997 GDP--roughly equivalent to US $12/household/month (Asian Development Bank, 1999). The low coverage is partly a result of the Government of Indonesia policy, which assigns the responsibility for sanitation to households (World Bank, 1993). This policy--which is a result of the poor past performance of large centralized sewer systems--has inhibited the evolution of effective local governmental institutions for the planning, implementing and operating of sewer systems. Currently, 73% of urban households have on-site sanitation through septic tanks. The partially- treated, or untreated, effluent from these facilities typically flows into soils, open drains, or directly into water bodies. Proper disposal of human waste, either sewage or sullage, is a rare exception. Given the scale of the problem, interest in neighborhood or community-based sewer systems is increasing. One of the more recent successes of the latter approach in Malang, Indonesia, is summarized in Foley, Soedjarwo, and Pollard, 2001. Source: Joshi, 2001 33 rely on discharge to an open field, a ground hole or water bodies, risking contamination of the surface and groundwater in dense urban areas (Figure II.8) (See Box II.1). Even in Jakarta, 30.3 percent of the poor discharge their waste directly into water bodies (Table IN13). Government statistics confirm that in DKI Jakarta, the water sources of 30 percent of households surveyed are within six meters of a source of contamination such as a sewage discharge outlet.83 Figure II.9. Vietnam ­ Urban type of toilet: 2.56 In Vietnam, the highest by per capita expenditure decile form of sanitation (flush toilet with septic tank) is, predictably, overwhelmingly more available in urban than in rural areas (62 percent versus 4 percent), but only 26 percent of the poorest urban decile have this facility (Tables VN17, 18). Overall, half or more of poor households do not have a protected toilet facility. The poorest urban residents report less availability of a "simple toilet" than rural population residents do, and a disturbing 31 percent of the poorest report have no toilet at all (Figure II.9). 2.57 Though this figure of 31 percent is equivalent to that for the Source: Table VN17 poorest rural decile, its environmental health implications can be much worse in denser settlements. The gradient of access to sanitation facilities across the expenditure deciles in urban areas is much steeper than the gradient in rural areas-- another element of the serious overall urban inequality in services that is evident even without taking the situation of nonregistered migrants into account. 2.58 In the Philippines the urban poor are better served with toilet facilities than are the rural poor, but certainly not adequately served in terms of environmental health. Only about half (52 percent) of the poorest urban decile have use of a water-sealed toilet, and 33 percent use an open pit, pile or "no toilet" (Table PH16). Again, the disparity in access to different types of toilet is wider in urban than in rural areas: while twice more rural poor than nonpoor use an open pit or no toilet, the ratio in urban areas is 4-to-1 (Table PH17). In Manila, the poorest group has considerably better access to water-sealed or closed-pit toilets, and relies less on unsafe disposal, than is the case in other urban areas (Table PH18). 2.59 The living standards surveys do not investigate solid waste disposal, but in urban areas it also is a serious environmental health hazard.84 Typically not more than half of household trash is collected; one random sample survey in kampungs of Indonesia found that only a quarter of poor households have such service.85 What trash is collected, often by informal private agents, is usually dumped in unprotected sites, including drains and canals, rather than in regulated landfills. Otherwise, it is burned, adding to ground-level air pollution. The dire waste disposal situation in Manila became international news in July 2000 when a literal mountain of refuse in the city's largest 83) Badan Pusat Statistik (BPS), Environmental Statistics of Indonesia, BPS Catalogue 2202, Jakarta. Cited in ADB Urban Chapter, p. 12. 84) Hardoy, Mitlin and Satterthwaite, 2001 85) Lindfield and Lanyon (2000), cited in Wegelin, 2000, p. 12. PART TWO An Analysis of Quantitative 34 and Qualitative Information Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam dumpsite (Payatas) collapsed on squatters' houses, killing dozens of people.86 Untreated or toxic industrial waste is often dumped, unregulated, in urban areas, and the poor often live in or near the dump areas. 2.60 Motorization is increasing rapidly in East Asian cities. They vary greatly in their Transport efforts to manage transport modes so as to balance the demands from economic and demographic growth, residents' rising expectations about travel, and transport's environmental costs. Experience across the world has shown that inadequate policy attention to public transport, land and housing policies that discourage in-fill development, and unrestrained use of motor vehicles contribute to geographic spread of urban areas, often extending the distances between where people live and where they work. In low-income cities of East Asia, a legacy of high density, mixed-use urban form based on nonmotorized transport, buses and "jitneys" is generally hospitable to the needs of their poor residents. But this land use pattern is changing, as public and private investment favors motor vehicles. Commutes lengthen and urban congestion intensifies. Access and mobility are worsened for the poor especially.87 Another problem is that the high density of Asian cities means that even at existing levels of motorization, polluting emissions are very high per hectare with local hazards particularly severe for the urban poor, who are least able to protect themselves. 2.61 According to the UN Habitat's Urban Indicators Program (UIP) database,88 the Asia/Pacific cities' residents, on average, take longer to reach work (42 minutes) than do urban residents in any other region, because they use nonmotorized transport (walking and bicycles), for 23 percent of work trips--higher use than in any other region but Africa.89 In cities of Indochina and Mongolia where urban growth is still light, travel times to work remain relatively low (30 minutes). Most East Asian cities have travel times similar to those of South Asian and Latin American cities (Annex 2, Table A5). Bangkok's average time, however, at 60 minutes, may foreshadow the future for many cities; Metro Manila's average travel speed of 10 km/hour is only slightly faster than that of Bangkok. Private motorized transport already accounts for well over half of the travel in some Indonesian, Thai and Vietnamese cities. Public transport is the dominant mode in Singapore, Seoul, Ulaan Baatar, Penang, and Yangon (Myanmar). Walking and bicycles, along with other nonmotorized modes, are major forms of travel in some countries, especially in China, but the sheer size, geographic spread, and congestion of many Asian cities make these modes both impractical and dangerous as long-term solutions, even for the poor. In major urban areas of Vietnam, motorcycles have overtaken both bicycle and public transport use, accounting for 80 percent of central area traffic in Ho Chi Minh City. Their dominance adds to congestion and compromises both road safety and the environment.90 2.62 Lower-income groups are the main users of public transport and nonmotorized alternatives. In Metro Manila, where socio-economic class is strongly associated with forms of transport, over 80 percent of the poor and poorest strata depend on public transit for work trips.91 According to the living standards survey examined here, half of the richest decile in Manila own a motor vehicle or motorcycle, but only one percent of the poorest do; for all urban areas the ratio of motor ownership between nonpoor 86) Washington Post, July 11, 2000. 87) Barter, 1998. 88) UN-Habitat Global Urban Indicators Database, 1998 89) UNCHS, 2001b, p. 43. Regional averages should be taken with caution because the UIP database does not contain a representative sample of cities. 90) World Bank, 1999, p. 6. 91) Social Weather Station Bulletin 97-22, cited in World Bank, 2001c. 35 and poor is 13-to-1 (Tables PH13,19). When low-income communities are dispersed throughout the city, as in Manila, public transport networks need to be dense and widespread to serve them well. 2.63 In most cities of Asian developing countries, the majority of trips by lower- income groups are on foot, which effectively limits their overall mobility.92 The vulnerable road users (pedestrians, bicyclists and motorcyclists), among whom the poor are disproportionately represented, account for a high proportion of road fatalities.93 Women's mobility is particularly hampered by inadequate public transport that runs infrequently and is plagued by crime. In Bangkok, evening buses for women-only have been instituted in response to their demands for safer means of travel.94 Then too, men often control most of the family's transport resources. In some countries, women are not permitted even to ride bicycles, although they must travel to school, health care, and shopping for the family, even if not to work outside the home.95 Information on household expenditure for transport was not available from the surveys reviewed here. Such information would enable a fuller analysis of the poor households' transport problems. Insecurity and 2.64 Asnotedearlier,povertyischaracterizedbyastateofinsecurityandvulnerability Urban Poverty to risks--personal, communal and financial--that undermine the asset base of the poor and their ability to cope. Insecurity of Home 2.65 The major source of insecurity for urban households and their communities is and Place lack of recognized tenure for residents in informal or illegal settlements, placing them at risk of being forcibly evicted with little notice, losing their home and belongings, and their connections to work and neighbors. In squatter areas of Metro Manila, official evictions are a threat, but private developers and contractors are also known to set fires ("hot demolition") to clear out shanties and open rights of way.96 As noted above, a detailed breakdown of de facto and de jure tenure status was not available to this desk study. Such analysis would be critical to a more thorough examination of urban poverty. 2.66 A second major source of insecurity is natural disasters, to which low-income settlements built on flood-prone land or hillsides, or near factories, are particularly vulnerable. The very dense and haphazard layout of such neighborhoods, without storm drainage, access roads or cadastres, makes emergency evacuation and service by emergency vehicles slow and difficult, raising the human toll of natural disasters and industrial accidents. In a 1999 survey of communities and, specifically, of the "Class D and E" (poor and poorest) residents in several urban areas of the Philippines, they ranked problems of the water system, drainage and floods, and roads in their top five concerns, above the national macrofinancial crisis.97 92) Barter, op cit. 93) Barter, op cit. 94) UNCHS (Habitat), 2001a. The service was instituted by the Bangkok Mass Transit Authority in 2000 on a pilot basis. 95) Racelis, op. cit., p. 22. Men surveyed in three communities in Metro Manila reported that they make 60-80 percent of the households' total expenditures on transportation, about equal to their share for "vices" (alcohol, cigarettes and other). 96) Racelis, 2000. 97) Social Weather Report Survey (SWRS) June 2-16, 1999, covering the national capital area (Metro Manila), Luzon, Visayas and Mindanao, cited in Racelis, 2000. PART TWO An Analysis of Quantitative 36 and Qualitative Information Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam 2.67 In the urban context, personal insecurity stems from such health and traffic risks Personal as described earlier, and from crime. In surveys of poor urban communities, the vast Insecurity majority of residents report feeling unsafe and subject to crime, violence, and harassment, much of it drug- and gang-related. In the same Philippine survey mentioned above, drug addiction ranked first or second (following only unemployment) as a major concern, and robbery was among the top five problems cited in two of the communities. Slum dwellers are victims not only of violent crimes, which tend to be concentrated in the poorer neighborhoods of cities, but also of the related negative stereotypes held about their neighborhoods and by association, about them. Crime therefore intensifies the exclusion and marginalization of the urban poor. 2.68 Financial risks for urban populations arise from their dependence on cash income Financial and expenditure, which are sensitive to vagaries of the economy at all levels. Developed Insecurity countries have various degrees of formal safety nets, and microcredit programs can be important in low- and middle-income countries. However, for urban as well as rural households in East Asia, as elsewhere in the developing world, reliance on informal private transfer mechanisms to supplement income is widespread, especially when times are difficult. Mutual financial support among families and communities, including informal credit groups, also may shrink in hard times, as reported from household interviews in Jakarta.98 2.69 A study of private financial exchanges in Vietnam found that virtually all such cash transfers were reported as sent to family members, and 70 percent are sent from a city/large town to a small town/rural area.99 An urban survey in the Philippines found that nearly 40 percent of urban respondents send money to their rural kin, but only 6 percent receive any financial transfers from other urban areas or abroad. The individuals providing remittances reported that the unavoidable social obligation to support the extended family, even when under economic pressure themselves, makes it difficult to save.100 The Philippines Poverty Assessment found that the relatively low poverty incidence for three categories of the population normally considered vulnerable-- female-headed households, the elderly and the unemployed--is due to significant cash transfers supporting them, almost all of which are from private sources. Transfers are large and increasing in the Philippines, accounting on average for 13 percent of pre-transfer household incomes in 1997. Just over half (57 percent) of the transfers originate from abroad, with the remainder mainly from domestic urban areas. The transfers are found to be highly progressive--benefiting more the households with low levels of pre-transfer income. 2.70 A multi-country study by the ADB also found that during the recent financial crisis, private transfers were a much more important safety net than public programs. In a sample of Indonesian communities surveyed, of the 44 percent of respondents who reported receiving cash transfers, 20 percent were obtained from the government and NGOs and 80 percent from family and friends. It was the better-off families who received more of the public funds.101 It is evident that the flow of funds from (mainly) urban to other urban and rural households through private transfers constitutes a valuable and well-targeted mechanism that alleviates poverty and financial insecurity. Financial risk-sharing through social and family networks, which is evidently important for the urban poor, is also an urban-based strategy that reduces poverty throughout the country. 98) Craig et al., p. 34. 99) Tabulations from the Vietnam VLSS (1997-98). Cox et al. 2001. 100) Survey in PNR Bangkal by Schelzig (1999 - incomplete reference) cited in Racelis, 2000. 101) Reyes et al., 1999, cited in Craig et al., 2001. 37 Empowerment, 2.71 It is often argued that the urban population has better access than the rural Social Capital population to channels for influencing government decisions and obtaining favors. and Urban Behind this superficial generalization, the realities of political voice and relations with Poverty government are much more complex, especially for the urban poor, who express as much sense of exclusion and powerlessness as do the rural poor. The urban poor therefore rely heavily on their own social organizations and networks, and on informal mechanisms for services. That reliance has many worthy aspects, but it is also a sign of the government's failure to acknowledge the urban poor and treat them fairly in its own backyard. Relations with 2.72 Living in geographic proximity to seats of government, the urban poor may Government have higher expectations than do the rural poor. In participatory poverty assessments, the urban poor often note their exclusion from information about government, their distrust or unawareness of its intentions, and their widespread experience of the most negative official behavior, corruption. In a "Consultations with the Poor" study of both rural and urban populations in Indonesia, respondents rated several institutions: government, religious, private commercial, and NGO. Neither government services nor programs, nor any NGOs (which often are rural-based) ranked among the top five institutions that the urban poor selected on criteria of importance, effectiveness, trustworthiness, and openness to community influence. Men and women, in both rural and urban areas, agreed that they could not influence government programs at all.102 2.73 Even in the Philippines, whose experience with democracy and decentralization has been longer than Indonesia's, the urban poor are disillusioned with government as not helping and even harming their interests, despite its agents' physical proximity. The failures of government programs to deliver as promised, for example, land tenure through the Urban Housing Act, gives rise to a sense of suspicion and betrayal.103 The demands of the poor may in some respects seem unrealistic; on the other hand, it is not clear that the government has seriously tried to educate the citizens about what they should expect from it. One study of the urban poor in Naga City in 1998 found that 74 percent were unaware of government programs intended to help them. Almost all were unaware of the relevant legislation on urban development and housing.104 2.74 The broader SWRS survey of several urban populations, and including specifically their poor households, that was cited earlier showed that they overwhelmingly expect from government "economic services" (provision of jobs, credit, lower prices, higher wages, projects for the poor), and to a lesser extent infrastructure, social services, and public order.105 The relatively low ranking of the latter elements, which some may consider government's major role, may reflect past populist governments' history of job programs and price controls, and their lack of interest in revamping the agencies that deliver core public services. That survey's respondents assigned responsibility for local problems to local, not national, government. About as many expressed the view that the government was not doing anything about problems, as gave it credit for "probably doing something." The survey revealed fairly high regard for the then- Administration's "readiness to cooperate with the poor to fight poverty."106 Respondents rated the national government's overall performance on housing programs as 102) Indonesia Poverty Assessment, Box A2.1. 103) Craig et al, 2001. 104) Ateneo Social Science Research Center (1998), "Socio-Economic Profile of the Urban Poor in Naga City", cited in Racelis, 2000. 105) SWRS June 2-16, 1999. 106) This survey was undertaken during the Administration of former President Enrique Estrada, who had campaigned to fight poverty and had strong political support among the urban poor. PART TWO An Analysis of Quantitative 38 and Qualitative Information Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam unsatisfactory, which is consistent with data on their regressive impact; however, the Community Mortgage Program, which is well-targeted and efficient in helping the poor acquire land and improve their housing, is held in much higher esteem.107 2.75 Corruption is widely perceived as an impediment to obtaining government services and benefits fairly and efficiently. In the above Philippine survey, only the very poor respondents expressed overall confidence that the national administration was trying to eradicate corruption (possibly reflecting faith in then President Estrada in particular). In areas of Jakarta, local corruption had interfered with an emergency rice subsidization program funded by the World Food Program at the height of the financial crisis, but residents had mobilized a boycott and forced corrective action.108 2.76 The social interactions and relationships among urban households are more Social Networks significant than their dealings with government or other institutions in helping them to among the Urban manage urban life. A large literature on developing countries has established that, to Poor pool risk, households devise various strategies of collaborating with others both within and outside the family.109 This research on risk, insurance, and uncertainty has been based largely on studies of rural households, and there are few such studies on urban poverty. Yet, risk sharing is likely to be a central feature in poor urban households, as Cox and Jimenez (1998), in a study of the Philippines, demonstrated. Several qualitative studies already cited here have highlighted the importance of networks in cities around the world, and particularly in East Asia, in helping urban households cope with risks.110 2.77 As discussed in Part One, the risks faced by the urban poor tend to be very different--based more on insecure property rights, environmental health, and unemployment than on climate and natural resources--from the risks faced by the rural poor. Like their rural brethren, the urban poor cope with such types of shocks by drawing on their social networks in conjunction with their personal assets. As Jellenik (1991) in her superb ethnography of a Jakarta slum points out, these social networks are likely to consist of both relatives and friends. Urban slums are very dense, often with many families living in the same house. The density tends to move social relationships away from the traditional forms that characterize village networks. Marriages are much less stable. Both women and men are more likely to engage in serial monogamy and consequently have several circles of relatives. Relationships are based more on the quality of reciprocal links between individuals and friends, rather than on familial obligations. Such reciprocal connections are built on helping each other to obtain housing and jobs, and to cope with illness or credit problems. 2.78 Networks in urban areas, besides being an asset for coping with shocks, are also an important source of mobility. Typically, individuals who migrate to the city come knowing a friend or relative with whom they can stay and who may help them find a job or connections to other people who might help. These social networks are indispensable for improving the living standards of the poor. In the extended peri- urban areas outside East Asian cities, the torrent of new workers into a relative social vacuum leaves community networks weaker and thinner than those in longer-established core urban settlements, and so there are fewer channels for informal housing, credit and other services.111 107) SWRS and other studies cited by Racelis 2000; the Bank's Philippines PA (Chapter 4) documents the regressive incidence of the public housing programs in general (not distinguishing among them). 108) Craig et al., p. 40. 109) Such literature includes: Besley, 1995; Coate and Ravallion, 1993; Cox and Jimenez, 1998; Rosenzweig and Stark, 1989. 110) Moser, 1996; Lim, Arce and Racelis, 2000; Craig et. al., 2001; Jellinek, 1991. 111) Webster, 2001. 39 2.79 A poor migrant's social networks, however, are not only in the city but also in the village. S/he often supports parents, spouses and other relatives in the village. When a migrant is successful in the city, that can improve his or her family's living standards in the village as well. On the other hand, during periods of adversity, the rural-based family may help the migrant with gifts of food and money.112 2.80 The nature of community in urban areas is also very different from village life. Newly arrived poor migrants may live in recently established squatter settlements, but they participate in a strong sense of community. They often live in enclaves of people from their home villages, or else with relatives who have developed strong ties within a more diverse community. Slum communities usually have well-defined leaders who mediate the relationship of the families in the slum to the State. As in rural areas, the ability of urban communities to engage in collective action may be important for determining how well they can obtain and access public services. This is particularly true in capital cities such Jakarta or Manila, where the urban poor--because of their numbers and the relative ease with which they can be organized--are an important element in the political culture. In a survey of poor people in Naga City in the Philippines, households reported a 91 percent participation rate in various organizations (averaging 1.4 members per household); of those participating, 71 percent belonged to organizations specifically of the poor.113 A slum community's willingness to engage in collective action can be mobilized by its leaders to obtain services. It also can be manipulated for private gain or unofficial corruption. It has been reported that among squatter communities in Manila, "toughminded syndicates often organized the occupation of such land and distributed sites for a price and monthly fees".114 One resident's mafia may be another's service provider. 2.81 In other words, networks may not always be beneficent. The ability of a community to engage in collective action can also make it an effective source of political violence. The obligations towards one's social network bring demands that the poor, who are constrained for time, may not always want to meet. Moreover, upward mobility can reduce the need to rely on old friends and relatives, so when a lucky member of the community moves ahead, ties with his or her roots may loosen; and thus webs of obligation can be lost. Then, too, networks can be means of exclusion as well as inclusion. In Vietnamese cities the government's harsh policy towards unregistered migrants makes them outcasts even to the poor residents. That makes it more difficult for the two groups to organize together and manage common problems of poverty.115 2.82 In short, urban social networks may be as central to the survival for urban poor as for the rural residents. The urban social fabric, though it may be more varied in texture and color, more easily torn and rewoven to fit changing needs, may still be indispensable as a basic shield and safety net for the urban poor. 112) Craig et al., 2001. 113) Racelis, 2000. 114) P. Strassman and Blunt, 1993. 115) Lim et al, 2000. PART TWO An Analysis of Quantitative 40 and Qualitative Information Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam PART THREE: IMPLICATIONS FOR POLICY, INSTITUTIONAL DEVELOPMENT, AND RESEARCH 3.1 This review of living standards survey data and other evidence points to certain Summary: Urban observations and preliminary conclusions about the directions and magnitudes of urban Poverty in the poverty trends and problems, at least in three major countries of the Region. The EAP Region findings here suggest recommendations about relative priorities and potential focus areas for policy and institutional follow-up, and also for research toward the Bank's more exact and comprehensive understanding of urban poverty, which will be useful among its clients and partners as well. 3.2 The main points that have emerged can be summarized as follows: · Over the last decade the countries of the Region have reduced urban poverty along with overall poverty in income/expenditure terms, but over the medium term, the urban poor are poised to increase both in numbers and as proportions of the total urban populations. In urban areas income inequality is relatively high (higher than in rural areas) and rising. Numerous sources have demonstrated great intra-urban inequities as well, in access to basic services and in welfare status across housing, land tenure, water and sanitation, and transport. · Income, or expenditure-based, measurements of urban poverty are highly sensitive to the assumptions used. Furthermore, much of the urban population is close to the estimated poverty line. The recent macrofinancial crisis demonstrated this: it hit urban populations particularly hard and sent many into poverty at least temporarily. Possibly suggesting the limitations of income- and expenditure-based measurements, the self-ratings of poverty status are high among urban populations. · Further caveats about the living standards surveys concern the limitations of the picture they can convey about urban poverty, for several methodological reasons: they do not fully capture the mixed livelihood sources (rural plus urban) of many households; the samples are typically too small to permit disaggregation among and within urban areas; sampling frames are often outdated, given the rapid changes in urban population; and the survey instrument is insufficiently tailored to the peculiarities of urban life. · The living standards survey data reveal that the poor and poorest among the urban populations of Indonesia, the Philippines and Vietnam are not the usually expected "vulnerable groups" (i.e., not female-headed households nor the elderly). They do include large families, and private transfers of urban origin may be a main reason that those vulnerable groups escape poverty. 41 · The migrants surveyed are not less well off than longtime residents. However, a major shortcoming of the Vietnam survey is its failure to capture the non-registered migrants, who suffer official exclusion from services and benefits. · The urban poor also are not necessarily unemployed, although the survey instruments are often unsuited to capturing the vagaries of urban earnings patterns, especially in the informal economy. Due to the commoditization of urban life, meaning that cash is required for essential goods and services, a stable source of cash income is essential to avoid poverty. · Although indicators improve generally with the increasing size of settlements, the largest urban areas (Manila and Jakarta) are not necessarily the most favored, and in some respects (e.g., water and sanitation in Jakarta), the poor residing there are worse off than in other types of settlements. The theoretical economic advantages of agglomeration can be lost in practice, as far as the poor are concerned, unless good urban management is in place. · The housing status of the urban poor is marked by a high degree of crowding, tenure insecurity and risk of forced eviction. There is a large quality differential across urban income groups, although in the Philippines housing conditions are surprisingly poor even for the urban middle class. · Although access to education, health facilities, water and sanitation is predictably higher overall in urban than in rural areas, data must be broken down to examine both effective access and quality of service, and outcomes--across income groups, across zones of the city, and among urban localities, if the welfare status of different groups, in absolute or relative terms, is to be determined. Such disaggregations are not feasible from most of the survey databases. · From the information available, large disparities in educational attainment are evident among the urban population. Moreover, in the urban setting, intermediate levels of education do not necessarily translate into higher incomes. · Health status in urban areas is harmed not only by behaviors, multiple stresses and environmental risks. In the Philippines, infant and child mortality are higher for the urban poor than for the rural poor. Health outcomes have been found to vary even more across zones of a city where this has been studied, although the living standards surveys examined here do not permit such an analysis. Undernutrition and hunger are evident in some urban areas of the Philippines and Vietnam. Risky behaviors like smoking, disaster-prone living conditions, incidence of crime, violence, and HIV-AIDS, and traffic accidents all affect health in urban areas, especially among the poor. · Sources of water supply and the quality of sanitation diverge greatly across income groups in the urban areas. The poor are more likely than the rich to pay for water in Jakarta and Manila. The large share of the urban poor without basic sanitation or safe waste disposal in Vietnam and in Jakarta pose major health risks for them and for the entire city populations. · The high density of East Asian cities has facilitated use of nonmotorized transport and walking for most transport, but this is changing with urban growth. The mobility and ease of access for poor urban residents are most likely to suffer from deteriorating public transport and increasing traffic congestion. Their welfare is also harmed by their susceptibility to traffic accidents, ground-level pollution, and transport-related crime. · The urban poor face a covariance of threats to their personal, financial and communal security, stemming from uncertain tenure, macroeconomic shocks that affect earnings and prices, crime and other social pathologies (e.g., drugs). Private financial transfers have more effect than public transfers in mitigating financial risks of the poor. PART THREE Implications for Policy and 42 Research Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam · Despite their physical proximity to seats of political power, the urban poor have little influence on policies or programs affecting them, unless they organize; generally the urban poor perceive themselves to be excluded by government and, at the same time, highly vulnerable to government corruption. · The urban poor have many complex social networks, and they serve many functions: social integration, mutual support, labor market facilitation, and collective action to obtain services. This social capital, which is highly diversified, is more important for the urban poor than are formal relations with government in helping them cope with urban life. Nonetheless, maintaining networks can also be burdensome on individuals, and in some cases networks can have malign effects 3.3 The preceding desk analysis is largely limited to describing the correlates of Policy and urban poverty in the focus countries, rather than the causes, which would require Institutional deeper investigation. The review is also limited mainly to identifying the urban poor Implications of in per capita income or expenditure terms, since the data sources provide less detailed Urban Poverty markers of non-income poverty. 3.4 An overall picture emerges of vulnerability as a key characteristic of urban poverty. Further study seems warranted on certain important underlying contributory or causal factors: tenure insecurity, problems of governance, dysfunction in labor and land markets, or factors in the investment climate impeding the growth of employment that could provide good livelihoods. Other areas to examine closely are key policy and institutional dimensions: the incidence of public expenditure, the effectiveness and efficiency of formal safety net programs, systems of political representation and legal protections, public access to information, as well as sectoral policies governing education, health, housing, and basic infrastructure services. Exploring those may help to identify approaches that could alleviate and reduce poverty in its multiple dimensions. But even the narrow focus of this desk study suggests some factors that may be driving urban poverty and should be addressed. 3.5 As discussed in Part One (D), one element of urban poverty could be considered a function of temporary mismatch between the supply and demand for jobs and for services during periods of rapid in-migration. If institutions are both responsive and efficient, they can reduce such gaps by providing more resources for services and by removing specific bottlenecks such as undue regulatory barriers. But in the countries viewed here, and arguably in many others, much of the poverty appears due to deep seated political and institutional factors that shut certain groups out from the opportunities and protections that others can expect from either markets or the government. The empirical finding of deep and pervasive inequalities in the urban areas indicates that fundamental roots of poverty lie in the structures of governance. 3.6 To relieve the former type of poverty--which may be characterized as waiting in a queue that moves by fair, well-known and accepted rules--external assistance can accelerate the queue by applying more public investment resources and helping to identify and establish measures to accelerate supply, such as support for private sector participation. Also useful would be providing information and removing other obstacles to satisfying demands for goods and services that an be met, at least in part, by well- functioning markets (e.g., housing). Such processes depend on donor assistance for infrastructure policy reform and improved investment, and for private sector development, which are very relevant to a strategy for relieving poverty. Those processes 43 remain a priority for the Region, particularly if they focus on constraints on services for low-income users, such as tariff structures or regulations that are obstacles to reaching such consumers. Another high priority is encouraging municipalities to change policies that obstruct land development, and to undertake flexible urban planning that can steer new development so as to forestall the creation of new slums as cities grow. 3.7 To get at the deeper problems that keep the urban poor disempowered and vulnerable and that block their taking advantage of improved opportunities, underlying governance issues need to be addressed directly. The insecurities of housing and land tenure, livelihood insecurity, and physical insecurity to which low-income urban residents are subjected, call for bold policies and programs to strengthen legal protections and property rights, to reduce corruption and arbitrary administrative actions, and to foster the communities' own social capital. Community-driven slum upgrading programs that are based on the active involvement of the residents to improve a wide range of physical and communal services in neighborhoods would have broad benefits. Many sources of vulnerability could be reduced by such grassroots programs, provided the local and national governments are committed to them as part of a framework of better governance that recognizes poor residents as full citizens with rights and responsibilities, rather than simply tolerating them through isolated actions. Along similar lines, it is important to plan interventions that raise the capacity of local governments to carry out their basic functions in a responsive, transparent and accountable manner. Local governments also should be encouraged to undertake participatory strategic planning of their investments and other activities. Those changes would transform the relations and rules of the game between the local government and poor residents--as well as others. 3.8 As also noted in Part One (D), how well urban growth is managed in the Region will matter greatly not only for levels of urban poverty, but even more for how well the urban areas will contribute to the growth of the national economy and hence, also, to the prospects for reducing rural poverty. The process of urban transition and the potential economies of agglomeration can raise productivity for both rural and urban residents, but only if the basic mechanisms--a well integrated internal market for labor and goods, with ease of movement and good information flows, and lowered production costs from shared infrastructure--are actually working. Donor support to policies and programs that make the urban economies function efficiently and thus raise the returns to private investment there would be good value to the nation as a whole. 3.9 Such policies and programs would, for example, be positive towards internal migration--and certainly remove residence restrictions; they would facilitate urban- rural remittances however appropriately as a main source of private transfers; favor internal marketing by improving transport and communications where needed; correct distortions in the financial markets and make possible credit and savings for the poor. Particular attention to housing sector reforms and to the Region's urban transport needs is crucial. The widespread inadequacies in housing and transport, which affect many urban residents but are especially debilitating to the poor, should be among the most urgent concerns of governments and donors because those failures undercut the basis of the urban economy, which is a fluid labor market. 3.10 Finally, the findings on urban poverty and inequality suggest that any policy or program interventions that are intended to be poverty reducing should be founded on more detailed information about targeting and about impacts among the urban population. Similarly, analysis of public expenditure should go beyond aggregate attributions to "rural" or "urban" beneficiaries and more accurately identify the distributional reach within the urban population and in different urban areas. PART THREE Implications for Policy and 44 Research Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam 3.11 The present analysis raises important issues about the common methods and Suggested sources for empirical research on poverty, such as living standards surveys, which are Priorities for often ill suited to the context of urban poverty. Alone, they are unlikely to reveal its Research into nature accurately or to discover the relevant distinctions with respect to rural poverty. Urban Poverty in Because the research methodology of living standards surveys has been developed the Region and applied most extensively in rural settings, the sampling design and the construction of survey instruments tend to be based upon rural notions of life. 3.12 The implications of both of those problems were discussed in Part II (A). Because urban populations are constantly changing, homes often contain several families. Because unregistered migrants are not counted as urban residents, decennial census- based sampling frames quickly become non-representative and they are likely to miss the poorest residents unless they are frequently updated. The survey instruments and methods are designed from rural experience with the nature of households, work, and living conditions, and are often not adapted adequately to capture the complexities of urban livelihoods and society, or the multi-spatial nature of households rooted in both places. 3.13 Techniques exist to ensure more accurate sampling and to design well-tailored survey instruments, but best practices are not always applied. Some alternative approaches to sampling and to survey design are outlined in Annex 1 to this report. In particular, conducting special-purpose surveys to supplement national living standards surveys, acquiring panel data, and combining quantitative and qualitative information would delve more adequately into many of the issues and dimensions of urban poverty, with sufficient detail across and within urban areas. 3.14 Using appropriate methods to obtain a more complete and accurate picture of urban poverty, certain specific issues and questions should be explored. From the preceding review, some research topics are suggested here: · The role of interspatial mobility (migration and multi-spatial livelihoods) in affecting poverty in both rural and urban areas. This would look at circular and temporary rural-to-urban migration as well as at longer-term movement, and would examine both the role of private transfers and the mechanisms by which rural migrants become integrated into the urban society and economy. · The dynamics of informal employment in urban areas--the quality of livelihoods and the patterns of occupational mobility in the context of policy and institutional conditions. · How social networks among the urban poor function to help members cope with poverty in its various aspects; the networks often have links with rural residents. · A disaggregated analysis of health outcomes by zones within some major cities, and comparisons among cities of different sizes, growth rates, and levels of service provision, to map health-related poverty in greater geographic and socio-economic detail than is currently available. · Evaluation of the impact of specific interventions or packages of interventions, such as neighborhood infrastructure improvements and tenure security, on the well- being of low-income residents. The role of social capital and how it affects or is affected by residents' participation in such interventions would also be a question for such research. 45 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam References Ahuja, V., B. Bidani, F. Ferreira, and M. Walton. 1998. Everyone's Miracle? Revisiting Poverty and Inequality in East Asia. Washington, D.C.: The World Bank. Anh, Truong Si, Patrick Gubry, Vu Thi Hong, and Jerrold W. Huguet. 1996. "Migration and Employment in Ho Chi Minh City." Asia-Pacific Population Journal., Vol. 11, No. 2, 3-22. Asian Development Bank. 2001. Cities Data Book: Urban Indicators for Managing Cities. Manila. April Draft. Asian Development Bank. 2000. "Fighting Poverty in Asia and the Pacific: The Poverty Reduction Strategy." available at: http://www.adb.org/Documents/Brochures/Fighting_Poverty Asian Development Bank. 1999. "Towards a Community Based Environmental Sanitation Program for Indonesia." Dillion Consulting and PT Dacrea Design and Engineering, for Asian Development Bank. (TA No. 2805- IN)). Jakarta, Indonesia. Barter, A. and Rahman Paul. 1998. "Transport and Urban Poverty in Asia: A Brief Introduction to the Key Issues," UNCHS (Habitat) Regional Symposium on Urban Poverty in Asia, Fukuoka, 27-29 October. Besley, Timothy. 1995. "Savings, Credit and Insurance," Chapter 36 in Handbook of Development Economics, Vol. 3A, Jere Behrman and T.N.Srinivasan (editors), Amsterdam: Elsevier. Betcherman, G., and R. Islam, 2000. East Asian Labor Markets: Impacts, Responses and Lessons. Washington D.C.: The World Bank. Brockerhoff, Martin and Ellen Brennan. 1998. "The Poverty of Cities in Developing Regions," Population and Development Review, 24(I):XX-XX (March), #p. 1-40. Campbell, Tim. 2001. "A Tale of Two Cities in Vietnam: Towards a Strategy for Growth, Poverty Reduction, and Environmental Quality in the Cities and Regions of Vietnam," The World Bank. China National Bureau of Statistics. 1999. "A Brief Introduction of Urban Household Survey." Memorandum. Coate, Stephen and Martin Ravallion. 1993. "Reciprocity Without Commitment: Characterization and Performance of Informal Insurance Arrangements." Journal of Development Economics, 40: 1-24. Cox, Donald and Emmanuel Jimenez. 1998. "Risk Sharing and Private Transfers: WhatAbout Urban Households?" Economic Development and Cultural Change, Volume 46, No. 3. Cox, Donald, Emanuela Galasso, and Emmanuel Jimenez. 2001. "Private Transfers in a Cross Section of Developing Countries". The World Bank. Cour, Jean-Marie. 2001. "Vietnam: Migrations, Urbanisation, and Changes in the Rural World," Paper prepared for the Franco-Vietnamese Economic and Financial Forum, May. Craig, Mary, Alejandro de la Fuente Meraz, Simin Ho, Hanayo Kuroda, Fredric Gilles Serpoul, Stephen Sokol, and Jocelyn Songco. 2001. "Krisis: How the Urban Poor Cope," Columbia University School of International and Public Affairs, Economic and Political Development Concentration, SIPA/World Bank Report, May. Dang, Ahn, Sidney Goldstein, and James McNally. 197l. "Internal Migration and Development in Vietnam." Center for Migration Studies of New York. Esguerra, J., A. Balisacan, and N. Confessor. 2000. "The Philippines: Labor Market Trends and Government Interventions following the East Asian Financial crisis." in East Asian Labor Markets and the Economic Crisis: Impacts, Responses and Lessons, edited by G. Betcherman, and R. Islam. Fan, Shenggen, Cheng Fang, and Xiaobo Zhang. 2001. "How Agricultural Research Affects Urban Poverty in Developing Countries: The Case of China," Paper presented at International Food Policy Research Institute, Washington, D.C., June 18, 2001. Fang, Cheng, Shenggen Fan, and Xiaobo Zhang. 2001. "Emergence of Urban Poverty and Inequality in China: Evidence from Household Survey." Paper presented at the International Conference on Urbanization in China: Xiamen, Fujian Province, China, June 26-28, 2001. 47 Foley, S., A. Soedjarwo, and R. Pollard. 2001. "Case Study: Of the People, By the People, For the People: Community-Based Sewer Systems in Malang, Indonesia." United Nations Development Program and World Bank, Water and Sanitation Program for East Asia and the Pacific, Jakarta. de Haan, Arjan. 2000. "Livelihoods and Poverty: the Role of Migration," Presentation at the World Bank, March 9, 2000. Hamid, Naved and Mildred Villareal, editors. 2001. Asian Cities in the 21st Century: Contemporary Approaches to Municipal Management. Volume V: Fighting Urban Poverty. Manila: Asian Development Bank Institute and Asian Development Bank. Hardoy, Jorge E., Diana Mitlin, and David Satterthwaite. 2001. Environmental Problems in the Developing World. London: Earthscan Publications. Henderson, Vernon. 1999. "The Effects of Urban Concentration on Economic Growth," Brown University working paper, December draft. Hentschel, Jesko and Jesse Bump. 1999. "Urban Poverty Dimensions ­ Some Cross-Country Comparisons," World Bank, December draft. Hugo, G. 2000. "The Crisis and International Population Movements in Indonesia." Asian and Pacific Migration Journal, Vol. 9 (1), 93-129. R. Islam, G. Bhattacharya, S. Dhanani, M. Iacono, F. Mehran, S. Mukhopadhyay, and P. Thuy, 2000. "The Economic Crisis: Labor Market Challenges and Policies in Indonesia." in EastAsian Labor Markets and the Economic Crisis: Impacts, Responses and Lessons, edited by G. Betcherman and R. Islam. Jellinek, Lea. 1991. The Wheel of Fortune: The History of a Poor Community in Jakarta: University of Hawaii Press. Jones, G.W. 1998. "Urbanization Trends in Southeast Asia: Some Issues for Policy," Journal of Southeast Asian Studies, March. Joshi, Shareen. 2001. "Urban Poverty in East Asia A Literature Review," Yale University. Background paper prepared for East Asia Region Urban Development Division (EASUR) Poverty Study, World Bank. June draft. Joshi, Shareen. 2001. "Urbanization and Migration in EastAsia :ALiterature Review," Yale University. Background paper prepared for East Asia Region Urban Development Division (EASUR) Poverty Study, World Bank. May 2 draft. Lim, Liza, Wilfredo F. Arce, and Mary Racelis. 2000. "Urban Poverty from the Perspectives of the Urban Poor in Vietnam." Ateneo de Manila University. Draft background paper prepared for East Asia Region Urban Development Division (EASUR), World Bank. Lindfield, M. and A. Lanyon. 2000. "Community Upgrading and Poverty Alleviation." Presentation to World Bank Seminar on Upgrading Projects, Jakarta, Indonesia. Luan, Trihn Duy, Nguyen Xuan Mai, and Vu Tuan Anh. 1999. "Poverty and Social Issues in Haiphong City: A Report on Socio-Economic Survey," Haiphong and Hanoi, March. Lucas, Robert E.B. 1999. "Internal Migration and Urbanization: Recent Contributions and New Evidence," Background Paper for World Development Report 1999/2000. Minot, Nicholas and Bob Baulch. 2001. "The Spatial Distribution of Poverty in Vietnam and the Potential for Targeting." DFID Poverty Trust Fund and World Bank Development Economics Research Group, Revised April 18. Moser, Caroline O.N. 1998. "TheAsset Vulnerability Framework: Reassessing Urban Poverty Reduction Strategies." World Development, 26(1): 1-19. Moser, Caroline. 1996. Confronting Crisis: A Comparative Study of Household Responses to Poverty and Vulnerability in Four Poor Urban Communities. Environmentally Sustainable Development Studies and Monographs Series No. 8. Washington, D.C.: The World Bank R e f e r e n c e s 48 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam Moser, Caroline, Michael Gatehouse and Helen Garcia. 1996. "Urban Poverty Research Sourcebook Module II: Indicators of Urban Poverty," UNDP/UNCHS (Habitat)/World Bank, Urban Management Program Working Paper No. 5, September. Ogawa, N. 1985. "Urbanization and Migration in SelectedASEAN Countries: Trends and Prospects," in Urbanization and Migration in ASEAN Development, edited by M. Hauser, D. B. Suits and N. Ogawa. Tokyo: National Institute for Research Advancement. Pernia, Ernesto M. 1994. Urban Poverty inAsia: ASurvey of Critical Issues. Hong Kong: Oxford University Press. Phan,Anbinh. 2001. "Peri-urbanization in Vietnam and Thailand," Unpublished thesis, Pomona College, California. Popkin, B.M. 1999. "Urbanization, Lifestyle Changes and the Nutrition Transition," World Development, 27(11): 1905-1916. Pradhan, Menno, Asep Suryahadi, Sumarto Sudarno and Lant Pritchett.2000. "Measurements of Poverty in Indonesia : 1996, 1999 and Beyond." Social Monitoring and Early Response Unit (SMERU) Working Paper, Jakarta, Indonesia. Pritchett, Lant, Asep Suryahadi, and Sudarno Sumarto. 2000. "Quantifying Vulnerability to Poverty: A Proposed Measure, with Application to Indonesia." SMERU Working Paper, Revised May. Prud'homme, Remy. 1994. "On the Economic Role of Cities". Joint OECD-Australian Government Conference: Cities and the New Global Economy, Melbourne, November 20-23. Racelis, Mary. 2000. "Urban Poverty from the Perspectives of the Urban Poor in the Philippines," Draft background paper prepared for East Asia Region Urban Development Division (EASUR), The World Bank. Rao, Vijayendra. 2001. "Potters and Slums: Two Qualitative/Quantitative Projects in India," in Qual-Quant: Qualitative and Quantitative Poverty Appraisal--Complementarities, Tensions and The Way Forward ­ Contributions to a Workshop Held at Cornell University, edited by Revi Kinbur. March. http:// www.people.cornell.edu/pages/sk145/papers/QQZ.pdf. Ravallion, Martin. 2000. "On the Urbanization of Poverty," World Bank Development Research Group Working Paper. Reyes, C.M., R.G. Manasan,A.C. Orbeta,. and G. de Guzman,1999. "The Social Impact of the Financial Crisis in the Philippines--Working Paper No. 5." Asian Development Bank. Rosenzweig, Mark and Oded Stark. 1989. "Consumption Smoothing, Migration and Marriage: Evidence from Rural India." Journal of Political Economy , 97(4): 905-927 Ruel, Marie T., Lawrence Haddad and James L. Garrett. 1999. "Some Urban Facts of Life: Implications for Research and Policy," World Development, Vol 27. No. 11, 1917-1938. Ruel, M.T., L. Haddad and J. L. Garrett. 1999. "Are Urban Poverty and Undernutrition Growing: Some Newly Assembled Evidence," World Development, Vol 27, No. 11, pp.1891-1905. Save the Children (UK). 1999. Ho Chi Minh City: A Participatory Poverty Assessment. Hanoi: World Bank. de Soto, Hernando. 2000. The Mystery of Capital : Why Capitalism Triumphs in the West and Fails Everywhere Else. New York: Basic Books. Stephens, Carolyn, Marco Akerman, Sebastian Avle, Paulo Borlina Maia, Paulo Campanario, Ben Doe and Doris Tetteh. 1997. "Urban Equity and Urban Health : Using Existing Data to Understand Inequalities in Health and Environment in Accra, Ghana and Sao Paulo, Brazil". Environment and Urbanization, Vol. 9, No. 1, April. Strassman, P. and A. Blunt. 1993. "Land, Income, Mobility and Housing, The Case of Metro Manila." Journal of Philippine Development, 36, XX, No. 1. Surbakti, Panjung. 1995 "Indonesia's National Socio-Economic Survey." Jakarta, Indonesia: Central Bureau of Statistics. Suryahadi, Asep, Sudarno Sumarto, Yusuf Suharso, and Lant Pritchett. 2000. "The Evolution of Poverty during the Crisis in Indonesia, 1996-1999 (Using Full Susenas Sample)." SMERU Working Paper, revised March. 49 Thomas, D. 2001. "Education in Crisis," Presentation at the Economic Trade and Development Workshop, Yale University. UNCHS (Habitat). 2001a. Cities in a Globalizing World: Global Report on Human Settlements 2001. United Nations Center for Human Settlements (Habitat). London: Earthscan Publications Ltd. UNCHS (Habitat). 2001b. The State of the World's Cities 2001. Nairobi, Kenya. UN Habitat. 2001c. The Global Urban Observatory Databases (CD-ROM). http://www.unhabitat.org/programmes/ guo/guo_databases.asp United Nations. 2000. World Urbanization Prospects: The 1999 Revision. Part 1: Urban and Rural Areas (POP/ DB/WUP/Rev.1999/1). United Nations Secretariat, Population Division of the Department of Economic and Social Affairs. Data set in digital form. Government of Vietnam-Donor-NGO Poverty Working Group: 1999. Joint Report Vietnam: Attacking Poverty, Vietnam Development Report 2000, December. Webster, Douglas R. 2001. "On the Edge: Shaping the Future of Peri-Urban EastAsia," Discussion Paper prepared for East Asia Region, Urban and Water Division, World Bank. June draft. Wegelin, Emiel. 2001. "Urban Sector Profile for Indonesia." Asian Development Bank. Draft. The World Bank. 2002. World Development Report 2003: Sustainable Development in a Dynamic Economy. The World Bank and Oxford University Press. The World Bank. 2001 a. "Vietnam Living Standards Survey (VLSS) 1997-98." Poverty and Human Resources Division. April. The World Bank. 2001b. World Development Report 2000/2001:Attacking Poverty. New York: Oxford University Press. The World Bank. 2001c. "World Bank Project Appraisal Document for Metro Manila Urban Transport Integration Project," Report No. 20767-PH. Transport Sector Unit, East Asia and Pacific Region, May 23. The World Bank. 2000 a. Philippines Poverty Assessment. Poverty Reduction and Economic Management Unit, East Asia and Pacific Region, Report No. 20498-PH. The World Bank. June 14. The World Bank. 2000 b. "Poverty Reduction in Indonesia: Constructing a New Strategy." Poverty Assessment, The World Bank, September draft. The World Bank. 2000 c. World Development Indicators (WDI). The World Bank. 1999 a. Gender Dimensions of the East Asia Crisis. Washington, D.C. The World Bank. 1999 b. Report No. 18748-VN. Vietnam Moving Forward:Achievements and Challenges in the Transport Sector, Transport Sector Unit, East Asia and Pacific Regional Office. The World Bank. 1993. "Indonesia: Urban Public Infrastructure Services." Report No. 12154-IND. Washington, DC. R e f e r e n c e s 50 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam ANNEX ONE: PROPOSED IMPROVEMENTS IN SAMPLING AND SURVEY DESIGN FOR ASSESSMENT OF URBAN POVERTY 1. We know much less about the urban poor than we know about the rural poor. The primary reason for this is that the analysis of risk and poverty has focused more on rural areas because in most low-income developing countries, the preponderance of the poor are in rural areas. As demonstrated in this review, however, that is rapidly changing; poverty is increasingly acquiring an urban face. But the rural analytical bias has resulted in a limited understanding of the urban experience of poverty. The bias results both from sampling design problems and from the construction of survey instruments--both tending to be based upon rural notions of life. This Annex first outlines the nature of the problems and then sketches possible solutions to them. 2. Poverty estimates are usually calculated from surveys of large, nationally Sampling Design representative samples. These samples are typically drawn from sampling frames based upon decennial censuses (e.g., Surbakti, 1995; World Bank 2001). In rural areas that is The Problem usually not a problem. Rural populations tend to be stable, with families that have lived several generations in the same village on ancestral land with well defined addresses and property rights. Family structures also are relatively stable, and mobility is low. Consequently a ten-year update on the sampling frame is adequate to track important changes in the size and composition of the rural population. Since the populations are relatively stable, samples drawn in subsequent years from this sampling frame are generally unbiased and representative. 3. The urban situation is quite different. Urban populations, particularly the urban poor, are characterized by high levels of mobility, unstable and poorly defined property rights, and fluid family and social structures. The poorest of the poor, in particular, often are squatters in temporary settlements subject to a high risks of demolition. Thus, the poorest are particularly mobile. Census-based sampling frames are consequently out of date almost by the moment that they are tabulated. Squatters are the most likely to be excluded from sampling frames, because of their poorly defined addresses and the high probability of changes in residence. Individuals who are often homeless and live on the streets are also unlikely to be captured within the sampling frame. Consequently a sample constructed on the basis of a census-based sampling frame is very likely to miss the poorest among the urban poor and to underestimate the extent of urban poverty. 4. Furthermore, urban homes often have multiple families resident in them, since new migrants typically stay with friends and relatives while they find a place to live. Thus, even if a household is picked up in the census, it is likely that there are now two 51 or more households resident in the same space, once again resulting in an undercount of the poor in the sample. 5. The problem is compounded in those countries where governments require citizens to register in order to be considered "legitimate" residents of a town or village. New migrants are likely to be unregistered, which leads to their automatic exclusion from censuses and from the samples based on them. This problem is widespread in East Asia, with both the Vietnam LSMS (World Bank, 2001) and the China Household Survey (China National Bureau of Statistics, 1999) reporting that their urban samples suffer from the problem. One Possible 6. Urban sampling frames thus need much more frequent updating than rural sampling Solution frames do. While it would be expensive to conduct a complete household listing before conducting a household survey, various statistically valid procedures can be employed to reduce the extent of bias. One possibility is to conduct a stratified random sample where sectors within a city are first selected--employing census-based information on population and housing to construct a community-level sampling frame. 7. Then the sampled neighborhoods can be mapped to identify new squatter settlements, and to incorporate changes in the size and structure of neighborhoods. Following that, a PRA (participatory rapid appraisal) or PPA (participatory poverty assessment) exercise can be conducted within those neighborhoods to obtain rough estimates of population size, housing quality, ethnic composition and other such broad indicators that will help determine the next stage of sampling and stratification. Once such information is obtained, the samples can be collected by randomly choosing neighborhoods from the updated list stratified by appropriate criteria. With the sample of neighborhoods chosen, households within them can be systematically listed to obtain a complete sampling frame for the subset of neighborhoods included in the multi-stage sample. Households can then be randomly chosen from this sampling frame. A sample conducted on the basis of this method is less likely to be biased against new migrants and the poor. 8. The method outlined above is just one way to improve the coverage of the urban poor in sample surveys. Although there may also be other methods, any technique employed will have to tackle the problem of updating the sampling frame to include transients and new migrants, and multiple households. The problem of the homeless is more difficult to deal with. To begin with, it may be important to conduct a case-study of homelessness in major metropolitan areas, to understand the magnitude of the problem, and then to develop ways of including the homeless in surveys of the urban poor. Survey 9. The experience of poverty in urban areas is quite different from that in rural areas. Instruments Theriskssufferedbytheurbanpooraresomewhatunique. Forinstance,theyareplagued by insecure property rights in housing and must constantly cope with the possibility of The Problem eviction or demolition. While climatic risks may dominate the lives of agricultural households, the urban poor are more likely to face macro-economic shocks that directly threaten urban jobs. This is particularly true of East Asian countries because of the recent economic downturn. 10. While it is well-known that occupations of the urban poor often are in the informal economy, it is less well-known that this sector can be highly organized. For instance, the sale of cooked food is a disciplined occupation that depends crucially on social ties and in inter-connections among grocers, cooks, and sellers (Jellinek, 1991). Occupational structures Annexes 52 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam urban areas are therefore very different from those in rural areas, and the risks they face not very well understood. For instance, garbage pickers dependent upon recycling plastic bags found in the trash may lose a lot of income if new environmental regulations banning plastic bags are instituted. The peculiar natures of urban occupations are typically not captured by national surveys, whose questions on income and occupation are usually based on traditional agricultural and salaried categories. 11. Urban social ties are also different from rural ones. Urban social groups sometimes import pre-existing relationships from the village, but more often urban social relationships are based upon unconventional connections peculiar to urban communities where individuals from diverse ethnic and economic backgrounds may live together in the same space. It is therefore not uncommon to observe marriage patterns in urban areas that are very different from those in rural areas. Typically, the lack of an extended family also makes young urban families much more dependent upon their own initiative, rather than on the endowments of social and cultural capital they obtain from their parents. All such differences affect how urban households cope with risk: their coping mechanisms depend much more on friends, co-workers, and employers than on close relatives as is the case in rural areas. 12. The life of urban migrants nonetheless remains closely linked to the families they leave behind in their villages. The migrant is often the rural family's most important source of support, and the rural family in turn may help with food and grain during crisis, and with marriage, pregnancy and other life events. The inter-connectedness between the rural and urban branches of a family is important enough that it is difficult to argue for a clear distinction between the rural and the urban poor. Rural and urban poverty are closely connected, and the inter-connectedness should be better taken into account when formulating policy, which tends to see rural and urban populations as distinct. Urban migration offers rural households a ticket to social and occupational mobility, so many urban migrants therefore spend as much effort on trying to "get ahead" as they do on "survival." Urban social networks are thus as geared to finding earning opportunities as they are to social protection strategies for the poor. Urban households, to coin a phrase, are the branch offices of their rural brethren. 13. National surveys typically do not take into account the complexities of urban life. The substantial differences between rural and urban households in their occupational composition, sources of risk, and the nature of their networks call for very different survey instruments. On the other hand, the close links between rural and urban households require a more integrated approach to survey design. Besides that, the rural experience of poverty is usually well covered in questionnaires, with several questions on agricultural occupations, investments and strategies for coping with climate-related shocks, but the urban poor often do not have their experience of poverty well measured or understood. 14. While qualitative and case-study research has substantially expanded our knowledge Possible Solutions of urban poverty, the nature of risks, social networks, and occupations may differ substantially from country to country and even from region to region. To develop survey instruments that can uncover some of the complex social and economic relationships outlined above, we need a method that first uncovers the full range of phenomena experienced by the urban poor, and then tries to formulate structured questions that can measure them in the context of a survey instrument. 15. One method that can be employed is to combine qualitative and quantitative techniques (Rao, 2001). Once the sampling frame has been constructed on the basis of the multi-stage method outlined above, a sub-sample of communities selected for the quantitative questionnaire can be the focus of an intensive qualitative investigation. 53 That would combine focus group discussions, Participatory Poverty Assessments at the community level, and in-depth interviews at the level of households. Community-level investigations should be conducted with sub-groups stratified into relevant categories: men and women, working age, students and the elderly. Participants in these discussions should, as far as possible, be randomly selected. Simultaneously, a small random sample of households can be given open-ended, in-depth interviews. The subject matter of the focus group discussions, PPAs and in-depth interviews should be kept as general as possible in order to allow new findings to be encountered and digested by the research team. Structured questionnaires often presuppose considerable knowledge on the part of the researcher, but when the subject matter--like the experience of urban poverty--is relatively less understood, it is best to allow the interviewee( in this case the poor) take the lead in informing researchers about the problems that they confront, so that the process becomes more consultative. This qualitative information should be recorded and ideally, the team that is conducting the interviews should meet at the end of every day to cross-check and dissect their findings. 16. For instance, the urban poor engage in occupations that are not always easy to predict or classify, such as organized panhandling or street performance or recycling discarded copper wire. Standard questionnaires would classify these either as "other" or as "self-employed." However, neither of those categories does justice to the complexity of the occupation, which may be quite formally organized and more in the nature of piece-work than self-employment. Consultative questionnaire development could obtain a comprehensive list of the types of occupations in the slum. A focus group would be very helpful in classifying these as wage-related or self-employed. The focus group's information could then be incorporated into the questionnaire, and it would better reflect the reality of occupational patterns among the urban poor, rather than preconceptions about what they might be. 17. Another example of the usefulness of this technique is to understand the role of networks--who do people seek help from, and for what purpose? Do new migrants access a different set of networks than older migrants do? How do urban households interact with their rural families? How do they interact with the slum leadership and with agents of the State? Qualitative work can generate insights into how those inquiries are best incorporated into the questionnaire--whether they should have a separate module, for instance, or be folded into modules on labor supply and housing. 18. Incorporating the qualitative information into the questionnaire is best done by bringing in the views of all the field staff, soon after the end of the qualitative field work. That may last from a week to several months, depending upon the size and experience of the team, and the complexity of the field site. The team should then have a conference to devise a structured quantitative questionnaire that emerges directly from the field experience. One way of doing this efficiently is to begin with a prototype survey instrument such as the LSMS or some other well-known household survey. Then each module in the questionnaire can be analyzed by the field team to examine how closely it fits what they learned in the field. The facilitator can direct the discussions towards specific structured questions that can be used to modify the prototype module. This process can develop an entirely new questionnaire informed by the field team's contributions. Once the new questionnaire is developed, it must be pre-tested and refined before being administered. 19. This mixed method approach is one solution for developing survey instruments that are attuned to the experience of urban poverty. It may mean that national surveys would have three modules--one that is common to all communities in the sample, and one each for rural and urban households, which would include questions on the inter- connections between rural and urban life. Annexes 54 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam ANNEX TWO: SELECTED INDICATORS Table A1. Selected economic and poverty indicators for cities in East Asia, South Asia and Latin America: 1998 57List of Tables Table A2. Gross educational enrollment (%) and literacy (%), male and female, in cities in East Asia, South Asia 58 and Latin America: 1998 Table A3. Comparative housing sector indicators for cities in East Asia, South Asia and Latin America: 1998 59 Table A4. Infrastructure service indicators for cities in East Asia, South Asia, and Latin America: 1998 60 Table A5. Average work-trip times and transport modes (%) for cities in East Asia, South Asia 61 and Latin America: 1998 Table IN1. INDONESIA ­ Food expenditure as a ratio of total household expenditure by per capita expenditure 62 Indonesia deciles, for all rural population and for urban settlements: 1999 Table IN2. INDONESIA - Household demographics (%) for rural and for urban settlements, distributions across 62 per capita expenditure deciles: 1999 Table IN3. INDONESIA ­ Rural and urban educational levels (%), within and across per capita 63 expenditure deciles: 1999 Table IN4. INDONESIA - School participation (%) of children 6-14 years old, for all rural locality and for urban 64 localities by size, per capita expenditure deciles: 1999 Table IN5. INDONESIA - Literacy rates (%) for all respondents age 5 years and older, for rural locality and for 65 urban localities by size, by per capita expenditure deciles: 1999 Table IN6. INDONESIA - Working status by gender for non-poor and poor (%) in rural localities and in urban 66 localities by size: 1999 Table IN7. INDONESIA - Employment status for rural localities and for urban localities by size, by gender and 66 poverty status (%): 1999 Table IN8. INDONESIA - Housing characteristics by poverty status (%), for rural and for urban settlements: 1999 67 Table IN9. INDONESIA - Housing characteristics (%) for rural and for urban settlements, by per capita 68 expenditure deciles: 1999 Table IN10. INDONESIA - Sources of drinking water (%) for rural and for urban settlements, by per capita 69 expenditure decile: 1999 Table IN11. INDONESIA ­ Sources of drinking water for rural localities and for urban localities by size, 70 by poverty status: 1999 Table IN12. INDONESIA - Sanitation characteristics (%) for rural and for urban settlements, by per capita 71 expenditure deciles: 1999 Table IN13. INDONESIA - Sanitation characteristics (%) for rural settlements and for urban settlements by size, 72 by poverty status: 1999 Table PH1. PHILIPPINES ­ Distribution of household characteristics (%) within per capita expenditure deciles, 73 The Philippines rural/urban: 1997 Table PH2. PHILIPPINES ­ Household characteristics (%) across per capita expenditure deciles, 74 rural/urban: 1997 Table PH3. PHILIPPINES ­ Distribution of employment status (%) within per capita expenditure deciles, 75 and across deciles, rural/urban: 1997 Table PH4. MANILA ­ Employment status (%), by per capita expenditure deciles: 1997 75 Table PH5. PHILIPPINES ­ Housing tenure status (%), within per capita expenditure deciles, rural/urban: 1997 76 Table PH6. PHILIPPINES ­ Type of housing (%), by per capita expenditure deciles, rural/urban: 1997 77 Table PH7. MANILA ­ Type of housing (%) by per capita expenditure deciles, rural/urban: 1997 78 55 Table PH8. PHILIPPINES ­ Type of roof (%) by per capita expenditure deciles, rural/urban: 1997 79 Table PH9. PHILIPPINES ­ Type of walls (%) by per capita expenditure deciles, rural/urban: 1997 80 Table PH10. MANILA ­ Type of roof (%) by per capita expenditure deciles, rural/urban: 1997 81 Table PH11. MANILA ­ Housing tenure status (%), by per capita expenditure decile: 1997 82 Table PH12. PHILIPPINES ­ Source of water and electricity, by per capita expenditure decile, rural/urban, 1997 83 Table PH13. PHILIPPINES ­ Consumer durables (%), by poverty status, rural/urban, 1997 83 Table PH14. MANILA ­ Source of water and electricity, by per capita expenditure decile, 1997 84 Table PH15. PHILIPPINES ­ Sources of water and electricity (%) for each poverty status, rural/urban: 1997 85 Table PH16. PHILIPPINES ­ Type of toilet (%), by per capita expenditure decile, rural/urban: 1997 86 Table PH17. PHILIPPINES ­ Type of toilet (%), by poverty status, rural/ urban: 1997 87 Table PH18. MANILA ­ Type of toilet (%), by per capita expenditure decile: 1997 87 Table PH19. MANILA ­ Households' consumer durables, by per capita expenditure decile: 1997 87 Vietnam Table VN1. VIETNAM ­ Household characteristics (%) by per capita expenditure decile, rural/urban: 1998 88 Table VN2. VIETNAM ­ Rural and urban household characteristics (%), by per capita expenditure decile and 89 poverty status: 1998 Table VN3. VIETNAM ­ Rural and urban ethnicity (%), by per capita expenditure decile: 1998 90 Table VN4. VIETNAM ­ Individual characteristics (%), by poverty status, rural/urban: 1998 90 Table VN5. VIETNAM ­ Individual literacy levels and gender (%), by per capita expenditure decile, 91 rural/urban: 1998 Table VN6. VIETNAM ­ Health status by poverty status, rural/urban: 1998 92 Table VN7. VIETNAM ­ Female and male smoking behaviors (%), by Poverty Status, rural/urban: 1998 92 Table VN8: VIETNAM ­ Types of housing: percentages within each per capita expenditure decile, and mean 93 percentages for all deciles, rural/urban: 1998 Table VN9. VIETNAM ­ Types of housing: percentages by per capita expenditure decile, and mean percentage 93 for all types in each decile: 1998 Table VN10. VIETNAM ­ For rural and for urban localities, types of housing (%) for non-poor 94 and for poor households, and means for both localities: 1998 Table VN11. VIETNAM ­ Types of walls (%) for each per capita expenditure decile, and mean percentages for 95 total deciles, rural/urban: 1998 Table VN12. VIETNAM ­ Types of roofs (%) for each per capita expenditure decile, and mean percentages for total 95 deciles, rural/urban: 1998 Table VN13. VIETNAM ­ Household characteristics, by poverty status, rural/urban: 1998 96 Table VN14. VIETNAM ­ Types of lighting (%) by per capita expenditure decile, rural/urban: 1998 97 Table VN15. VIETNAM ­ Sources of drinking water (%) for each poverty status, rural/urban: 1998 97 Table VN16. VIETNAM ­ Sources of drinking water (%) within each per capita expenditure decile, rural/urban: 1998 98 Table VN17. VIETNAM ­ Types of toilets (%) within each per capita expenditure decile, rural/urban: 1998 98 Table VN18. VIETNAM ­ Type of toilet (%) by poverty status, rural/urban: 1998 99 Explanatory Note to These tables contain detailed data from the national household surveys for Indonesia, the Philippines and Vietnam Tables IN1 through as described in the text. All data are percentages. VN18 Per capita expenditure deciles range from 1 (poorest) to 10 (richest). The classification of Poor and Non-Poor households is based on poverty lines estimated by the surveys, separately for the rural and the urban populations. The tables show: (i) distribution within per capita expenditure deciles (column data for each set of variables add to 100 percent), or (ii) distribution across per capita expenditure deciles (row data for each variable add to 100 percent) for rural and for urban household samples. Annexes 56 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam s tor 1998 ictimsV /000 0.17 0.16 .000 0.02 0.01 0.10 0.24 0.01 0.05 0.04 0.80 0.15 0.03 11.0 homicide 20.40 0.23 0.08 data indica of 1993 b . ty erv America: po tin La latoTfo pop. eht sector preliminary ni and % .003 are 70.5 0.0 15.4 29.7 14.4 8.8 .071 7.6 0.36 34.2 .096 10.3 .035 31.3 and data a employed informal Asia economic ) South %( 5.3 0.21 .001 10.2 4.7 5.4 3.2 1.9 4.3 .011 0.8 8.9 7.8 4.7 4.4 0.6 8.7 0.4 0. 50.3 23 1.31 15.4 10.2 Unemployment Selected Asia, A1. East lebaT in below line 16.4 2.0 6.6 24.8 0.9 0.0 0 0 19.0 6.1 30. 4.0 1.5 2.1 1.1 15.9 9.73 2.1 10.6 44.3 18.6 20. .081 4.4 .084 15.4 cities locally-defined poverty orf Households % the DC. a $) 756 991 084 490 verageA income (in 3,584 1,366 1,167 1,077 3,193 27,047 19,933 19,933 23,500 8,521 3,384 1,920 13,026 5,391 3,179 14,748 household ashington,W an Bank, .0 .0 area (000) 999.8 562 0.726 0.421 .0 .0 .0 b 0 .0 0. 499 orldW 1,806.4 9,489.4 1,076.2 2,373.3 Population 25,130 3,691.9 2,188.9 3,163.5 3,843.0 10,392 5,647.0 5,063.0 10,000 4,472. 4,600 2,996 2,317.4 7,431.0 1,669.5 metropolit data). 2001, and ar (19982 2000 Penh Mai a Minh Aires Chi City Phnom Bandung Jakart Semarang Surabaya okyoT ientianeV Penang Ulaanbaat angonY Database Cebu Singapore Hanam Pusan Seoul Bangkok Chiang Hanoi Ho Dhaka Bangalore Pokhara Colombo Buenos Guayaquil Lima Montevideo Database, Indicators Korea Indicators Country Cambodia Indonesia napaJ PDR al Lanka Urban Lao Malaysia Mongolia Myanmar Philippines Singapore South Thailand ietnamV Bangladesh India Nep Sri Argentina Ecuador Peru Uruguay Global Development UNCHS orldW Asia Asia America Region East South Latin Source: 57 (%), y 1998 ela M 79.5 99.2 97.5 ac 100.0 92.2 91.0 99.1 90.6 96.8 98.4 93.0 95.1 95.1 60.3 66.2 98.2 98.6 (%) liter Rate America: and tin (%) La Literacy 57.0 97.3 91.4 100.0 78.9 82.0 97.1 88.7 89.2 95.1 90.0 89.0 89.5 60.3 42.0 97.8 98.3 Female and ollment Asia enr South (%) 45.2 64.7 53.3 99.9 99.8 99.9 65.9 26.6 68.6 93.9 tional (male) 12.66 100.0 100.0 Secondary Asia, educa Enrollment East ossrG in Gross 8.26 95.7 92.9 100.0 54.9 74.2 57.7 100.0 99.9 99.7 99.9 62.3 35.4 75.3 100.0 (female) Secondary A2. cities lebaT in, emalef data). Penh Mai Minh and (19982 Chi City Phnom Jakarta Semarang okyoT Vientiane Penang Ulaanbaatar angonY Singapore Hanam Pusan Seoul Bangkok Chiang Hanoi Ho Dhaka Pokhara Guayaquil Montevideo male Database Indicators Korea PDR Urban Country Cambodia Indonesia Japan Lao Malaysia Mongolia Myanmar Singapore South Thailand ietnamV Bangladesh Nepal Ecuador Uruguay Global UNCHS Asia Asia America Region East South Latin Source: Annexes 58 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam cities 1998 ratio orf % 19.0 2.5 10.0 4.9 15.4 2.0 13.9 22.2 25.0 34.0 16.1 31.1 s Rent-to- income tor America: tin indica La and price- sector % 8.9 7.6 14.6 3.4 5.6 23.2 7.2 7.8 8.3 13.3 3.1 3.7 4.0 5.7 8.8 6.8 16.7 13.8 21.6 5.1 3.4 8.7 5.6 Asia to-income House housing evtia South Asia, Other 61.8 4.5 19.8 0.4 0.4 0.0 1.11 0.0 0.0 0.0 1.4 0.5 12.5 1.0 13.0 2.2 5.2 0.6 10.7 East Compar in A3. lebaT %% 16.4 0.0 0.0 2.0 7.4 48.4 1.5 5.0 0.0 0.0 0.0 17.9 3.0 5.0 8.0 14.0 0.0 5.7 49.0 18.8 1.3 Squatters ypesT enureT 0.0 enancyT 34.1 16.2 31.0 6.6 33.4 37.4 9.0 6.6 56.0 58.9 41.4 19.5 30.0 17.2 9.0 29.0 22.5 17.4 23.4 21.8 %% Formal 21.8 61.4 64.0 66.6 85.6 18.2 50.0 86.0 93.4 44.0 39.7 35.8 77.0 62.3 76.0 58.0 69.6 28.4 57.2 66.2 ownership data). (19982 Penh Mai Aires City Phnom Bandung Jakarta Semarang Surabaya okyoT Vientiane Penang Ulaanbaatar angonY Cebu Singapore Hanam Pusan Seoul Bangkok Chiang Hanoi Dhaka Bangalore Pokhara Colombo Buenos Guayaquil Lima Montevideo Database Indicators Korea PDR Lanka Urban Country Cambodia Indonesia Japan Lao Malaysia Mongolia Myanmar Philippines Singapore South Thailand ietnamV Bangladesh India Nepal Sri Argentina Ecuador Peru Uruguay Global UNCHS Asia Asia America Region East South Latin Source: 59 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 cities 1998 6.0 5.1 other 22.3 25.7 3.0 1.0 20.0 15.0 24.7 4.2 orf s 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5.0 tor America: burned openly 16.3 7.4 1.2 2.0 (%) tin indica La re- 0 0 0 0 14 0 0 0 0 15 0 0 0 cycled 15.0 30.0 10.3 10.0 30.0 44.3 38.0 35.0 14.5 15.9 0.3 7.0 vice and Disposal 0 0 0 0 0 0 0 0 0 0 0 0 0 ser aste 0.1 0.3 43 e Asia, W open dump 74.0 80.0 90.0 86.0 50.0 60.8 76.7 100.0 99.8 Solid uctur 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 South inci- eratedn 78.0 10.0 5.0 66.3 3.0 14.5 5.0 2.0 0.2 astr Infr Asia, sani- yrat 0 0 0 0 0 0 0 0 78.6 77.7 74.3 70.0 8.6 5.0 landfill 100.0 33.7 67.0 41.2 57.0 99.0 98.0 65.0 100.0 94.0 57.0 A4. East lebaT in tele- 40.0 70.8 99.2 86.8 98.0 90.0 17.3 25.0 7.0 phone 100.0 100.0 100.0 59.7 75.0 60.0 21.2 44.5 1.31 26.0 70.4 44.0 75.1 (%) city 75.5 99.0 99.0 85.2 89.2 100.0 electri- 100.0 100.0 100.0 85.0 80.0 100.0 100.0 100.0 100.0 99.8 100.0 100.0 99.7 90.0 98.3 75.0 96.0 100.0 99.0 99.0 99.7 ega 74.9 55.0 64.8 55.8 60.0 81.2 92.3 67.9 69.4 98.6 60.0 50.0 30.0 22.0 20.8 76.0 98.1 42.0 71.5 79.1 Connections sewer- 100.0 100.0 100.0 water 44.7 50.3 34.0 40.9 100.0 87.0 99.1 60.0 77.8 41.4 100.0 81.1 97.9 99.9 99.0 95.0 70.0 59.0 60.0 22.5 41.4 22.9 100.0 70.0 75.2 97.6 200m access water % 85.4 90.0 91.4 89.7 94.3 to 100.0 95.0 99.9 90.3 95.0 98.2 100.0 81.1 99.9 99.9 99.0 100.0 90.0 99.1 82.9 80.0 100.0 77.0 81.1 99.3 data). Households with within (19982 Penh Mai Minh Aires Database Chi City Phnom Bandung Jakarta Semarang Surabaya okyoT Vientiane Penang Ulaanbaatar angonY Cebu Singapore Hanam Pusan Seoul Bangkok Chiang Hanoi Ho Dhaka Bangalore Pokhara Colombo Buenos Guayaquil Lima Montevideo Indicators Korea PDR Lanka Urban Country Cambodia Indonesia Japan Lao Malaysia Mongolia Myanmar Philippines Singapore South Thailand ietnamV Bangladesh India Nepal Sri Argentina Ecuador Peru Uruguay Global UNCHS Asia Asia America Region East South Latin Source: Annexes 60 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam (%) 1998 ) 0 2.2 3.0 8.8 0.9 3.2 0.9 bicycle/ other %( 12.5 56.1 10.0 14.7 21.7 23.8 13.3 33.6 24.0 86.2 24.7 75.0 13.5 walking/ modest America: ansportr tin ) 5 2 2 La %( 0.2 17.8 2.1 72 55.0 59.0 65.0 38.7 32.5 38.8 9.2 35.7 14.0 65.0 42.2 89.3 82.2 59.6 and and bus/minibus Modes times Asia ransportT ) 0 0 0 0 1 0 0 0 0 0 0 0 0 0 k-trip South %( 21.0 3.7 14.5 6.6 32.3 8.1 16.4 train/tram orw ega Asia, ervA East ) in %( 87.3 82.0 80.0 41.8 42.0 10.0 16.7 25.1 37.1 20.1 58.7 94.1 64.4 74.0 4.6 39.6 1.01 23.7 33.5 10.7 16.9 26.9 private motorized A5. lebaT cities orf time work- (mn) 45 30 35 54 72 40 30 45 35 30 42 60 60 03 03 52 45 30 20 52 24 45 54 verageA travel per trip data). (19982 Penh Mai Minh Aires Chi Database City Phnom Bandung Surabaya okyoT ientianeV Penang Ulaanbaatar angonY Cebu Singapore Pusan Seoul Bangkok Chiang Hanoi Ho Dhaka Bangalore Pokhara Colombo Buenos Guayaquil Lima Montevideo Indicators Korea PDR Lanka Urban Country Cambodia Indonesia Japan Lao Malaysia Mongolia Myanmar Philippines Singapore South Thailand ietnamV Bangladesh India Nepal Sri Argentina Ecuador Peru Uruguay Global UNCHS Asia Asia America Region East South Latin Source: 61 e , * 1999 1999 001 001 001 001 001 001 001 001 001 001 001 001 001 001 001 001 Mean Mean* 0.747 0.69 0.66 0.639 0.62 0.56 01 xpenditure 15.2 9.1 13.5 26.6 9.5 7.0 9.4 9.7 9.7 7.7 7.7 17.7 25.4 2.4 13.1 5.4 settlements 01 deciles: 9.5 9.4 8.1 9.8 7.8 3.9 7.7 0.662 0.561 0.52 0.502 0.518 0.45 e 13.0 14.9 18.3 10.4 10.5 10.4 14.9 18.0 13.6 settlements: urban 1.61 9.7 18.9 1.41 10.2 9.3 10.0 10.3 10.2 12.2 9.9 13.4 14.4 4.9 12.8 8.7 household orf Deciles urban 10.7 9.9 13.5 10.0 10.4 9.3 10.1 9.7 1.51 12.6 10.1 1.71 1.11 6.0 12.5 9.8 0.723 0.627 0.578 0.563 0.572 0.543 total orf and xpenditure 9.4 al 10.1 13.5 8.1 10.4 10.2 9.8 10.3 10.0 12.6 10.5 10.7 8.5 7.9 10.6 10.7 of and urr pita Urban tioar ca 8.6 10.2 5.4 6.9 10.5 10.6 9.7 9.9 10.3 12.3 9.8 8.3 7.1 9.3 10.4 1.11 0.737 0.653 0.614 0.612 0.585 0.554 tion orf Expenditurea a per 8.4 8.1 5.5 9.5 7.9 5.5 8.9 (%) Capit 10.3 10.2 10.7 10.5 10.4 1.41 1.01 10.7 1.81 as 1 oss Per 7.5 5.4 5.3 9.7 9.3 9.6 5.8 4.4 7.9 e popula Deciles 3456789 10.5 10.2 1.31 10.3 1.01 13.4 12.0 0.748 0.672 0.642 0.615 0.61 0.569 al urr phicsarg acr 2 7.5 10.4 4.1 4.5 10.4 1.41 9.9 9.9 9.2 7.3 1.51 5.1 3.3 17.0 6.1 12.0 xpenditure all utions 1 8.1 10.3 2.7 3.5 8.9 12.2 10.5 9.9 9.6 3.8 10.8 4.4 2.2 24.6 4.0 10.7 oodF­ orf, Expenditurea 0.753 0.686 0.662 0.651 0.629 0.567 demo * Capit distrib 001 001 001 001 001 001 001 001 001 001 001 001 001 001 001 001 Mean Per deciles 0.757 0.703 0.676 0.663 0.627 0.59 01 15.0 9.3 18.2 15.4 9.7 8.4 9.9 1.31 1.01 1.41 8.3 20.7 24.0 2.7 14.2 6.4 e Household- 13.2 9.6 16.7 12.8 9.4 8.5 10.0 1.71 12.1 13.0 10.3 18.3 17.5 3.9 12.8 8.0 INDONESIA 0.764 0.71 0.686 0.679 0.646 0.592 1.61 9.8 18.2 1.41 10.0 8.8 9.9 1.11 1.41 14.0 10.5 14.4 13.3 5.1 12.8 9.2 xpenditure Deciles 10.6 9.9 12.1 10.2 9.8 9.5 9.9 10.5 1.11 12.6 1.41 12.2 10.6 6.7 12.1 10.1 IN1. INDONESIA lebaT pitaca 9.3 0.77 0.723 0.701 0.685 0.658 0.608 10.1 7.6 9.6 10.2 9.8 9.8 10.1 10.6 1.21 1.11 8.8 9.3 8.3 10.3 10.9 Rural IN2. 4.5 7.8 7.9 9.9 per yb lebaT Expenditurea 8.6 10.2 9.5 10.3 10.1 9.9 9.6 10.1 10.0 10.4 9.7 1.21 456789 8.5 3.0 8.5 9.9 9.8 9.1 9.1 6.3 6.4 9.4 0.772 0.733 0.712 0.699 0.67 0.619 Capit 10.2 10.1 10.7 10.8 1.51 1.31 Per 3 7.5 10.4 10.6 8.7 10.4 10.7 10.1 9.0 8.9 8.1 9.4 5.3 5.0 13.6 7.9 1.51 123456789 2 7.3 0.775 0.746 0.729 0.712 0.704 0.634 10.4 6.1 7.3 10.2 1.51 10.3 8.6 8.0 6.2 9.4 3.4 3.6 16.6 6.4 1.51 deciles 1 8.3 groups all 10.2 3.0 6.5 9.9 12.1 10.3 8.4 7.6 4.2 8.6 2.9 2.5 22.0 4.3 10.0 all sampled, sampled, household ­999K headship household children families head old old old old old of Settlement population old 3+ adults population of <250th 250-499K 500K 1M-2M >2M the household family total of years years years years years years of parent adult with of types total ypeT Rural Urban Urban Urban Urban Urban for* Household Female-headed Male-headed Age <15 15-25 26-35 36-45 46-55 56-65 +56 ypeT Nuclear Single Elderly Single Family Family Other for* Annexes 62 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam * * (%), 1999 Mean 24.4 25.2 19.0 18.4 6.9 0.8 1.9 3.3 0.1 001 001 001 001 001 001 001 001 001 Mean elsvle 01 1.31 14.7 16.8 30.4 7.4 1.6 5.4 1.71 0.8 01 3.3 4.2 6.3 1.81 7.7 13.8 20.7 25.4 41.4 deciles: 8.5 1.5 4.0 7.2 0.3 5.0 6.0 8.2 tional e 14.6 18.0 18.5 27.2 12.4 10.4 15.5 18.2 18.4 20.1 17.2 19.7 19.8 24.3 8.9 1.3 3.2 5.5 0.2 6.3 6.9 9.2 1.71 1.51 14.4 15.1 14.7 14.2 educa Deciles xpenditure 19.4 22.5 20.5 21.9 8.3 1.1 2.4 3.7 0.1 Deciles 7.5 8.4 10.2 1.21 1.41 12.1 12.3 10.6 8.1 urban 21.5 24.4 20.3 19.9 7.9 0.9 1.9 3.2 0.1 8.7 9.6 10.6 10.7 1.41 10.7 9.9 9.6 6.5 pita Urban Urban and ca Expenditurea 23.9 26.3 20.0 17.8 7.5 0.8 1.4 2.2 0.0 Expenditurea 10.1 10.7 10.8 10.0 1.31 10.4 7.4 7.0 3.2 al urR­ per 456789 Capit 25.6 27.6 20.3 16.1 7.0 0.6 1.0 1.8 0.0 Capit 1.21 1.61 1.31 9.3 10.8 8.1 5.6 5.7 2.6 oss Per 3 28.7 28.9 19.2 14.4 6.1 0.5 0.9 1.3 0.0 Per 3456789 13.1 12.7 1.21 8.7 9.8 7.2 5.2 4.3 2.6 acr Level 2 32.0 30.3 18.8 12.1 5.2 0.3 0.6 0.8 0.0 3.6 2.8 0.6 and Decile 2 15.4 14.1 1.61 7.7 8.9 4.5 Each at 1 INDONESIA 38.0 31.7 16.3 9.4 3.7 0.2 0.3 0.4 0.0 Each 1 19.5 15.7 10.7 6.4 6.8 3.3 1.9 1.6 0.6 age within * in * IN3. Mean 42.1 35.6 12.3 5.9 2.8 0.4 0.3 0.5 0.0 age 001 001 001 001 001 001 001 001 001 Mean lebaT Percent 01 28.0 30.0 16.0 13.0 7.2 1.6 1.5 2.6 0.1 Percent 01 5.1 6.5 10.1 17.0 19.7 30.8 35.3 36.5 55.8 Decile, 34.0 34.2 15.3 9.1 4.8 0.8 0.7 1.0 0.0 6.7 8.0 Level, 10.4 12.9 14.3 17.4 16.9 16.0 3.8 Each 37.1 35.4 14.2 7.7 3.9 0.6 0.4 0.8 0.0 7.9 9.0 Each 10.5 1.71 12.6 12.2 12.0 12.9 9.6 ithin W Deciles 38.8 36.2 13.7 6.8 3.2 0.4 0.3 0.6 0.0 For Deciles 8.8 9.7 10.7 1.01 10.9 10.2 9.6 9.7 9.6 41.0 36.2 12.9 6.1 2.9 0.3 0.2 0.4 0.0 9.7 10.2 10.5 10.3 10.4 7.8 7.0 7.1 1.9 Rural Rural Expenditurea 42.2 36.8 12.2 5.5 2.5 0.3 0.2 0.3 0.0 Expenditurea 10.3 10.6 10.2 9.6 9.1 7.0 6.0 6.4 9.6 456789 Capit 43.9 37.0 1.91 4.6 2.0 0.2 0.2 0.2 0.0 456789 Capit 1.11 1.11 10.3 8.4 7.5 5.4 5.1 4.4 1.9 Per 3 45.8 36.7 1.21 4.1 1.8 0.2 0.1 0.2 0.0 Per 3 12.0 1.41 10.0 7.6 7.0 4.1 4.2 3.5 3.8 deciles 2 49.1 36.3 9.7 3.3 1.3 0.1 0.1 0.1 0.0 2 13.4 1.71 9.1 6.3 5.1 3.3 2.0 2.3 1.9 or 1 52.6 35.5 8.4 2.5 0.8 0.1 0.1 0.1 0.0 1 15.0 12.0 8.2 5.1 3.4 1.8 2.0 1.3 1.9 levels all sampled, JH JH school school I/II III IV population high high school school I/II III IV high high completed completed total Education Not primary Primary Junior ocationalV Senior Diploma Diploma Diploma Master/Ph.D. Education Not primary Primary Junior ocationalV Senior Diploma Diploma Diploma Master/Ph.D. for* 63 old, s 1999 la otT 001 001 001 001 001 001 001 001 001 001 001 001 001 001 001 100U 001 001 eary 6-14 deciles: 01 2.5 4.9 1.9 1.5 3.2 0.8 2.3 3.9 2.2 2.2 4.4 2.9 2.0 5.5 3.4 1.6 5.4 6.8 e en 9 hildrc 4.1 6.2 4.0 3.4 5.1 1.7 4.6 6.4 2.2 4.7 5.6 2.0 4.5 7.4 3.4 3.7 7.8 6.8 of xpenditure (%) pita 8 5.4 7.6 5.1 4.5 6.4 3.5 3.9 7.1 3.7 3.0 6.0 1.0 3.7 7.7 0.7 6.3 7.7 6.8 ca tion ticipa per,e 7 6.8 8.7 6.5 5.0 7.5 4.6 6.5 7.6 4.1 4.7 7.5 4.9 5.1 8.9 2.0 12.0 1.21 4.1 siz par yb Deciles 6 8.2 9.7 8.0 6.9 9.1 3.7 8.1 9.5 4.4 8.6 9.2 2.0 6.8 9.2 2.0 8.9 hool 12.0 4.1 Sc- localities Expenditurea 5 9.8 10.4 9.0 9.5 10.1 6.7 1.21 10.1 7.0 10.3 9.7 3.9 9.0 10.1 4.7 8.4 1.31 8.1 urban Capit INDONESIA orf Per 4 1.41 1.31 1.51 12.8 12.1 10.8 9.1 1.31 7.7 1.61 1.21 8.8 9.3 9.6 8.8 8.9 9.5 8.1 IN4. and lebaT 3 13.5 12.2 13.7 14.2 13.4 1.01 15.0 12.5 9.6 12.1 12.1 9.8 1.61 1.11 8.1 14.7 12.1 18.9 locality al 2 urr 16.0 13.6 16.4 15.6 15.0 18.4 15.0 14.4 19.9 19.4 16.3 26.5 15.0 13.5 23.6 16.8 10.7 8.1 all orf 1 22.3 15.4 23.9 26.6 18.0 39.0 24.3 17.2 39.1 23.3 18.0 38.2 33.1 17.0 43.2 18.8 12.5 28.4 school school school school school school ation in in in in in in school school school school school school in longer in longer in longer in longer in longer in longer School articipp school school school school school school Not In No Not In No Not In No Not In No Not In No Not In No 2M- of 250th< 250K-499K 500K-999K 1M 2M> ypeT settlement Rural Urban Urban Urban Urban Urban Annexes 64 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam , older 1999 latoT , 81.2 1.4 17.4 92.9 0.5 6.6 91.3 0.9 7.8 93.9 0.5 5.7 94.2 0.5 5.3 95.9 0.6 3.4 older 1999 latoT 001 001 001 001 001 001 001 001 001 001 001 001 001 001 001 001 001 001 and and s deciles: eary 01 s e 88.0 1.1 10.9 97.3 0.4 2.4 96.9 0.3 2.8 98.0 0.1 1.9 98.1 0.1 1.8 97.9 0.6 1.5 deciles: 7.9 5.6 4.6 5.5 3.4 1.9 6.8 2.4 2.3 8.5 1.7 2.7 9.7 1.5 3.2 9.9 5.0 eary 01 e 1.31 5 5 ega 9 84.4 1.3 14.3 96.3 0.5 3.2 95.7 0.7 3.5 97.1 0.4 2.5 96.8 0.3 2.9 97.6 0.5 1.9 ega 9 8.5 7.8 6.7 7.3 6.2 3.4 8.6 6.5 3.7 8.8 7.6 3.8 10.0 7.4 5.3 10.9 7.9 5.9 xpenditure xpenditure pita 8 83.6 1.5 14.9 95.1 0.5 4.5 94.9 0.8 4.3 96.5 0.3 3.2 96.7 0.3 3.0 96.6 0.6 2.8 pita 8 9.1 9.1 7.6 8.3 6.8 5.5 9.5 7.9 5.0 8.4 5.1 4.6 9.4 5.9 5.1 9.2 8.5 espondentsr ca espondentsr ca 10.5 per per all orf yb,e 7 all 7 82.8 1.3 15.9 94.6 0.4 5.0 93.1 0.7 6.2 96.0 0.4 3.6 96.0 0.4 3.6 96.2 0.6 3.2 orf yb,e 9.6 8.9 8.7 8.9 6.5 6.6 9.5 7.1 7.4 9.2 6.8 5.7 10.4 8.1 6.9 1.11 10.5 10.3 (%) siz siz tesar yb 6 (%) 82.3 1.4 16.2 94.4 0.4 5.3 92.1 0.9 7.1 94.2 0.4 5.5 94.7 0.6 4.7 96.3 0.6 3.1 9.2 9.9 6.8 7.8 9.6 9.0 9.1 6.8 8.8 9.4 8.2 1.61 Deciles tesar yb 6 Deciles 10.0 10.1 10.0 12.6 10.5 10.3 y y ac 5 81.0 1.4 17.5 93.8 0.6 5.7 91.5 1.1 7.4 92.8 0.5 6.8 94.6 0.8 4.7 96.1 0.5 3.4 ac 5 10.3 10.6 10.4 10.8 1.51 9.2 10.1 12.2 9.5 9.3 9.3 1.21 9.6 16.3 8.3 10.5 8.6 10.4 localities Liter- Expenditurea localities Liter- Expenditurea 4 4 urban Capit 80.5 1.4 18.2 93.1 0.5 6.4 90.6 1.0 8.4 93.2 0.5 6.2 93.4 0.5 6.1 94.9 1.0 4.1 urban Capit 10.6 10.6 1.21 1.61 10.8 1.31 10.5 1.61 1.41 10.6 1.91 1.71 9.1 1.11 10.5 8.6 13.8 10.4 orf Per orf Per INDONESIA and 3 79.7 1.4 18.9 91.9 0.6 7.5 89.3 1.0 9.7 92.3 0.7 7.1 92.9 0.5 6.6 94.6 0.7 4.7 INDONESIA and 3 10.9 1.01 12.2 12.0 13.0 13.8 10.9 12.4 13.8 10.9 15.3 13.9 9.5 1.11 1.91 9.5 1.21 13.1 IN5. lebaT locality 2 IN5. 2 78.6 1.4 19.9 91.2 0.6 8.2 88.9 0.9 10.1 92.4 0.7 6.9 92.2 0.5 7.2 93.7 0.7 5.6 7.9 9.2 al urr lebaT locality 1.31 12.0 13.4 12.5 14.5 15.9 1.71 12.4 15.7 12.6 19.5 15.5 10.8 13.3 14.9 13.1 al urr orf 1 76.1 1.6 22.3 87.7 0.8 1.51 85.6 1.2 13.2 89.9 0.6 9.5 88.9 0.4 10.7 93.8 0.7 5.5 orf 1 1.81 14.4 16.1 13.3 20.5 24.6 12.3 17.7 22.2 12.5 16.1 22.0 12.1 12.6 25.6 8.1 9.2 13.2 R/W R/W R/W R/W R/W R/W R/W R/W R/W R/W R/W R/W Latin Other Latin Other Latin Other Latin Other Latin Other Latin Other Latin Other Latin Other Latin Other Latin Other Latin Other Latin Other naC read/write R/W R/W Cannot R/W R/W Cannot R/W R/W Cannot R/W R/W Cannot R/W R/W Cannot R/W R/W Cannot naC read/write R/W R/W Cannot R/W R/W Cannot R/W R/W Cannot R/W R/W Cannot R/W R/W Cannot R/W R/W Cannot of 250th< 250-499K 500K-1M 1M-2M 2M> of 250th< 250-500K 500-1M 1M-2M 2M> ypeT settlement Rural Urban Urban Urban Urban Urban ypeT settlement Rural Urban Urban Urban Urban Urban 65 tus 3.0 2.9 5.9 6.8 5.4 7.1 5.1 8.7 6.1 8.3 5.9 9.3 sta enderg 1999 latoT 97.0 97.1 94.1 93.2 94.6 92.9 94.9 91.3 93.9 91.7 94.1 90.7 localities (%): yment ,e yb 2.8 2.7 6.7 8.5 5.8 8.9 5.6 7.3 Poor 97.2 97.3 93.3 91.5 94.2 91.1 94.4 89.4 10.6 92.7 89.0 1.01 89.9 10.1 87.4 12.6 tus urban siz Emplo- f or yb sta ty and ver 96.9 3.1 96.9 3.1 94.4 5.6 93.8 6.2 94.8 5.2 93.5 6.5 95.0 5.0 91.8 8.2 94.2 5.8 92.4 7.6 94.4 5.6 90.9 9.1 po Non-poor and ficial) ficial) ficial) ficial) ficial) ficial) ficial) ficial) ficial) ficial) ficial) ficial) INDONESIA localities (Of (Of (Of (Of (Of (Of (Of (Of (Of (Of (Of (Of al IN7. lebaT urr orf Other Unemployed Other Unemployed Other Unemployed Other Unemployed Other Unemployed Other Unemployed Other Unemployed Other Unemployed Other Unemployed Other Unemployed Other Unemployed Other Unemployed Female Male Female Male Female Male Female Male Female Male Female Male of 250th< 250-499K 500K-999K 1M-2M 2M> ypeT settlement Rural Urban Urban Urban Urban Urban tus (%) and sta 1999 latoT 56.7 43.3 26.4 73.6 66.1 33.9 37.3 62.7 63.3 36.7 35.8 64.2 63.1 36.9 37.9 62.1 68.3 31.7 39.4 60.6 67.2 32.8 36.8 63.2 poor e: king siz orW- and localities yb Poor 56.6 43.4 27.0 73.0 71.1 28.9 41.1 58.9 64.8 35.2 38.5 61.5 64.4 35.6 39.5 60.5 71.8 28.2 42.7 57.3 74.2 25.8 42.5 57.5 al urr 56.7 43.3 25.8 74.2 64.4 35.6 36.0 64.0 62.8 37.2 35.1 64.9 62.7 37.3 37.5 62.5 67.5 32.5 38.7 61.3 66.7 33.3 36.4 63.6 non-poor in localities Non-poor orf INDONESIA urban IN6. enderg in working working working working working working working working working working working working lebaT yb Not orkingW Not orkingW Not orkingW Not orkingW Not orkingW Not orkingW Not orkingW Not orkingW Not orkingW Not orkingW Not orkingW Not orkingW Female Male Female Male Female Male Female Male Female Male Female Male of 250th< 250-499K 500-999K 1M-2M 2M> ypeT settlement Rural Urban Urban Urban Urban Urban Annexes 66 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam (%), 1999 Mean* 68.1 10.8 9.9 4.2 5.5 1.5 2.6 53.1 2.7 35.4 3.9 0.2 1.8 0.2 72.0 22.0 3.9 2.1 20.6 22.1 43.8 10.4 0.2 2.8 0.1 19.8 2.9 7.3 20.1 24.0 25.9 tus sta ty erv Urban 76.5 7.7 6.2 1.2 6.4 1.9 1.6 57.4 1.2 33.5 1.4 0.4 4.4 0.2 57.0 29.9 10.5 2.6 7.1 16.8 55.4 10.7 0.6 9.2 0.1 10.3 0.9 6.4 27.2 29.6 25.5 po settlements: Poor yb urban orf 66.6 1.31 10.6 4.7 5.4 1.5 2.8 52.4 3.0 35.7 4.4 0.1 1.3 0.2 74.7 20.6 2.7 2.0 23.0 23.1 41.7 10.3 0.2 1.6 0.1 21.5 3.3 7.4 18.8 23.0 26.0 Non-poor acteristics and al harc urr orf Mean* 90.7 0.9 1.5 2.6 3.0 1.2 1.0 46.2 3.6 36.5 0.7 0.3 1.01 0.6 41.4 39.1 14.9 6.0 4.6.0 10.0 36.5 24.1 1.7 21.2 0.7 13.9 0.3 3.2 22.2 31.5 28.8 Housing- Rural Poor 94.5 0.3 0.5 1.4 2.2 1.0 0.7 47.4 2.1 31.9 0.3 0.4 16.0 1.1 31.3 39.2 22.4 7.1 2.2 6.6 31.4 21.5 2.3 34.9 1.1 1.71 0.3 3.7 25.4 32.7 26.2 INDONESIA IN8. 88.8 1.2 2.0 3.2 3.4 1.3 1.2 45.7 4.4 38.8 0.9 0.3 8.4 0.4 46.4 39.0 1.21 3.4 7.8 1.61 39.0 25.3 1.3 14.5 0.4 15.0 0.4 3.0 20.6 30.9 30.0 Non-poor lebaT sampled tile fiber roof plaster house pulp sheeting tile m2 m2 m2 m2 of m2 population Private Lease Rent ficialfO roof walls floor Free Other of Concrete Corrugated Shingle Iron Asbestos Sugar Leaves Other of Brick oodW Bamboo Other of Marble Floor Cement oodW Bamboo Earth Other area 0 10< 10-19 20-39 40-59 60+ total Status ypeT ypeT ypeT Floor for* 67 , * 1999 Mean 68.1 10.8 9.9 4.2 5.5 1.5 2.6 53.1 2.7 35.4 3.9 0.2 1.8 0.2 72.0 22.0 3.9 2.1 20.6 22.1 43.8 10.4 0.2 2.8 0.1 19.8 2.9 7.3 20.1 24.0 25.9 01 56.5 14.8 15.5 7.2 4.6 1.4 4.3 57.9 2.7 29.8 4.7 0.0 0.4 0.2 86.7 10.9 0.7 1.7 40.5 25.9 27.6 5.5 0.1 0.4 0.1 37.2 6.8 1.11 1.41 12.3 21.1 settlements deciles: e 62.8 13.0 12.6 5.8 4.5 1.2 3.7 55.4 2.9 31.9 5.1 0.1 0.7 0.3 81.9 15.2 1.4 1.5 32.1 25.5 34.5 7.1 0.1 0.6 0.1 29.6 5.4 8.2 13.0 18.6 25.2 urban 66.7 1.31 10.2 5.1 5.6 1.1 2.9 54.3 2.5 34.3 5.0 0.0 0.8 0.2 79.4 16.8 1.8 1.9 26.0 25.0 39.6 8.2 0.1 1.0 0.1 24.6 3.6 6.9 16.1 20.5 28.2 orf Deciles 9.1 4.4 5.7 1.5 2.6 2.5 4.2 0.1 1.2 0.3 2.7 2.2 0.1 1.6 0.1 2.9 6.5 and xpenditure 68.4 10.9 53.3 35.7 75.2 20.0 21.8 23.6 42.8 10.0 20.6 17.9 24.6 27.5 al 70.0 10.1 8.5 4.3 5.3 1.6 2.8 51.8 3.0 36.6 4.3 0.1 1.3 0.1 74.0 20.9 3.0 2.1 19.7 22.8 45.3 9.8 0.2 2.1 0.1 18.8 2.0 6.3 19.6 25.5 27.8 urr pita Urban ca orf Expenditurea 70.1 9.7 8.6 4.0 5.9 1.6 2.2 52.9 2.5 37.2 3.4 0.1 1.5 0.2 70.9 23.0 3.9 2.2 16.7 21.3 47.8 1.11 0.2 2.9 0.1 16.2 1.8 6.4 21.4 27.3 27.0 per Capit 71.9 9.3 8.3 3.5 5.3 1.6 2.2 2.7 37.5 3.9 0.2 2.3 0.2 67.6 25.4 5.0 2.0 13.7 21.1 49.7 1.71 0.3 3.4 0.0 13.4 1.4 5.9 23.1 28.3 27.8 (%) yb 51.0 Per 71.2 10.0 8.3 2.8 6.0 1.7 1.9 51.6 3.0 37.6 3.2 0.2 2.3 0.2 64.0 27.0 6.3 2.7 12.1 20.4 51.1 12.6 0.2 3.5 0.1 12.5 1.4 6.2 25.2 28.1 26.5 72.7 8.9 8.3 2.1 6.4 1.7 1.7 51.1 2.9 37.6 3.0 0.3 3.3 0.1 60.3 30.3 7.0 2.5 10.4 17.6 51.0 14.6 0.5 5.8 0.1 1.11 1.2 6.9 27.1 29.1 24.6 acteristics 123456789 73.0 8.8 8.5 1.4 6.4 1.9 1.7 50.7 2.5 37.4 2.1 0.4 5.1 0.2 53.3 34.9 9.3 2.6 7.1 16.2 52.9 15.0 0.7 7.9 0.2 8.7 1.6 8.1 30.1 28.0 23.4 harc * Mean 90.7 0.9 1.5 2.6 3.0 1.2 1.0 46.2 3.6 36.5 0.7 0.3 1.01 0.6 41.4 39.1 14.9 4.6 6.0 10.0 36.5 24.1 1.7 21.2 0.7 13.9 0.3 3.2 22.2 31.5 28.8 01 Housing- 82.1 2.4 4.1 6.1 3.8 1.5 1.5 46.2 3.9 40.4 1.6 0.2 5.7 0.5 56.8 30.9 7.9 4.4 14.1 15.9 38.7 19.4 0.8 10.8 0.3 21.6 0.8 3.9 17.4 25.3 31.0 88.5 1.3 2.0 3.3 3.7 1.2 1.2 45.7 3.8 39.7 1.0 0.2 7.8 0.5 49.1 35.6 10.9 4.4 8.8 12.7 39.5 22.2 1.2 15.2 0.4 16.5 0.5 2.8 19.3 30.5 30.4 90.2 1.0 1.5 3.1 2.9 1.3 1.2 45.9 3.8 37.6 0.8 0.4 9.6 0.6 45.0 37.0 13.5 4.5 6.7 10.9 39.0 22.6 1.8 18.4 0.6 14.6 0.2 2.7 20.2 31.2 31.1 INDONESIA Deciles 91.2 0.8 1.3 2.4 3.1 1.2 1.0 45.2 3.7 38.2 0.7 0.2 10.4 0.6 42.8 38.2 13.9 5.2 5.8 10.2 38.8 23.6 1.4 19.7 0.6 13.1 0.3 2.8 21.2 32.8 29.6 91.5 0.7 1.2 2.3 3.0 1.2 0.9 45.5 3.4 36.9 0.7 0.2 1.81 0.6 40.2 40.0 15.1 4.8 5.0 9.5 37.7 24.6 1.7 20.7 0.6 13.2 0.2 2.6 22.4 33.3 28.3 IN9. Rural lebaT Expenditurea 92.1 0.5 1.2 2.2 2.9 1.0 0.9 46.3 3.7 35.9 0.5 0.4 1.71 0.6 40.2 39.3 16.0 4.6 4.4 9.0 37.5 24.4 1.7 22.5 0.6 12.7 0.3 2.9 22.8 32.2 29.1 Capit 93.2 0.6 1.0 1.6 2.6 1.1 0.9 46.5 3.3 35.3 0.6 0.4 12.4 0.6 36.9 41.6 16.9 4.6 3.6 8.7 34.7 25.9 1.8 24.7 0.7 12.0 0.3 2.8 23.3 33.9 27.7 Per 93.7 0.4 0.6 1.7 2.4 1.2 0.8 46.5 3.9 34.5 0.4 0.3 12.9 0.8 34.8 43.4 17.6 4.2 3.0 7.1 35.1 26.4 2.0 25.7 0.7 1.31 0.2 3.2 25.3 32.8 27.2 93.9 0.5 0.7 1.2 2.7 1.1 0.8 47.2 3.7 32.8 0.4 0.4 14.2 0.6 31.5 44.0 19.8 4.7 2.5 6.9 32.2 26.4 2.1 28.9 0.9 10.9 0.2 4.0 25.6 32.6 26.6 123456789 94.5 0.3 0.5 0.9 2.8 0.9 0.6 47.9 2.7 30.8 0.3 0.5 16.4 0.8 27.9 45.3 22.0 4.8 1.8 5.7 28.0 27.6 2.7 32.6 1.7 9.9 0.3 4.7 deciles 28.3 32.0 24.7 all sampled, house tile fiber of roof roof walls floor plaster population of pulp sheeting of of tile area m2 m2 m2 m2 m2 total Status Private Lease Rent ficialfO Free Other ypeT Concrete Corrugated Shingle Iron Asbestos Sugar Leaves Other ypeT Brick oodW Bamboo Other ypeT Marble Floor Cement oodW Bamboo Earth Other Floor <10 10-19 20-39 40-59 >60 for* Annexes 68 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam , * 0.3 2.1 1999 Mean 46.6 16.9 24.4 5.3 1.5 0.3 0.6 1.8 51.9 48.1 65.6 21.0 6.2 7.2 01 0.4 5.2 60.9 16.9 12.8 2.0 0.7 0.1 0.2 0.8 68.4 31.6 67.7 21.7 2.6 8.0 settlements decile: e 0.4 2.8 55.4 18.2 18.2 2.7 0.8 0.1 0.4 1.1 61.1 38.9 69.8 19.4 4.4 6.3 urban 0.3 2.4 50.4 18.5 21.8 4.1 0.8 0.1 0.3 1.4 55.7 44.3 70.3 18.2 4.7 6.7 orf xpenditure Deciles 0.4 2.2 48.1 18.1 23.2 4.4 1.2 0.3 0.4 1.7 53.7 46.3 68.2 19.2 5.5 7.0 and 0.2 1.8 al 46.3 18.6 24.7 4.7 1.4 0.2 0.4 1.6 51.0 49.0 68.4 18.9 6.1 6.6 pita Urban urr ca orf Expenditurea 0.2 1.3 43.5 17.6 27.4 5.4 1.3 0.3 0.8 2.1 47.7 52.3 65.6 20.7 6.4 7.3 per (%) yb 0.2 1.4 6.3 1.8 0.4 0.9 1.9 6.7 6.7 Capit 42.0 17.2 28.0 46.1 53.9 65.8 20.8 Per teraw 3456789 0.3 1.3 40.4 15.2 29.8 7.3 2.2 0.4 0.8 2.2 45.0 55.0 62.0 23.4 7.3 7.3 2 0.4 1.2 38.4 14.7 30.8 8.0 2.4 0.6 1.1 2.5 43.2 56.8 60.1 23.7 8.4 7.7 1 0.4 1.0 35.7 13.4 31.2 10.3 3.3 0.6 1.1 2.9 41.5 58.5 54.8 25.6 1.51 8.0 drinking * of 0.4 0.4 Mean 1.11 7.9 33.2 17.6 12.4 7.0 5.4 4.7 14.5 85.5 48.1 25.9 19.7 6.3 ces 01 0.4 1.0 19.2 12.0 33.5 12.6 9.8 4.5 3.1 3.9 23.7 76.3 59.5 21.9 13.5 5.1 Sour- 0.5 0.4 13.8 9.8 35.0 14.6 1.71 5.6 4.1 4.5 17.7 82.3 53.1 24.4 16.7 5.8 0.5 0.4 12.2 8.7 35.6 16.2 1.51 6.1 4.0 4.7 15.7 84.3 52.4 24.5 17.2 5.9 Deciles 0.5 0.4 1.61 8.7 34.6 16.6 1.51 6.5 5.2 4.4 14.8 85.2 49.6 25.5 18.8 6.2 INDONESIA 0.4 0.2 10.6 8.2 21 33.1 18.6 6.8 5.3 4.7 14.3 85.7 47.9 26.4 19.5 6.2 Rural IN10. Expenditurea 0.5 0.3 10.1 7.5 33.3 17.9 12.7 7.2 5.6 4.9 12.9 87.1 46.9 26.4 20.3 6.5 lebaT 456789 0.3 0.2 8.6 6.4 Capit 33.4 19.8 12.4 7.9 6.2 4.6 1.61 88.4 44.9 27.2 21.1 6.7 Per 3 0.3 0.2 7.6 6.1 31.9 20.3 13.9 8.0 7.1 4.5 10.4 89.6 42.6 28.1 22.7 6.5 2 0.4 0.1 7.3 5.1 30.4 21.2 14.6 8.9 7.4 4.5 10.2 89.8 40.3 28.3 24.5 6.8 1 0.2 0.2 5.5 3.9 29.1 21.3 15.2 10.7 7.3 6.5 8.7 91.3 37.3 27.9 26.5 8.3 deciles all sampled, well spring of water well spring water population water purchase water to total Source drinking Others Bottled paT not Pump Protected Unprotected Protected Unprotected River Rain How obtain Purchase Do Drinking facility Private Shared Public None for* 69 0 . . 0.5 3.7 46.9 38.5 8.3 1.7 0.3 54.0 46.0 65.9 15.6 9.8 8.7 1999 Mean* 2M > . . . . localities tus: 1.8 Poor 52.3 19.2 19.8 5.4 1.5 63.1 36.9 52.6 20.4 20.7 6.3 al urr sta Urban ty Non- p roo 0 . . 0.5 3.8 46.7 39.5 7.7 1.5 0.2 53.5 46.5 66.6 15.4 9.2 8.8 orf erv teraw po 0.7 2.6 yb,e 53.9 17.7 18.5 3.9 2.0 0.2 0.2 0.2 60.9 39.1 65.5 18.9 5.3 10.3 Mean* 1M-2M 0.8 0.9 siz Poor 41.5 10.5 29.1 10.0 6.1 0.4 0.5 0.3 50.0 50.0 54.4 26.1 10.9 8.5 drinking yb Urban 0.7 2.9 Non- poor 56.0 18.9 16.8 2.9 1.3 0.2 0.1 0.2 62.7 37.3 67.3 17.8 4.4 10.5 of ces 0.1 2.4 localities Mean* 52.3 16.3 23.8 3.9 0.6 0.2 0.4 0.1 57.8 42.2 64.5 20.0 6.6 9.0 Sour­ . . 500K-999K urban Poor 0.8 32.6 17.4 40.4 6.3 1.1 0.3 1.1 36.4 63.6 58.1 23.7 12.2 6.0 orf Urban 0.1 2.6 Non- poor 55.5 16.1 21.1 3.5 0.6 0.1 0.3 0.1 61.2 38.8 65.5 19.4 5.7 9.4 and INDONESIA Settlement 0.3 1.8 of 34.3 13.4 36.2 7.6 1.0 0.2 0.5 5.0 39.3 60.7 62.4 23.1 5.5 8.9 Mean* ypeT IN11. 250K-499K 0.3 0.5 Poor 18.6 13.7 48.5 12.6 1.9 0.5 0.6 2.8 22.2 77.8 61.0 25.2 8.6 5.1 lebaT Urban Non- poor 0.3 2.0 37.1 13.3 33.9 6.7 0.8 0.1 0.5 5.3 42.4 57.6 62.7 22.8 5.0 9.6 0.2 1.8 Mean* 49.5 13.3 24.2 5.9 2.2 0.5 1.0 1.5 53.9 46.1 67.4 22.2 5.8 4.6 <250th Poor 0.2 1.2 37.3 13.0 32.5 9.1 3.9 0.7 1.0 1.1 41.8 58.2 57.9 26.5 1.31 4.3 Urban 0.2 1.9 Non- poor 52.2 13.3 22.4 5.1 1.9 0.4 1.0 1.5 56.5 43.5 69.5 21.3 4.6 4.6 0.4 0.4 1.11 7.9 33.2 17.6 12.4 7.0 5.4 4.7 14.5 85.5 48.1 25.9 19.7 6.3 Mean* 0.3 0.2 7.0 5.2 Rural Poor 31.1 19.3 16.4 10.6 5.9 4.0 10.0 90.0 38.1 29.8 26.2 5.9 noN 0.5 0.5 -poor 13.1 9.3 34.2 16.7 10.4 5.3 5.1 5.0 16.7 83.3 53.1 23.9 16.4 6.6 sampled well spring of water well spring obtain water population purchase water to total Source drinking Others Bottled paT not Pump Protected Unprotected Protected Unprotected River Rain How water Purchase Do Drinking facility Private Shared Public None for* Annexes 70 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam , * 1999 Mean 77.1 17.2 5.7 78.7 1.81 7.3 2.3 72.6 1.5 9.0 14.7 0.4 1.7 01 77.5 21.0 1.5 91.4 6.4 1.9 0.3 85.0 0.4 5.0 7.9 0.1 1.5 settlements deciles: e 80.3 17.0 2.7 88.2 8.0 3.0 0.8 81.9 0.9 6.4 9.4 0.1 1.3 urban 81.3 15.2 3.4 86.0 9.0 3.6 1.3 79.2 1.1 6.5 1.41 0.2 1.6 orf Deciles 4.5 5.7 1.7 1.2 8.7 0.3 1.8 and xpenditure 80.5 15.0 82.4 10.3 74.9 13.1 al 80.2 15.1 4.7 80.7 1.61 5.9 1.7 74.1 1.3 8.2 14.4 0.2 1.9 urr pita Urban ca orf Expenditurea 78.4 16.2 5.4 77.3 12.6 7.7 2.4 71.5 1.4 9.2 15.7 0.4 1.8 per Capit 77.6 16.0 6.5 75.2 13.3 8.9 2.6 69.1 1.7 9.4 17.6 0.6 1.7 (%) yb Per 3456789 73.8 18.4 7.8 71.6 14.7 10.5 3.1 65.9 2.0 1.91 17.9 0.5 1.9 2 71.6 18.7 9.7 67.0 16.3 12.6 4.1 61.5 2.6 13.4 20.0 0.7 1.9 acteristics 1 67.6 19.5 12.9 60.4 17.9 15.9 5.8 57.5 3.4 13.5 23.1 0.7 1.7 harc * Mean 71.5 16.8 1.71 38.7 10.6 37.0 13.8 32.3 5.6 1.61 45.2 2.5 2.7 tion 01 79.3 14.6 6.2 60.7 10.8 20.9 7.7 52.1 3.6 8.0 32.8 1.6 1.9 Sanita- 75.2 16.0 8.7 49.6 1.21 28.7 10.4 41.5 4.4 9.6 39.7 2.2 2.6 73.5 16.7 9.8 44.3 1.41 32.9 1.41 37.4 4.6 10.7 42.3 2.6 2.5 Deciles 72.4 17.0 10.5 41.3 10.9 35.0 12.8 35.0 4.6 1.51 43.1 2.8 3.0 INDONESIA 71.2 16.8 12.0 38.0 10.3 37.7 14.1 31.7 5.6 1.71 45.3 2.8 2.9 Rural IN12. Expenditurea 70.5 17.0 12.5 34.9 10.7 39.6 14.8 29.2 5.8 12.4 46.6 2.9 3.1 lebaT 456789 Capit 68.6 17.7 13.7 31.4 10.6 42.4 15.6 25.8 6.6 12.5 49.5 2.7 2.8 Per 3 68.1 17.4 14.5 29.0 10.0 43.9 17.1 23.9 6.9 13.0 50.7 2.6 2.9 2 66.5 18.2 15.4 25.1 10.0 47.5 17.4 19.8 7.7 14.2 52.9 2.4 3.0 1 65.3 18.1 16.6 19.6 9.4 50.4 20.7 15.2 7.6 14.9 56.4 3.0 2.8 deciles all sampled, field toilet field population facility of disposal tank latrine total oiletT Private Shared Public ypeT Squatter Throne Dry Others Final Septic Pond/rice River/lake/ocean Hole Shore/open Others for* 71 0 75.4 18.4 6.3 84.7 12.6 2.0 0.7 74.0 1.0 12.5 9.7 2.7 1999 Mean* 2M > . tus: Poor 64.6 24.0 1.41 71.8 23.7 3.9 0.6 61.6 1.5 28.8 4.2 3.9 settlements sta Urban 0 al urr ty 6.0 1.9 0.7 1.0 1.61 2.7 erv Non- poor 76.0 18.1 85.4 12.0 74.7 10.0 orf po 5.3 4.4 1.1 1.5 6.3 9.9 0.2 1.4 (%) ,e yb 77.4 17.3 83.4 1.01 80.7 Mean* 1M-2M siz Poor 68.4 20.8 10.8 66.0 19.4 12.4 2.2 65.7 2.7 1.71 17.8 0.5 1.6 yb Urban Non- poor 78.8 16.7 4.4 86.3 9.6 3.1 1.0 83.2 1.3 5.4 8.5 0.2 1.4 acteristics harc Mean* 73.2 17.3 9.6 80.5 12.0 5.2 2.3 75.9 1.2 12.0 9.7 0.1 1.2 tion settlements 500K-999K Poor 63.4 17.8 18.9 72.5 15.5 8.2 3.8 67.6 2.1 16.2 12.3 0.3 1.4 Sanita- urban Urban orf Non- poor 74.7 17.2 8.1 81.7 1.51 4.7 2.1 77.2 1.1 1.31 9.2 0.1 1.2 and Settlement of 76.4 19.2 4.4 76.9 1.91 8.6 2.6 74.0 1.6 6.8 15.4 0.3 1.9 Mean* INDONESIA ypeT 250K-499K Poor 68.0 21.4 10.6 58.3 15.8 18.3 7.6 55.0 4.0 12.6 25.7 0.5 2.2 Urban IN13. Non- poor 77.9 18.8 3.3 80.2 1.21 6.8 1.8 77.5 1.1 5.8 13.6 0.3 1.8 lebaT Mean* 78.8 15.7 5.4 76.2 1.71 9.2 2.8 68.5 1.7 9.4 18.2 0.6 1.5 <250th Poor 72.5 17.4 10.1 59.7 17.3 16.9 6.1 55.4 2.5 12.7 26.6 1.1 1.7 Urban Non- poor 80.2 15.4 4.4 79.9 10.5 7.6 2.1 71.4 1.6 8.7 16.4 0.5 1.5 Mean* 71.5 16.8 1.71 38.7 10.6 37.0 13.8 32.3 5.6 1.61 45.2 2.5 2.7 Rural Poor 69.4 17.6 13.0 23.7 10.6 48.3 17.4 19.3 5.4 1.01 56.9 3.5 3.9 noN -poor 72.6 16.5 1.01 46.1 10.5 31.3 12.0 38.8 5.7 1.91 39.4 2.1 2.1 sampled field toilet field population facility of disposal tank latrine total oiletT Private Shared Public ypeT Squatter Throne Dry Others Final Septic Pond/rice River/lake/ocean Hole Shore/open Others for* Annexes 72 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam * 0 (%) 1997 18.1 81.9 001 3.1 20.3 28.1 21.9 15.4 1.21 001 73.5 25.8 0.6 001 Mean 01 26.7 73.3 001 . 2.9 14.1 26.4 24.3 18.6 13.7 001 75.9 23.0 1.1 001 acteristics al/urban: 23.8 76.2 001 0 3.4 17.4 25.9 23.3 17.4 12.5 001 74.4 24.3 1.2 001 harc urr, 21.5 78.5 001 . 3.1 18.0 26.6 22.1 18.0 12.1 001 72.4 27.0 0.6 001 Deciles 19.6 80.4 001 . 2.7 17.8 26.2 24.2 16.0 13.2 001 73.1 26.2 0.7 001 deciles e 18.2 81.8 001 . 3.4 20.4 26.8 21.6 15.6 12.1 001 71.7 27.8 0.5 001 household Urban of Expenditurea 16.7 83.3 001 0 3.3 22.2 25.9 22.3 15.3 10.9 001 72.1 27.4 0.5 001 ution xpenditure 001 0.1 001 Capit 16.4 83.6 001 2.2 20.5 29.0 22.2 14.2 1.81 73.4 26.5 Per pita 3456789 13.6 86.4 001 4.3 24.0 29.1 19.1 14.5 9.1 001 73.8 25.5 0.7 001 Distrib­ ca 2 12.4 87.6 001 2.9 25.7 30.8 20.2 12.6 7.8 001 74.2 25.4 0.3 001 per 1 1.31 88.7 001 .... 2.5 23.1 34.2 19.9 1.71 8.6 001 74.3 25.2 0.4 001 * within 12.2 87.8 001 0 3.0 20.7 26.9 21.0 16.2 12.3 001 80.7 18.9 0.3 001 Mean PHILIPPINES 01 19.3 80.7 001 . 2.3 13.3 23.8 23.6 21.4 15.7 001 78.4 20.6 1.0 001 PH1. 17.0 83.0 001 . 2.3 14.9 21.8 22.0 22.4 16.6 001 77.9 21.4 0.7 001 lebaT 14.6 85.4 001 1.0 2.6 15.9 19.8 24.2 21.6 15.8 001 78.0 21.9 0.1 001 Deciles 12.7 87.3 001 3.7 17.3 25.2 20.1 19.6 14.1 001 80.5 19.0 0.5 001 12.6 87.4 001 2.9 20.5 24.8 21.0 17.9 12.8 001 80.1 19.4 0.4 001 Rural Expenditurea 9.7 90.3 001 2.7 23.0 25.3 21.3 16.0 1.71 001 81.3 18.4 0.2 001 .... 456789 Capit 10.5 89.5 001 3.7 23.3 25.6 22.0 14.5 10.9 001 80.9 19.0 0.1 001 Per 3 9.2 90.8 001 0.1 4.1 24.2 32.2 18.7 10.7 10.1 001 82.3 17.4 0.2 001 2 8.2 91.8 001 . 3.7 27.8 33.2 18.4 8.7 8.2 001 0 85.3 14.6 001 1 8.8 91.2 001 . 1.8 26.9 36.7 18.3 9.4 6.9 001 82.4 17.5 0.1 001 deciles all sampled, head the head old old old old old non-relatives old of the household family population of years years years years years years of family more or Gender Female Male otalT Age <15 15-24 25-34 35-44 45-54 55-64 +56 total otalT ypeT Single Extended woT otalT *for 73 (%) lat 1997 To 001 001 001 001 001 001 001 001 001 001 001 001 001 001 001 01 . 14.9 9.0 10.1 9.5 7.0 9.5 1.11 12.1 12.3 10.1 10.4 9.0 18.4 10.1 acteristics al/urban: 13.3 9.4 10.1 87.4 1.11 8.6 9.3 10.7 1.41 1.31 10.1 10.2 9.5 20.0 10.1 harc urr, . 12.0 9.6 10.0 10.1 8.9 9.5 10.1 1.71 10.8 10.0 9.9 10.5 9.8 10.0 . Deciles 10.8 9.8 10.0 8.9 8.7 9.3 1.01 10.4 1.71 10.0 9.9 10.1 10.9 10.0 deciles . e 10.1 10.0 10.0 1.11 10.1 9.6 9.9 10.2 10.9 10.0 9.8 10.8 7.5 10.0 Household­ Urban Expenditurea 9.2 10.2 10.0 12.6 10.7 10.9 9.2 10.2 9.9 9.8 10.0 9.8 10.6 8.3 10.0 xpenditure 9.1 Capit 10.2 10.0 7.3 10.1 10.4 10.1 9.3 10.6 10.0 10.0 10.3 1.8 10.0 Per 7.5 pita 10.5 10.0 13.9 1.81 10.4 8.7 9.4 8.1 10.0 10.0 9.9 1.21 10.0 PHILIPPINES ca 6.9 10.7 10.0 9.4 11 12.6 9.2 8.1 7.0 10.0 10.1 9.8 5.3 10.0 per .... PH2. 123456789 6.1 10.6 9.8 8.0 1.21 1.91 8.9 7.4 7.5 9.8 9.9 9.6 6.8 9.8 lebaT oss acr lat To 001 001 001 001 001 001 001 001 001 001 001 001 001 001 001 01 . 15.2 8.9 9.6 7.3 6.2 8.5 10.9 12.7 12.3 9.6 9.4 10.5 28.9 9.6 . 13.9 9.4 10.0 7.8 7.2 8.1 10.5 13.8 13.5 10.0 9.6 1.31 20.3 10.0 1.91 9.8 10.0 47.3 8.8 7.7 7.4 1.61 13.4 12.9 10.0 9.7 1.61 2.2 10.0 Deciles 10.5 10.0 10.1 12.5 8.4 9.5 9.7 12.2 1.61 10.1 10.0 10.1 15.3 10.1 10.4 10.1 10.1 10.0 10.0 9.3 10.1 1.11 10.6 10.1 10.0 10.4 13.7 10.1 Rural Expenditurea 7.9 10.3 10.0 9.0 1.11 9.5 10.2 9.9 9.6 10.0 10.1 9.8 7.4 10.0 .... 8.7 Capit 10.3 10.1 12.5 1.31 9.6 10.6 9.0 9.0 10.1 10.1 10.1 1.8 10.1 Per 7.6 10.4 10.1 52.7 13.7 1.71 12.1 9.0 6.7 8.3 10.1 10.3 9.3 6.4 10.1 . 6.8 10.6 10.1 12.6 13.6 12.5 8.9 5.4 6.8 10.1 10.7 7.8 1.0 10.1 . 123456789 7.1 10.2 9.9 5.9 12.8 13.5 8.6 5.7 5.6 9.9 10.1 9.1 3.2 9.9 deciles all sampled, head the head old old old old old non-relatives old of the household family population of years years years years years years of family more or Gender Female Male Mean* Age <15 15-24 25-34 35-44 45-54 55-64 +56 total Mean* ypeT Single Extended woT Mean* *for Annexes 74 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam * (%) 1997 Mean 81.2 18.8 001 latoT 001 001 001 (%), 1997 nae tus 78.8 21.2 100 latoT 001 001 001 sta 01 77.3 22.7 001 tus M 01 9.6 12.1 10.1 sta al/urban: 77.3 22.7 001 9.6 12.2 10.1 deciles: 01 100 01 yment r, ur 75.4 24.6 10.3 12.5 10.8 76.7 23.3 001 yment e 9.5 10 12.4 emplo 78.6 21.4 001 9.6 1.41 10 75.3 24.7 100 9.8 12.0 10.3 of deciles 79.4 20.6 001 11 Emplo­ 9.8 10 xpenditure oss Urban Urban ution acr 82.2 17.8 001 10.1 9.5 10 pita 89 89 9.3 MANILA ca 72.2 27.8 100 13.3 10.1 Distrib­ and, 456789 81.8 18.2 001 10 10.1 9.7 per Deciles 3 85.8 14.2 001 Deciles PH4. 3456789 10.6 7.5 10 yb Decile 75.8 24.2 100 Decile 9.3 1.11 9.7 deciles 2 e 87.1 12.9 001 2 10 10.7 6.9 lebaT 9.9 1 85.9 14.1 1 76.1 100 9.5 1.11 PHILIPPINES Expenditurea 001 Expenditurea 10.4 7.4 9.8 23.9 Expenditurea Expenditure * PH3. xpenditure Capit 9.5 001 Capit 001 001 001 Capit Capita Mean 90.5 latoT 8.6 9.8 Per Per Per 81.3 18.7 100 Per 10.1 lebaT pita 01 83.7 16.3 001 01 8.9 16.4 9.6 ca Within 86.8 13.2 001 Across 9.6 10 Within Across 13.8 4567 4567 per 80.3 19.7 100 10.0 9.1 9.8 87.5 12.5 001 9.7 10 13.1 within Distribution 90.3 9.7 001 Distribution 10.1 10.2 10.1 Distribution Distribution 84.1 15.9 100 10.4 7.3 9.8 89.8 10.2 001 01 10.8 10.1 Rural Rural 92.2 7.8 001 56789 10.2 8.2 10 86.9 13.1 100 1.01 6.2 10 456789 92.2 7.8 001 4 10.3 8.2 10.1 3 93.9 6.1 001 3 10.5 6.5 123 10.1 81.2 18.8 100 123 10.2 8.8 9.9 2 94.3 5.7 001 deciles 2 10.6 6.1 10.1 or 1 93.5 6.5 001 1 10.2 6.7 9.9 levels all sampled, status head job status head job status head job status head job have have population have have household job not household job not total household job not job not Employment of Has Does otalT Employment of Has Does Mean for* household Employment of Has Does otalT Employment of Has Does Mean 75 (%), lat 1997 To 60.0 12.4 5.1 1.91 4.3 6.0 0.4 001 tus 01 sta 76.4 1.81 2.4 3.4 1.6 4.2 0.2 001 e ur al/urban: 71.0 13.1 3.2 6.0 1.5 5.0 0.2 001 ten urr, 66.5 13.6 4.5 8.1 2.2 5.0 0.2 001 Deciles 62.1 14.4 4.5 9.4 3.5 5.9 0.1 001 Housing­ deciles e 57.3 14.1 4.8 12.2 4.4 6.5 0.8 001 Urban Expenditurea 55.6 13.3 6.4 13.5 5.1 5.8 0.2 001 xpenditure Capit 55.4 13.1 7.0 14.2 4.4 5.5 0.4 001 PHILIPPINES Per pita 53.1 1.91 6.1 15.7 5.6 7.2 0.4 001 ca PH5. 49.9 10.8 6.3 18.1 7.0 7.6 0.5 001 lebaT per 123456789 25 7.3 5.9 19.1 7.5 7.4 0.8 001 within lat To 68.7 0.9 3.1 21.9 1.7 3.6 0.1 001 01 77.3 1.6 3.7 13.1 0.5 3.7 0.1 001 . 72.4 1.5 3.4 17.4 1.8 3.6 001 . 71.6 1.2 3.7 19.2 1.0 3.2 001 Deciles 66.6 0.8 3.9 23.2 1.5 3.8 0.2 001 64.8 1.0 4.1 23.9 2.2 4.0 0.1 001 Rural Expenditurea 68.1 0.4 2.4 23.9 1.8 3.2 0.2 001 Capit 68.6 0.9 3.0 22.6 1.8 3.1 0.1 001 Per 66.2 0.5 2.8 24.1 2.2 4.1 0.2 001 0 66.5 0.4 2.2 24.4 2.4 4.1 001 123456789 65.9 0.4 1.9 26.3 2.0 3.4 0.1 001 lot, lot, lot lot lot and and and and rent rent-free, rent-free, house house tenure charge charge of house house house, house charge house free charge free ypeT Own Rent Own Own with Own without Rent: with Rent: without otalT Annexes 76 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam * (%), 1997 Mean 84.1 4.0 8.6 2.7 0.7 0 001 latoT 001 001 001 001 001 001 001 01 . 81.4 4.9 12.2 0.2 1.4 001 01 . 9.7 12.3 14.3 0.9 21.2 10.1 housing 0 of al/urban: 83.6 4.2 10.3 0.3 1.6 001 10.0 10.6 12.1 1.2 24.2 5.5 10.1 ypeT­ urr, . 83.7 4.0 1.01 0.5 0.8 001 . 10.0 10.1 12.9 1.9 12.9 10.0 0 84.5 3.9 10.0 1.0 0.6 001 10.0 9.7 1.71 3.6 9.7 3.7 10.0 deciles 0 e 84.1 4.1 9.9 1.2 0.5 001 10.0 10.5 1.71 4.6 8.4 5.5 10.0 Urban Urban 85.9 3.7 8.1 1.8 0.3 0.2 001 10.2 9.3 9.5 6.7 5.0 39.7 10.0 PHILIPPINES . xpenditure 456789 85.0 4.6 7.2 2.8 0.4 001 . 10.1 1.51 8.5 10.5 6.7 10.0 . PH6. pita Deciles 3 85.0 3.7 6.9 4.0 0.3 001 . Deciles 3456789 10.1 9.4 8.1 15.1 5.1 10.0 lebaT ca 2 84.7 3.5 6.1 5.4 0.2 0.2 001 2 10.1 8.7 7.1 20.0 3.0 35.0 10.0 per 0 yb Expenditurea 1 83.1 3.2 3.7 9.7 0.3 001 Expenditurea 1 9.7 7.9 4.2 35.5 3.9 10.4 9.8 * Capit Mean 98.1 0.9 0.3 0.5 0.1 0.1 001 Capit latoT 001 001 001 001 001 001 001 Per 01 . Per 96.9 1.5 0.5 0.5 0.6 001 01 . 9.5 15.8 19.5 9.0 43.2 9.6 Within 9 96.8 1.6 0.7 0.4 0.4 0.2 001 Across 9.9 17.6 25.5 8.0 28.3 17.1 10.0 0 . 8 97.9 1.2 0.4 0.6 001 . 10.0 13.0 15.5 10.8 1.3 10.0 Distribution 7 98.2 0.8 0.2 0.4 0.2 0.2 001 Distribution 10.1 9.2 7.3 8.3 12.9 18.8 10.1 0 6 98.5 0.7 0.3 0.2 0.3 001 10.1 8.1 10.2 4.0 3.3 25.1 10.1 Rural Rural . 5 98.8 0.5 0.1 0.5 0.2 001 . 10.1 5.8 2.3 9.1 12.6 10.0 . 98.5 0.7 0.1 0.6 0.1 001 . 456789 10.1 8.5 4.3 1.31 6.3 10.1 34 98.0 0.9 0.2 0.7 0.1 0.1 001 3 10.1 9.9 7.5 12.8 9.2 12.3 10.1 . 98.5 0.4 0.1 0.8 0.1 001 . deciles 2 10.2 4.8 5.1 15.8 7.8 10.1 or 0 . 12 98.7 0.7 0.1 0.6 001 . 1 9.9 7.3 2.9 1.01 1.9 9.9 levels all sampled, house house boat house house boat building building population of house cave, of house cave, total ypeT Single Duplex Apartment Improvised Commercial Other: otalT ypeT Single Duplex Apartment Improvised Commercial Other: Mean* for* 77 (%) 1997 Mean* 64.6 6.5 21.7 6.3 0.9 1.0 001 latoT 001 001 001 001 001 001 001 housing of al/urban: 01 64.4 6.7 27.6 0.6 0.8 001 01 10.8 1.01 13.7 1.0 9.0 10.8 ypeT­ urr, 9 64.3 7.3 24.5 1.0 2.9 001 10.2 1.51 1.61 1.7 32.4 10.3 deciles e MANILA 8 64.1 5.8 28.2 0.8 1.1 001 89 10.1 8.9 13.2 1.2 1.71 10.1 PH7. lebaT epita xpenditur .... 7 Decile 62.9 6.3 26.6 2.5 1.7 001 .... Decile 9.5 9.4 1.91 3.9 17.4 9.7 ca . . per 6 yb 63.1 7.2 25.8 3.6 0.4 001 9.6 5.6 3.9 9.9 Expenditurea Expenditurea 10.8 1.81 Capit Capit 5 567 Per 67.4 6.5 20.8 4.3 0.4 0.6 001 9.8 Per 10.2 9.7 9.4 6.7 4.3 001 ithin W . 4 66.4 7.9 19.1 5.7 0.9 001 Across . 10.1 1.91 8.7 9.0 9.5 9.8 Distribution . . 3 65.6 6.1 17.6 10.2 0.5 001 Distribution 9.9 9.1 7.9 15.9 5.6 9.8 . 2 65.3 4.9 16.2 13.4 0.1 001 . 234 10.1 7.6 7.5 21.5 1.1 10.0 . 1 62.3 6.6 9.4 21.2 0.5 001 . 1 9.6 10.0 4.3 33.6 5.0 9.9 deciles all sampled, house house boat house house boat building building population of house cave, of house cave, total ypeT Single Duplex Apartment Improvised Commercial Other: otalT ypeT Single Duplex Apartment Improvised Commercial Other: Mean* *for Annexes 78 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam * (%) 1997 Mean 62.8 14.6 1.6 13.3 6.3 1.3 001 latoT 001 001 001 001 001 001 001 oofr 01 91.4 2.1 0.2 5.2 0.9 0.1 001 01 14.7 1.4 1.4 3.9 1.5 1.0 10.1 of ypeT­ al/urban: 83.3 4.7 0.4 9.0 2.5 0.1 001 13.4 3.3 2.6 6.8 4.0 0.7 10.1 urr, 77.5 7.1 0.5 1.31 3.3 0.2 001 12.4 4.9 3.2 8.5 5.2 1.9 10.0 73.2 10.1 0.5 1.61 4.0 0.5 001 1.61 6.9 2.8 8.7 6.3 4.2 10.0 deciles e 65.3 13.2 0.6 14.7 5.5 0.7 001 10.5 9.1 3.7 1.11 8.7 5.3 10.0 Urban Urban PHILIPPINES 59.4 16.3 1.4 15.5 6.9 0.5 001 9.5 1.11 8.6 1.61 10.8 4.2 10.0 PH8. xpenditure 456789 53.9 18.0 1.7 16.4 8.5 1.4 001 8.6 12.4 10.1 12.3 13.5 1.21 10.0 lebaT pita Deciles 3 46.9 21.0 2.6 18 9.6 1.8 001 Deciles 3456789 7.5 14.4 15.8 13.6 15.1 14.3 10.0 ca 2 44.0 23.5 3.2 16.5 10.3 2.5 001 2 7.0 16.0 19.3 12.4 16.2 19.3 10.0 per yb Expenditurea 1 31.7 30.7 5.5 12 15.2 5.0 001 Expenditurea 1 5.0 20.6 32.5 1.21 18.6 37.8 9.8 * Capit Mean 44.2 40.8 0.7 8.3 5.8 0.2 001 Capit latoT 001 001 001 001 001 001 001 Per 01 . Per 74.7 13.3 0.6 8.7 2.8 001 01 . 16.3 3.1 7.9 10.1 4.6 9.6 Within 9 63.7 21.8 0.4 9.4 4.4 0.2 001 Across 14.4 5.3 5.9 1.41 7.6 13.2 10.0 8 53.6 29.6 0.7 1.01 5.1 0.1 001 12.2 7.3 9.8 13.3 8.7 3.5 10.0 Distribution 7 48.8 35.8 0.8 8.6 5.9 0.1 001 Distribution 1.11 8.8 1.21 10.5 10.2 4.2 10.1 6 43.2 39.0 0.5 10.2 6.9 0.2 001 9.9 9.6 7.1 12.4 12.0 10.3 10.1 Rural Rural . 5 41.9 43.0 0.7 8.0 6.4 001 . 9.5 10.6 9.7 9.7 10.9 10.0 35.6 47.5 0.8 9.4 6.5 0.2 001 456789 8.1 1.71 10.6 1.51 1.21 10.5 10.1 34 31.3 54.2 0.8 6.6 6.9 0.2 001 3 7.1 13.4 1.01 8.1 12.0 1.81 10.1 27.7 58.5 1.3 6.3 5.8 0.5 001 deciles 2 6.3 14.5 17.5 7.7 10.0 28.6 10.1 or 12 22.5 64.6 0.7 4.4 7.5 0.3 001 1 5.0 15.6 9.3 5.3 12.7 17.9 9.9 levels all sampled, iron iron nipa strong light salvaged nipa strong light salvaged Roof materials concrete, Roof materials population of cogon, Concrete, of cogon, total peyT Strong: Light: Salvaged Mixed: predominantly Mixed: predominantly Mixed: predominantly otalT ypeT Strong: Light: Salvaged Mixed: predominantly Mixed: predominantly Mixed: predominantly Mean* for* 79 * (%) 1997 Mean 59.4 14.5 1.8 15.5 7.4 1.4 001 latoT 001 001 001 001 001 001 001 allsw 01 89.9 2.1 0.2 6.4 1.3 0.1 001 01 15.3 1.4 1.3 4.2 1.7 0.4 10.1 of al/urban: 81.1 4.0 0.4 1.51 2.9 0.1 001 13.8 2.8 2.0 7.5 4.0 0.7 10.1 ypeT­ urr, 75.3 6.1 0.4 14.0 3.8 0.3 001 12.7 4.2 2.3 9.1 5.1 2.0 10.0 71.4 8.8 0.6 13.9 4.7 0.5 001 12.0 6.0 3.3 9.0 6.3 3.7 10.0 deciles e 62.2 12.7 1.0 16.9 6.3 0.9 001 10.5 8.8 5.4 1.01 8.6 6.3 10.0 Urban Urban PHILIPPINES 55.3 15.5 1.5 19.0 8.0 0.7 001 9.3 10.7 8.4 12.2 10.7 5.1 10.0 PH9. xpenditure 50.2 17.6 2.1 18.2 10.2 1.7 001 8.5 12.2 1.41 1.81 13.7 12.0 10.0 lebaT pita Deciles 43.1 21.8 2.7 19.5 10.9 2.0 001 Deciles 7.3 15.1 14.8 12.6 14.6 14.7 10.0 ca 23456789 37.5 24.6 3.3 19.0 12.6 2.9 001 23456789 6.3 17.0 18.1 12.2 16.9 21.1 10.0 per yb Expenditurea 1 26.5 31.9 6.2 16.6 13.9 4.8 001 Expenditurea 1 4.4 21.6 33.1 10.5 18.4 34.1 9.8 * Capit Mean 39.1 41.0 0.9 1.31 7.3 0.4 001 Capit latoT 001 001 001 001 001 001 001 Per Per 01 72.4 1.51 0.7 1.31 3.8 0.3 001 01 17.8 2.7 7.5 9.7 5.0 6.7 9.6 Within 58.2 21.4 0.5 14.8 4.8 0.2 001 Across 14.9 5.2 6.0 13.1 6.5 6.5 10.0 46.4 29.9 0.9 15.5 6.9 0.3 001 1.91 7.3 10.6 13.7 9.5 7.9 10.0 Distribution 45.3 35.1 0.8 1.21 7.6 0.1 001 Distribution 1.71 8.6 8.8 10.0 10.4 2.2 10.1 37.6 39.1 0.4 14.4 8.2 0.3 001 9.7 9.6 4.9 12.9 1.31 7.3 10.1 Rural Rural 37.1 43.4 0.9 10.0 8.4 0.2 001 9.5 10.6 10.2 8.9 1.51 5.5 10.0 30.4 47.0 0.8 12.2 8.8 0.7 001 7.8 1.61 9.7 10.9 12.1 18.4 10.1 25.6 55.9 1.3 9.2 7.8 0.3 001 6.6 13.7 14.8 8.2 10.7 7.2 10.1 23.2 58.8 1.2 8.2 7.9 0.8 001 deciles 6.0 14.5 13.9 7.4 10.8 21.7 10.1 or 123456789 16.3 67.0 1.2 5.9 9.1 0.6 001 123456789 4.1 16.1 13.5 5.2 12.2 16.5 9.9 levels all sampled, iron iron nipa strong light salvaged nipa strong light salvaged allsW materials concrete, allsW materials population of cogon, concrete, of cogon, total ypeT Strong: Light: Salvaged Mixed: predominantly Mixed: predominantly Mixed: predominantly otalT ypeT Strong: Light: Salvaged Mixed: predominantly Mixed: predominantly Mixed: predominantly Mean* for* Annexes 80 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam (%) 1997 1.3 3.1 6.4 3.3 001 001 001 001 001 001 001 001 oofr Mean* 66.4 19.5 latoT of ypeT­ al/urban: 01 urr, 93.6 0.6 0.4 4.2 1.0 0.2 001 01 15.2 5.1 1.3 2.3 1.7 0.5 10.8 84.7 0.3 0.7 1.51 2.4 0.3 001 13.1 2.7 2.3 6.1 3.8 1.0 10.3 MANILA deciles e . . 89 89 PH10. 80.8 0.4 15.0 3.4 0.4 001 12.3 1.2 7.8 5.4 1.2 10.1 lebaT xpenditure pita Decile 76.9 1.5 1.1 15.3 3.9 1.4 001 Decile 1.21 1.41 3.5 7.6 5.9 4.0 9.7 ca per yb 71.1 2.3 1.5 18.3 4.7 2.0 001 4.8 9.3 7.3 6.2 9.9 Expenditurea Expenditure 10.6 17.4 Capit Capita Per 63.6 1.6 2.7 23.4 7.1 1.6 001 9.4 Per 12.1 8.4 1.71 10.8 4.9 9.8 Within 4567 59.9 1.5 2.3 25.3 7.3 3.8 001 Across 4567 8.9 1.71 7.1 12.7 1.11 1.41 9.8 Distribution 46.9 1.2 5.3 30.9 10.7 5.0 001 Distribution 6.9 9.1 16.5 15.5 16.3 14.9 9.8 46.9 1.9 6.8 26.9 10.5 7.0 001 7.1 15.0 21.8 13.8 16.4 21.5 10.0 123 36.2 2.0 10.5 26.2 13.8 1.31 001 123 5.4 15.7 33.0 13.3 21.3 34.4 9.9 deciles all strong light salvaged strong light salvaged sampled, iron iron nipa nipa roof materials materials concrete, population of cogon, predominantly predominantly predominantly roof concrete, of cogon, predominantly predominantly predominantly total ypeT Strong: Light: Salvaged Mixed: Mixed: Mixed: otalT ypeT Strong: Light: Salvaged Mixed: Mixed: Mixed: otalT *for 81 (%), 1997 3.8 5.5 6.0 7.9 0.9 001 001 001 001 001 001 001 001 001 tus Mean* 50.0 25.9 latoT sta e decile: ur e 01 70.6 18.8 1.4 1.4 2.4 5.1 0.3 001 01 15.2 7.8 4.0 2.8 4.3 6.9 4.3 10.8 ten 9 5.3 3.8 3.4 8.9 6.1 Housing­ xpenditure 60.7 26.0 2.0 2.1 2.0 6.8 0.5 001 12.5 10.3 10.3 pita ca 8 56.6 27.7 3.5 4.0 1.9 6.1 0.3 001 89 1.51 10.8 9.1 7.3 3.2 7.8 3.0 10.1 MANILA per yb 7 Decile 51.2 29.7 3.4 4.6 4.5 6.3 0.3 001 Decile 9.9 1.11 8.7 8.1 7.3 7.8 3.0 9.7 PH11. lebaT 6 48.4 28.5 2.2 5.9 6.5 6.8 1.7 001 9.5 5.7 8.5 9.9 Expenditurea Expenditurea 10.9 10.7 10.6 19.2 Capit Capit 5 567 Per 46.5 29.5 5.5 4.6 6.3 7.0 0.6 001 9.1 8.2 8.6 6.7 9.8 Per 1.11 14.1 10.1 ithin W 4 45.5 28.4 6.1 7.1 5.6 6.5 0.9 001 Across 8.9 10.8 15.6 12.6 9.1 8.1 10.4 9.8 Distribution 3 41.1 28.1 4.5 6.7 8.3 10.5 0.9 001 Distribution 8.0 10.6 1.51 1.91 13.4 13.0 9.9 9.8 2 36.8 25.2 5.4 9.5 10.8 1.01 1.2 001 234 7.4 9.8 10 14.1 17.4 17.9 14.0 14.2 1 40.4 17.7 4.6 9.5 12.7 13.1 2.0 001 1 8.0 6.8 1.91 17.2 20.8 16.4 23.1 9.9 charge deciles charge charge all charge with without w/c wo/c with without lot lot w/c wo/c lot lot sampled, lot lot lot and and lot lot lot and and and and rent rent-free, rent-free, house house and and rent rent-free rent-free house house tenure tenure population of house house house, house house free free of house house house house house free free total ypeT Own Rent Own Own Own Rent: Rent: otalT ypeT Own Rent Own Own Own Rent: Rent: Mean *for Annexes 82 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam , * 1997 Mean 37.3 21.4 13.2 15.0 5.7 2.0 0.1 5.3 001 90.3 9.7 001 (%), 1997 Mean* 0.78 0.77 0.27 0.35 0.51 0.06 0.12 01 electricity 66.0 8.7 16.9 4.2 2.1 0.5 0.2 1.4 001 99.4 0.6 001 lesba and al/urban, 57.2 13.2 15.7 7.0 2.9 0.8 0.2 3.0 001 98.2 1.8 001 dur al/urban, 0 0.07 teraw , urr 0.67 0.30 0.02 0.06 0.01 52.0 15.3 15.6 8.5 3.9 0.7 0.1 3.9 001 97.5 2.5 001 urr Urban Poor , of decile 45.4 17.7 15.5 1.11 5.1 1.2 0.2 3.8 001 96.2 3.8 001 tus ce e 0 37.7 20.4 13.8 15.1 5.6 1.4 5.9 001 94.2 5.8 001 Consumer­ sta ty 0.79 0.81 0.30 0.38 0.55 0.06 0.13 Sour­ 32.4 25.3 12.4 16.8 5.8 1.7 0.2 5.4 001 91.5 8.5 001 erv Deciles Non-poor Urban po xpenditure 0 456789 27.4 27.2 1.61 18.8 6.1 2.7 6.2 001 89.3 10.7 001 yb pita 3 24.4 26.5 1.61 20.6 7.0 2.9 0.1 6.8 001 0.73 0.34 0.08 0.14 0.17 0.01 0.04 86.2 13.8 001 PHILIPPINES Expenditure Mean ca Capita PHILIPPINES 2 8.0 3.4 0.2 7.2 per Deciles 18.1 28.8 10.6 23.7 001 79.8 20.2 001 PH13. Per yb 1 1.51 31.0 8.2 23.9 1.11 5.2 0.1 9.0 001 70.5 29.5 001 PH12. lebaT 1 0 Rural Poor 0.64 0.1 0.01 0.03 0.02 0.01 * lebaT Expenditurea 7.2 Mean 20.3 16.9 22.8 17.3 14.0 0.6 1.0 001 52.1 47.9 001 Capit 01 22.0 17.2 27.8 14.3 10.5 6.4 0.6 1.3 001 87.2 12.8 001 1 Per 13.2 18.7 24.6 19.3 12.3 10.2 0.5 1.1 001 74.0 26.0 001 0.77 0.45 0.1 0.19 0.24 0.01 0.05 Non-poor 10.0 21.9 20.4 20.4 15.5 10.2 0.6 0.9 001 64.7 35.3 001 7.1 21.4 18.3 24.6 14.3 12.5 1.0 0.9 001 59.6 40.4 001 deciles all 5.3 20.6 17.2 23.8 17.7 13.8 0.3 1.2 001 54.6 45.4 001 set recorder motorcycle 4.0 22.3 15.6 24.2 17.2 14.8 0.7 1.3 001 50.5 49.5 001 tape sampled, condition jeep,, Rural radio television video stereo refrigerator air car 456789 4.7 19.8 14.0 23.0 20.0 16.7 0.9 1.0 001 41.9 58.1 001 population owns owns owns owns owns owns owns 3 2.8 22.3 12.2 25.9 18.9 16.8 0.5 0.6 001 36.5 63.5 001 total Family Family Family Family Family Family Family *for 2 2.0 19.6 10.4 26.4 22.8 17.8 0.4 0.7 001 31.0 69.0 001 1 1.8 18.7 8.6 25.6 23.4 20.5 0.3 1.1 001 22.4 77.6 001 deciles all sampled, well water well of faucet piped river population faucet piped building well water total Source Own Shared Own: Shared: Dug Spring, Rain Peddler otalT the Electricity in seY oN otalT *for 83 , 1997, Mean* 49.2 24.0 5.6 9.0 1.9 10.3 001 99.2 0.8 001 latoT 001 001 001 001 001 001 001 001 001 001 electricity decile and e 01 77.1 7.5 9.8 3.0 0.5 2.2 001 001 . 001 01 . 16.9 3.4 18.7 3.6 3.0 2.3 10.8 10.9 10.8 teraw of xpenditure 9 68.1 14.0 6.7 5.2 1.6 4.4 001 99.7 0.3 001 14.2 6.0 12.2 6.0 8.5 4.4 10.3 10.3 3.9 10.3 ce pita Sour­ ca 8 63.8 15.6 7.5 5.8 1.0 6.3 001 001 . 001 . 89 13.1 6.6 13.4 6.5 5.3 6.2 10.1 10.2 10.1 per yb 7 MANILA Decile 58.2 17.5 8.3 7.0 0.7 8.4 001 001 . 001 . Decile 1.51 7.1 14.2 7.5 3.5 7.9 9.7 9.8 9.7 PH14. 6 54.2 19.8 3.9 10.6 2.1 9.4 001 001 . 001 . 8.1 6.8 9.0 9.9 9.9 9.9 lebaT Expenditurea Expenditurea 10.9 1.61 1.11 Capit Capit 5 Per 45.4 567 29.3 4.0 8.9 2.5 9.9 001 99.6 0.4 001 9.0 Per 1.91 6.9 9.6 13.0 9.4 9.8 9.8 4.2 9.8 ithin W 4 37.7 31.8 5.0 9.8 2.5 13.2 001 99.4 0.6 001 Across 7.5 13.0 8.8 10.7 12.8 12.6 9.8 9.8 7.8 9.8 Distribution 3 37.6 31.3 5.4 10.1 1.7 13.9 001 99.3 0.7 001 Distribution 7.5 12.7 9.3 10.9 8.9 13.2 9.8 9.8 8.1 9.8 2 26.6 35.4 3.3 15.3 3.8 15.6 001 97.7 2.3 001 234 5.4 14.8 5.9 17.0 20.2 15.2 10.0 9.9 27.7 10.0 1 19.8 39.6 2.2 15.1 2.6 20.6 001 96.0 4.0 001 1 4.0 16.4 3.9 16.6 13.5 19.9 9.9 9.6 48.2 9.9 deciles all building building sampled, well the well the water well in water well in of faucet piped of faucet piped population faucet piped well faucet piped well Source Own Shared Own: Shared: Dug Peddler otalT Electricity seY oN otalT Source Own Shared Own: Shared: Dug Peddler Mean Electricity seY total oN Mean *for Annexes 84 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam (%) 1997 Mean* 37.3 21.4 13.2 15.0 5.7 2.0 0.1 5.3 001 90.3 9.7 001 electricity al/urban: Urban Poor 10.0 30.1 9.8 23.2 15.3 6.4 0.1 5.0 001 56.9 43.1 001 and , urr teraw tus sta of ty erv 40.1 20.5 13.6 14.1 4.8 1.6 0.1 5.3 001 93.7 6.3 001 ces Deciles Non-poor po h Sour­ eac Expenditure orf Mean 7.2 20.3 16.9 22.8 17.3 14.0 0.6 1.0 001 52.1 47.9 001 Capita Per PHILIPPINES Rural Poor 2.3 21.0 8.3 23.7 23.6 19.5 0.4 1.1 001 26.8 73.2 001 PH15. lebaT 9.5 19.9 20.9 22.4 14.3 1.41 0.7 1.0 001 63.9 36.1 001 Non-poor deciles all building building sampled, the well the in well in faucet piped river population faucet piped well water Electricity Own Shared Own: Shared: Dug Spring, Rain Peddler otalT Electricity seY total oN otalT *for 85 * (%), 1997 Mean 77.4 8.3 4.5 4.1 5.8 001 latoT 001 001 001 001 001 001 toilet 01 95.1 2.5 0.6 1.2 0.6 001 01 12.4 3.1 1.3 3.0 1.0 10.1 of al/urban: 90.8 4.1 1.3 2.9 0.9 001 1.81 4.9 2.9 7.3 1.6 10.1 ypeT­ r, ur 88.3 5.4 1.8 2.5 1.9 001 1.41 6.6 4.1 6.2 3.3 10.0 decile 85.9 6.5 2.4 2.6 2.6 001 1.11 7.8 5.3 6.3 4.5 10.0 e 81.8 7.3 3.1 3.9 3.8 001 10.6 8.9 7.0 9.5 6.7 10.0 Urban Urban PHILIPPINES 76.5 8.7 4.6 4.6 5.7 001 9.9 10.5 10.2 1.21 9.8 10.0 xpenditure 73.4 9.6 5.4 4.7 7.0 001 9.5 1.61 12.1 1.61 12.1 10.0 PH16. pita Deciles lebaT 67.2 1.11 7.0 5.7 8.9 001 Deciles 8.7 13.4 15.8 14.0 15.5 10.0 ca 23456789 per 62.0 13.3 8.5 5.6 10.6 001 8.0 16.0 19.0 13.7 18.3 10.0 yb Expenditurea 1 52.2 14.5 10.2 7.1 16.1 001 Expenditurea 123456789 6.6 17.2 22.3 17.1 27.2 9.8 * Capit Mean 52.0 15.4 14.3 3.4 14.9 001 Capit latoT 001 001 001 001 001 001 Per Per 01 82.7 6.5 4.1 1.7 5.0 001 01 15.3 4.0 2.8 4.9 3.3 9.6 Within 70.5 10.3 7.5 3.1 8.6 001 Across 13.6 6.7 5.2 9.3 5.8 10.0 64.4 13.1 9.1 2.5 10.8 001 12.4 8.5 6.4 7.5 7.3 10.0 Distribution 57.1 13.7 13.6 3.8 1.71 001 Distribution 1.11 9.0 9.6 1.21 7.9 10.1 51.0 15.7 13.9 4.0 15.5 001 9.9 10.3 9.8 1.91 10.5 10.1 Rural Rural 48.5 17.5 16.5 3.7 13.9 001 9.4 1.41 1.61 10.9 9.3 10.0 45.5 19.1 16.8 2.8 15.7 001 8.8 12.5 1.81 8.5 10.6 10.1 39.1 20.7 17.1 4.0 19.1 001 7.6 13.5 12.1 1.81 12.9 10.1 34.1 19.0 20.7 4.3 21.8 001 deciles 6.6 12.5 14.6 13.0 14.8 10.1 or 123456789 27.9 18.3 23.5 3.7 26.6 001 123456789 5.3 1.71 16.2 10.9 17.6 9.9 levels all sampled, toilet pit pile toilet population pile of pit pit of pit total ypeT ater-sealedW Closed Open Others: None otalT ypeT ater-sealedW Closed Open Others: None otalT for* Annexes 86 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam (%), 1997 Mean* 84.6 7.2 1.4 5.0 1.8 001 latoT 001 001 001 001 001 001 , toilet . 1997 01 0.82 0.92 0.38 0.44 0.66 0.10 0.14 95.1 3.1 0.3 1.5 001 . 01 12.1 4.6 2.3 3.3 lesba 10.8 Mean* of decile: dur e ypeT­ 9 90.7 3.6 0.8 3.9 0.9 001 9 1.01 5.2 5.7 8.1 5.3 10.3 decile: Decile Decile e 01 0.87 0.98 0.73 0.76 0.93 0.47 0.54 8 91.6 4.4 0.7 3.2 0.1 001 8 10.9 6.2 5.3 6.5 0.6 10.1 consumer xpenditure 7 89.1 6.1 0.6 3.2 1.1 001 7 10.2 8.2 4.0 6.2 5.7 9.7 0.89 0.96 0.64 0.59 0.84 0.22 0.29 MANILA pita Expenditurea Expenditure xpenditure ca 6 86.4 6.5 0.6 5.5 1.1 001 6 8.9 4.1 5.9 9.9 Capit Capita 10.1 10.8 pita PH18. per Per 0.89 0.98 0.58 0.58 0.84 0.12 0.19 lebaT yb 5 83.8 7.5 1.7 5.8 1.2 001 Per 9.7 10.2 12.2 1.31 6.6 9.8 Households'­ ca 89 per Within 4 83.9 7.4 1.5 5.7 1.5 001 Across 9.7 10.1 10.3 1.31 8.1 9.8 yb 0.84 0.96 0.48 0.51 0.80 0.07 0.09 3 80.2 8.9 2.3 6.1 2.5 001 MANILA 345 9.2 12.0 16.2 12.0 13.6 9.8 Distribution 2 73.9 13.3 2.1 6.7 4.0 001 Distribution 8.7 18.5 15.2 13.4 22.5 10.0 PH19. 0.85 0.95 0.39 0.50 0.72 0.07 0.07 1 70.5 1.71 3.5 8.6 5.7 001 12 8.3 16.1 24.8 17.1 31.6 9.9 lebaT sampled 0.82 0.91 0.28 0.40 0.67 0.01 0.05 toilet toilet of pit pile population pit of pit pile pit 4567 total 0.81 0.91 0.29 0.35 0.61 0.01 0.03 ypeT ater-sealedW Closed Open Others: None otalT ypeT ater-sealedW Closed Open Others: None otalT *for ­ 0.78 0.91 0.22 0.33 0.50 0.01 0.02 (%), 8.3 4.5 4.1 5.8 1997 Mean* 77.4 001 otalT 001 001 001 001 001 001 5.5 9.2 0 toilet Urban Poor 46.4 15.3 13.8 5.7 18.8 001 Urban Poor 17.1 28.3 13.0 29.9 0.78 0.86 0.09 0.20 0.34 0.01 PHILIPPINES of urban:/ Status ypeT al Status urr, 80.6 7.6 3.5 3.9 4.5 001 94.5 82.9 71.7 87.0 70.1 90.8 Non-poor Poverty Non-poor 123 0 0.69 0.80 0.05 0.13 0.28 0.01 Poverty PH17. lebaT tus sta ithinW Mean 52.0 15.4 14.3 3.4 14.9 001 Across Mean 001 001 001 001 001 001 ty deciles Rural Poor 32.6 18.1 20.4 3.7 25.2 001 Rural Poor 19.9 37.1 45.2 35.2 53.5 31.7 all erv Distribution po Distribution set recorder motorcycle yb 60.9 14.2 1.51 3.2 10.2 001 or 80.1 62.9 54.8 64.8 46.5 68.3 tape sampled, Non-poor Non-poor sampled condition jeep,, radio television video stereo refrigerator air car population owns owns owns owns owns owns owns toilet pit pile toilet pile population total of pit pit of pit total Family Family Family Family Family Family Family *for ypeT ater-sealedW Closed Open Others: None otalT ypeT ater-sealedW Closed Open Others: None Mean *for 87 (%) 1998 Mean* 37.8 62.2 001 0.6 13.2 29.9 25.0 18.0 13.3 5.1 19.8 21.0 18.1 8.7 7.6 1.41 8.4 01 42.5 57.5 7.9 0.6 10.7 44.2 19.9 12.6 12.1 0.6 7.8 14.6 15.8 14.1 3.6 13.3 30.2 acteristics al/urban: 36.4 63.6 9.0 0.6 13.3 37.1 21.4 19.9 7.7 3.3 8.5 20.8 20.1 1.31 6.8 18.7 10.6 harc urr, 27.5 72.5 9.1 0.2 1.91 28.9 29.8 17.4 1.81 1.4 12.2 16.2 16.1 10.7 7.9 15.8 19.7 . Deciles decile 43.0 57.0 10.2 8.4 35.3 19.2 21.9 15.2 3.0 22.9 19.9 14.7 1.51 8.8 10.9 3.8 e 34.3 65.7 9.5 0.8 1.31 24.9 34.5 12.0 16.5 4.7 15.3 24.9 13.9 1.31 10.5 13.0 3.6 Household­ Urban . Expenditurea 42.4 57.6 10.0 15.5 22.1 19.8 27.2 15.4 3.7 21.1 23.7 16.7 6.6 9.2 14.3 6.4 AM xpenditure Capit 36.0 64.0 10.2 0.9 9.3 25.7 26.6 18.4 19.1 6.7 23.3 19.7 14.9 1.21 8.7 6.5 1.9 VIETN pita Per 36.8 63.2 10.3 1.7 14.2 30.0 26.9 16.0 1.21 4.9 22.1 17.3 23.4 7.5 7.8 14.4 5.2 ca . VN1. per 46.8 53.2 1.01 1.51 25.6 31.5 17.7 13.6 1.21 23.7 24.7 21.8 5.4 5.3 6.4 6.1 lebaT yb . 123456789 32.9 67.1 13.0 1.1 22.5 29.2 21.0 15.9 10.3 8.2 32.5 24.7 21.5 1.7 6.6 4.8 Mean* 16.9 83.1 001 1.3 20.6 33.5 19.9 14.8 9.9 8.8 29.7 23.7 23.6 4.1 4.1 4.8 0.1 01 22.5 77.5 7.3 0.8 12.3 28.4 28.6 17.4 12.5 5.3 22.1 21.1 22.2 7.9 6.4 10.2 9.4 17.5 82.5 8.2 1.2 15.9 30.9 21.7 17.9 12.4 4.2 24.9 25.7 25.8 5.0 4.6 5.8 3.1 16.8 83.2 8.6 1.0 12.7 32.7 23.8 18.1 1.71 5.9 30.1 22.8 24.3 2.5 5.2 1.7 1.2 Deciles 16.3 83.7 9.4 1.0 16.3 34.5 19.8 16.7 1.71 6.5 27.5 23.2 26.9 3.4 5.9 6.5 1.1 17.2 82.8 9.5 1.1 19.6 36.5 19.1 15.4 8.2 6.6 29.8 23.1 25.2 6.7 5.4 6.2 6.0 Rural Expenditurea 16.9 83.1 10.3 2.0 24.9 26.5 21.1 15.0 10.5 5.6 31.4 24.5 26.2 4.1 3.0 7.4 6.0 Capit 17.2 82.8 1.01 1.2 21.2 33.9 17.8 15.7 10.2 7.6 30.7 25.0 23.9 3.4 4.1 8.4 5.0 Per 15.4 84.6 1.61 0.7 20.9 36.7 18.8 14.2 8.7 8.7 31.9 22.3 24.9 4.0 3.2 4.4 6.0 15.5 84.5 1.81 2.8 26.3 34.1 17.3 1.41 8.2 13.1 30.8 24.6 22.5 3.6 3.2 3.2 . 123456789 15.5 84.5 12.4 1.2 28.1 37.7 15.9 9.6 7.5 19.3 33.7 24.4 16.5 2.3 1.7 7.1 4.0 deciles all head sampled, the headship head old old old old old old of the population of years years years years years years I I II III SC total Household Female-headed household Male-headed household otalT Age 15-24 25-34 35-44 45-54 55-64 65> Education Never Cap< Cap Cap Cap Nghe THCN DHCD *for Annexes 88 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam decile 1998 8.37 Mean* 37.84 62.16 0.61 13.18 29.89 25.05 17.96 13.30 5.08 19.84 20.96 18.09 8.73 7.56 1.391 e household tus: . sta Urban Poor 27.63 72.37 1.60 24.01 30.74 18.87 18.26 6.52 8.73 34.31 27.75 18.48 1.14 5.25 4.33 urban ty xpenditure ver and po 38.87 61.13 0.51 12.09 29.80 25.67 17.93 13.99 4.71 18.37 20.27 18.05 9.49 7.79 12.10 9.21 al pita urR­ ca and Non-poor per 1 AM yb Mean 16.86 83.14 1.33 20.59 33.49 19.90 14.77 9.92 8.83 29.75 23.73 23.64 4.1 4.08 4.83 1.03 VIETN (%), Rural Poor 15.59 84.41 1.43 25.35 34.68 17.43 12.5 8.61 12.46 32.24 23.82 21.71 3.24 2.84 3.40 0.28 VN2. lebaT 1 acteristics 17.91 82.09 1.24 16.61 32.51 21.97 16.66 1.011 5.79 27.67 23.66 25.26 4.84 5.1 6.02 1.66 deciles Non-poor all harc head sampled, the headship head old old old old old old of the population of years years years years years I years I II III SC total Household Female-headed household Male-headed household Age 15-24 25-34 35-44 45-54 55-64 65+ Education Never Cap< Cap Cap Cap Nghe THCN DHCD *for 89 * . 0 0 (%), 0.7 1.6 1998 Mean 91.7 1.0 1.0 7.3 0.6 2.0 Mean Mean 86.0 63.7 01 . 90.0 10.0 ethnicity decile: . Poor 0.71 1.43 Poor 89.3 73.9 e 92.6 7.4 . urban 94.1 5.9 .... .... .... .... .... .... 0.70 1.62 83.1 62.6 and xpenditure Deciles 85.1 14.5 0.4 Non-poor Non-poor al . . . . . urR­ 87.1 5.0 12.0 4.0 pita Urban . . ca Expenditurea 89.8 9.5 0.7 AM per .. . expenditure) yb Capit 93.8 5.3 0.4 0.5 total VIETN Per 93.1 2.0 . . expenditure of here 5.0 1.6 % . .... . Health (as household Rural Urban Born Rural Urban VN3. 23456789 92.1 5.1 2.5 0.3 lebaT 1 97.4 1.0 ..... . ..... 0.5 1.0 0.6 0.4 ­ 1.9 4.3 3.1 1.0 2.1 1.5 * AM (%), 1998 otalT 85.1 85.4 85.3 67.1 74.0 70.4 otalT Mean 81.6 4.2 1.4 0.5 2.5 3.1 2.4 0.1 5.2 VIETN 01 96.9 3.0 . . 2.6 4.9 3.7 1.5 2.2 1.8 0.1 0.8 0.8 0.6 0.6 88.5 89.3 88.9 79.4 81.5 80.4 al/urban: Poor Poor 93.1 8.1 . VN4. acteristics 0.4 0.7 2.3 0.3 0.7 0.6 urr, 93.7 4.1 . . 1.3 3.8 0.9 2.1 0.4 1.2 0.4 1.1 1.9 lebaT harc tus 82.2 82.2 82.2 65.8 73.3 69.4 2.5 1.5 Non-poor Non-poor Deciles 90.1 6.0 0.2 0.4 2.7 2.2 2.3 1.0 sta 1.3 vidual ty 92.4 9.2 0.4 0.5 1.7 0.1 1.5 2.0 Indi v er 0.3 Rural po Expenditurea 84.8 1.2 1.8 1.3 2.2 2.4 2.6 3.0 2.5 yb Female Male Mean Female Male Mean Female Male otalT Female Male otalT Capit 76.0 4.4 2.6 0.3 3.3 4.1 4.5 2.0 4.5 Per 79.0 7.3 2.1 0.3 2.8 3.4 4.4 1.1 3.3 72.6 4.2 1.8 0.2 2.5 5.0 3.1 1.3 9.2 job 123456789 54.8 0.3 3.1 0.5 3.9 8.7 1.9 1.3 deciles 20.9 all job secondary Has Rural Urban Has Rural Urban sampled, (Kinh) group population Ethnic ietnameseV uaT Thai Chinese Khmer Muong Nung oeM total Other *for Annexes 90 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam (%), 1998 nae 0 M 73.0 10.9 16.1 80.5 9.0 10.0 0.5 nae M 51.0 49.0 52.3 47.7 latoT 001 001 001 001 enderg al/urban: 01 . 7.6 8.4 6.0 9.8 1.1 01 01 8.3 9.0 8.6 7.8 and elsvle urr, 84.0 83.2 48.9 51.1 54.6 45.4 . y decile 80.0 9.5 10.4 87.1 2.9 9.3 0.7 50.3 49.7 51.0 49.0 8.7 8.9 8.8 9.3 ac e liter . 89 79.0 9.3 1.71 84.8 6.7 7.4 1.1 89 Deciles 50.8 49.2 51.5 48.5 Deciles 89 9.1 9.2 9.2 9.5 vidual e xpenditur . Indi­ pita Deciles 77.5 10.0 12.4 78.3 1.01 9.9 0.8 50.3 49.7 52.2 47.8 9.0 9.3 ca Expenditurea Expenditurea 10.1 10.1 AM per Capit Capit VIETN yb 81.3 7.7 10.9 0.1 84.6 5.7 8.1 1.6 50.5 49.5 51.8 48.2 9.5 9.7 9.6 9.8 Expenditurea Per Per ithin VN5. Capit . W Across 76.3 1.01 12.8 81.4 10.3 7.7 0.5 51.5 48.5 53.5 46.5 10.2 9.9 10.1 9.6 lebaT Per by Gender Gender . . by by 4567 4567 4567 Levels 74.8 1.81 13.5 80.2 9.1 10.7 50.6 49.4 51.3 48.7 10.3 10.5 10.0 10.4 Literacy 72.6 10.2 17.1 0.1 81.2 8.7 9.8 0.2 Distribution Distribution 51.1 48.9 53.6 46.4 10.9 10.8 10.5 10.0 . 65.1 14.5 20.4 77.1 1.41 1.41 0.2 51.9 48.1 50.1 49.9 1.31 10.9 10.3 1.31 . 123 48.9 15.0 36.1 0.1 75.0 12.1 12.8 123 52.9 47.1 53.8 46.2 123 12.7 1.71 12.8 12.1 ficulty language ficulty language read dif dif Can seY With oN ferent Dif seY With oN ferent Dif Gender Female Male Female Male Gender Female Male Female Male Rural Urban Rural Urban Rural Urban 91 , male tus 1998 Mean 31.1 68.9 24.9 75.1 25.2 74.8 25.0 75.0 20.5 79.5 20.6 79.4 otalT 001 001 001 001 001 001 001 001 001 001 001 001 Sta Status Status and ty emaleF­ ervoP Poverty Poor 28.3 71.7 23.3 76.7 23.5 76.5 16.7 83.3 16.0 84.0 16.0 84.0 Poverty Poor 34.2 39.3 32.7 35.7 32.8 35.8 8.3 13.9 5.2 7.1 5.3 7.3 al/urban: yb ur r within Across AM (%), 32.9 67.1 25.8 74.2 26.0 74.0 26.2 73.8 20.8 79.2 21.0 79.0 65.8 60.7 67.3 64.3 67.2 64.2 91.7 86.1 94.8 92.9 94.7 92.7 s Non-poor Non-poor Distribution Distribution VIETN vior VN7. beha lebaT smoking smoking smoking smoking smoking smoking smoking smoking smoking smoking smoking smoking not smoking not smoking not smoking not smoking not smoking not smoking not smoking not smoking not smoking not smoking not smoking not smoking smoking Female Male otalT Female Male otalT Female Male otalT Female Male otalT Rural Urban Rural Urban tus sta 1998 001 001 001 001 Mean 57.3 42.7 62.3 37.7 otalT Health­ 57.7 42.3 54.2 45.8 43.9 43.1 7.7 10.8 al/urban: Poor Poor AM urr, Status tus Status 56.9 43.1 63.1 36.9 56.1 56.9 92.3 89.2 VIETN sta Non-poor Poverty Non-poor Poverty ty VN6. erv within between lebaT po yb sick sick sick sick not sick not sick not sick not sick Distribution asW asW asW asW Distribution asW asW asW asW Rural Urban Rural Urban Annexes 92 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam , * , 3.0 1998 Mean 16.2 5.2 6.8 7.6 50.0 13.9 001 1998 latoT 001 001 001 001 001 001 001 decile e 01 0.3 43.9 3.0 9.7 8.9 30.4 1.1 001 decile e 01 82.0 21.4 4.5 1.21 9.1 4.8 0.6 . al/urban: 32.5 2.9 13.7 5.1 43.6 2.2 001 decile: . h 18.0 4.9 18 5.9 7.8 1.4 xpenditure urr, 6.0 28.5 5.1 10.0 7.3 46.8 1.7 001 xpenditure eac 8.9 8.6 8.5 1.1 . in 18.0 16.0 13.4 pita Deciles 26.3 6.6 6.6 8.2 46.2 6.1 001 . pita Deciles 16.6 12.9 9.9 10.9 9.4 4.4 ca deciles . . 16.7 5.6 1.9 ca types 9.9 9.1 7.2 per all 10.3 47.7 10.5 001 10.3 12.7 12.9 Urban all Urban h f or Expenditurea 12.7 10.3 5.6 8.0 51.9 10.6 001 per yb orf Expenditurea 7.8 19.7 9.6 10.5 10.4 7.6 eac esg e 456789 4.9 7.9 0.5 456789 3.1 7.6 8.8 Capit 1.41 58.7 12.0 001 esg g Capit 15.5 15.3 12.0 within centa Per 3 5.6 6.5 7.7 5.8 60.8 13.6 001 centa centa Per 3 3.5 12.8 1.61 7.8 12.5 10.0 esg per 2 3.7 1.5 9.2 3.8 62.7 25.5 001 per per 2 2.5 3.2 4.7 5.4 13.7 20.1 ..... ..... centa mean 1 1.4 2.9 7.0 8.0 45.5 41.5 001 mean 1 1.1 7.2 1.4 13.6 1.81 38.6 per * and 0 0.5 1.3 3.0 housing: and 6.7 Mean 61.8 29.4 001 of latoT 001 001 001 001 001 001 001 01 1.0 3.4 2.3 6.1 9.4 69.1 14.1 001 housing: . of 0.5 4.7 7.0 9.5 65.5 19.2 001 ypesT­ 01 001 54.9 12.7 42.9 10.2 8.2 3.5 . 9.0 29.3 20.8 1.51 8.7 5.3 . 0.5 0.8 8.5 AM . ypesT­ 70.5 19.6 001 10.1 5.4 10.9 9.8 5.7 . Deciles 3.2 8.0 7.6 67.5 20.9 001 . VIETN Deciles 23.1 27.7 10.7 10.2 6.7 AM . 0.7 8.3 64.7 26.3 001 . 5.2 9.9 8.5 Rural . VN9. 1.71 Rural VIETN Expenditurea 0.5 7.0 61.0 31.6 001 . lebaT Expenditurea 3.6 10.7 10.1 1.11 456789 0.4 2.1 6.6 VN8: Capit 61.5 29.4 001 456789 8.7 Capit 17.6 10.8 1.01 1.01 lebaT . Per 3 0.2 7.3 61.2 31.3 001 . Per 3 1.7 12.5 1.41 12.3 ....... ..... 2 0.5 0.2 4.7 59.7 35.0 001 ....... ..... 2 12.0 1.4 8.2 1.41 14.1 . . 1 0.2 2.0 1.6 46.3 51.8 001 deciles . . all 1 5.3 6.8 2.9 9.2 deciles 21.8 all 1T 2T sampled, 1T 2T sampled, house house toilet toilet house house house toilet house house house house toilet house house population house house house population of house+garden private separate total of house+garden private separate total ypeT City Multistoried Multistoried One-story with One-story with Semi-permanent emporaryT otalT *for ypeT City Multistoried Multistoried One-story with One-story with Semi-permanent emporaryT *for 93 , , 1998 0.3 Mean* 16.2 5.2 6.8 7.6 50.0 13.9 001 otalT 001 001 001 001 001 001 001 001 localities households . urban poor localities: Urban Poor 1.3 2.4 0.5 6.3 42.7 46.8 001 . Urban Poor 0.7 4.3 0.6 7.6 7.8 30.9 9.2 orf orf both and and f or al urr 0.3 17.7 5.5 7.5 7.8 50.7 10.6 001 001 99.3 95.7 99.4 92.4 92.2 69.1 90.8 orF­ means Non-poor Non-poor non-poor AM orf and 0 0.5 1.3 0.3 6.7 (%) Mean* 61.8 29.4 001 otalT 001 001 001 001 001 001 001 001 VIETN housing . VN10. 23 of Rural Poor 0.3 0.4 0.1 4.7 56.3 38.2 001 . Rural Poor 25.9 15.2 8.6 41.4 59.2 45.5 lebaT types 0 0.6 2.0 0.5 8.4 66.5 22.0 001 001 86 74.1 84.8 91.4 58.6 40.8 54.5 Non-poor Non-poor toilet toilet toilet toilet 1T 2T private separate with with 1T 2T private separate sampled house with with house house house house house house house house house house house house house population of house+garden of house+garden total ypeT City Multistoried Multistoried One-story One-story Semi-permanent emporaryT otalT ypeT City Multistoried Multistoried One-story One-story Semi-permanent emporaryT otalT *for Annexes 94 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam , , 1.4 4.3 1998 Mean* 74.9 5.1 1.6 1.0 4.5 1.9 5.3 001 1998 Mean* 24.7 26.2 35.5 6.8 0.3 0.1 6.3 2.0 001 decile decile e 01 . . . . 7.4 87.9 3.1 0.5 1.1 001 e 01 . . . 38.4 1.61 41.7 7.1 1.2 001 . . . al/urban: 8.9 85.2 3.9 1.0 0.3 0.7 001 . . al/urban: 32.3 1.91 50.5 4.4 0.6 0.3 001 xpenditure urr, . . . 9.1 86.0 1.4 0.8 0.8 1.9 001 xpenditure urr, . . . 29.9 18.9 39.5 10.5 1.2 001 pita Deciles 0.7 3.5 85.1 4.4 0.2 1.3 2.7 0.6 1.4 001 . . . pita Deciles 32.5 17.9 42.7 5.0 2.0 001 ca deciles . deciles . . . 1.1 6.3 78.4 4.4 0.5 6.3 0.7 2.4 001 ca 31.0 22.5 36.9 7.5 2.1 001 per Urban Urban h total . . . . Expenditurea 1.6 3.2 76.2 9.4 0.8 3.3 0.4 5.2 001 per h total 5.6 4.7 001 eac orf Expenditurea 28.9 30.1 30.8 2.5 1.9 4.9 2.3 0.5 6.7 0.9 3.3 eac orf . orf esg Capit 77.1 001 orf esg Capit 23.8 33.7 31.6 3.2 0.7 6.2 7.0 001 Per (%) 3456789 2.5 2.4 72.7 6.7 1.3 1.3 5.4 2.0 5.7 001 . . Per 9.2 3.7 4.0 centa (%) 3456789 18.3 39.1 29.3 001 allsw per 2 0.4 2.8 60.6 6.5 2.6 1.9 7.7 6.8 10.6 001 centa . . 5.2 0.3 oofsr per 2 10.1 36.4 32.8 15.1 001 of 1 3.9 0.6 52.7 5.6 5.7 2.4 6.9 4.8 17.3 001 . of 1 1.71 31.9 25.6 9.6 0.8 20.0 5.0 001 mean mean 0 ypesT­ and 1.5 0.8 Mean* 49.0 5.9 7.0 0.7 14.0 8.4 12.7 001 ypesT­ and 7.5 Mean* 53.3 1.01 3.9 0.1 0.4 23.7 001 01 . . 0.4 2.2 AM 66.2 4.9 1.1 0.6 16.5 2.9 5.2 001 01 AM 12.5 49.8 22.1 5.7 0.3 9.5 001 0.7 2.6 61.5 3.4 3.3 1.1 15.2 2.6 9.4 001 . 13.2 51.8 20.1 4.7 0.4 9.8 001 VIETN 0 0 0.8 57.7 4.2 2.8 0.6 18.6 3.8 1.31 001 VIETN 7.8 58.8 13.2 5.0 0.1 15.0 001 Deciles 0.6 0.8 VN11. 60.1 6.6 5.9 1.7 9.3 3.6 1.41 001 . ... 9.6 5.8 0.2 VN12. Deciles 57.8 10.9 15.6 001 lebaT 1.6 0.8 51.7 7.7 5.0 1.1 13.2 3.9 15.0 001 8.0 56.7 10.8 3.1 0.1 0.3 20.5 5.0 001 Rural Rural Expenditurea 0.7 0.6 47.8 7.4 6.5 0.4 13.4 7.2 15.9 001 lebaT . Expenditurea 7.7 53.5 10.4 4.4 0.1 23.8 001 . 456789 3.2 0.4 Capit 46.2 5.3 7.9 0.3 14.9 7.9 13.9 001 456789 8.3 Capit 53.9 8.0 4.8 0.8 24.2 001 Per 3 1.2 0.5 47.7 5.8 5.1 0.4 13.4 1.11 14.8 001 Per 3 7.3 56.9 7.3 3.5 0.2 0.4 24.4 001 2 1.7 0.3 39.4 8.0 10.5 0.4 13.2 10.3 16.3 001 .... 2 3.9 51.6 8.7 2.9 0.1 0.6 32.2 001 . 1 3.3 0.2 27.8 4.8 16.2 0.4 13.7 23.2 10.3 001 deciles all 1 1.5 deciles 44.1 6.3 0.7 0.2 0.9 46.4 001 all sampled, sampled, walls population roof population of stone brick total of bamboo leaves, total ypeT Other Concrete Brick, Unfired Earth Iron oodW Bamboo Leaves/branches otalT *for ypeT Cement ileT Iron Panles Canvas Wood, Straw Other otalT *for 95 2.68 2.07 9.46 Mean* Mean* 10.37 Mean* 14.36 15.07 , Poor 3.06 3.27 Poor 7.66 6.00 Poor 1.381 8.85 1998 2.32 1.92 acteristics 10.97 10.82 16.84 15.70 deciles Non-poor Non-poor Non-poor all harc al/urban: urr , sampled, pp tus pp person m2 m2 Household­ sta per population area, ty area, total AM erv Rooms Rural Urban Living Rural Urban otalT Rural Urban *for po yb VIETN 2.37 1.61 2.20 2.68 2.07 5.55 5.22 VN13. Mean* Mean* Mean* lebaT 2.89 3.05 2.90 3.06 3.27 6.08 6.76 Poor Poor Poor 1 1.93 1.46 1.78 2.32 1.92 5.1 5.07 Non-poor Non-poor Non-poor children 0.455 0.092 0.374 with hhs (%) total among Count population children, children Head of of total size Poverty Rural Urban Mean, Number Rural Urban otalT Number Rural Urban HH Rural Urban Annexes 96 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam . (%) . . 1998 Mean 98.2 0.2 1.6 001 latoT 001 001 001 (%) 1998 34 01 3 Mean 4.8 13.6 12.9 0.1 5.1 3.1 4.5 001 01 001 . . . 001 01 . . . 8.0 lighting teraw of al/urban: 001 . . . 001 . . . 9.1 urr, al/urban: . . 001 . . . 001 . . . 41 4.9 9.3 Urban Poor 8.2 8.3 5.7 13.7 27.2 7.4 10.6 001 ypesT­ drinking , urr . . of tus decile 99.5 0.5 001 . . 10.3 25.0 AM e . . sta 99.6 0.4 001 . . 9.7 2.5 ces . 4.7 0.1 2.5 4.2 2.9 3.9 Urban . Urban ty 46.5 10.2 13.6 1.51 001 VIETN 99.2 0.8 001 . 10.1 5.0 Sour­ ervpo Non-poor xpenditure 99.1 0.4 0.5 001 10.3 18.8 3.1 AM h VN14. eac 5 1.1 0.5 0.3 0.5 0.2 lebaT pita Deciles 3456789 1.2 1.8 Deciles 3456789 13.5 41.4 10.3 12.6 14.7 001 97.0 001 10.1 56.3 1.71 VIETN ca orf Mean . . 2 per 95.5 4.5 001 2 10.7 30.9 yb . ..... Expenditurea 1 94.3 5.7 001 . ..... VN15. Expenditurea 1 12.4 46.8 8 lebaT Rural Poor 0.8 0.3 0.3 7.9 43.2 0.5 21 10.2 16.6 0.2 001 Capit Mean 70.5 1.7 26.5 1.2 001 Capit latoT 001 001 001 001 Per Per 01 93.0 2.0 4.5 0.5 001 01 9.6 8.4 1.3 3.1 1.3 0.7 0.3 18.1 39.8 0.5 8.9 14.6 13.1 0.2 2.5 001 Within 85.7 3.1 01 1.3 001 Across 9.9 14.5 3.1 8.6 Non-poor 77.3 2.8 19.2 0.7 001 9.4 14.1 6.2 4.6 Distribution 83.3 0.8 15.5 0.4 001 Distribution 1.11 4.3 5.5 2.9 73.2 2.1 22.5 2.2 001 9.8 1.71 8.0 17.0 Rural Rural water pump 69.6 2.4 27.6 0.4 001 10.2 14.4 10.7 3.5 tap tap with water well pond 456789 69.8 2.2 25.6 2.3 001 drinking well water 456789 10.9 14.2 10.7 20.3 of private private standpipe spring drilled well water lake,, 3 68.1 1.0 29.6 1.3 001 3 1.21 6.8 12.9 12.0 Source Inside Outside Public Deep Hand-dug Filtered Other Rain River Container Other otalT 2 57.6 0.1 41.6 0.7 001 2 9.7 0.9 18.6 6.2 1 74 1.5 49.3 2.2 001 1 8.2 10.8 23.0 21.9 lamp lamp lighting lamp kerosene lighting lamp kerosene of oil, of oil, ypeT Electricity Battery Gas, Other otalT ypeT Electricity Battery Gas, Other 97 . (%) 1998 Mean* 43.0 10.0 4.8 13.6 12.9 1.0 3.0 5.1 3.1 4.5 001 (%) 1998 Mean* 61.5 6.9 15.4 2.5 1.0 12.7 001 teraw 01 . . . 77.0 6.7 2.0 10.8 1.3 2.1 001 01 . toilets 94.8 2.1 0.9 1.1 1.1 001 . . al/urban: 76.3 6.5 3.2 9.9 1.0 1.2 1.9 001 of al/urban: 90.8 3.4 2.9 0.3 0.4 2.3 001 drinking urr, . 62.1 7.4 3.6 13.9 5.1 3.2 1.5 3.3 001 of .... .... Deciles ypesT­ , urr 81.5 6.0 3.3 1.1 0.8 7.4 001 59.2 10.6 3.5 14.9 7.0 1.4 1.6 0.8 1.0 Deciles 4.8 3.1 1.2 1.3 8.2 ces decile 001 decile 81.4 001 e 51.6 10.3 3.4 12.0 12.0 6.0 . 3.1 2.3 1.8 2.9 001 AM e 68.0 7.8 12.1 3.1 1.5 7.4 001 Sour­ Urban Urban Expenditurea 42.3 1.11 6.9 18.6 1.51 1.2 3.9 2.6 1.9 001 VIETN Expenditurea 61.9 14.1 8.3 4.5 1.3 9.9 001 AM xpenditure Capit 30.0 15.1 5.1 17.2 20.6 3.1 3.9 1.5 3.6 001 xpenditure 9.0 1.4 2.0 001 VN17. Capit 51.9 20.7 15.1 VIETN pita Per 26.0 10.1 4.5 12.9 18.7 4.6 1.31 5.2 6.7 001 pita Per 46.5 8.0 26.4 2.8 1.2 15.2 001 ca ca 20.5 5.7 13.9 19.3 2.7 9.2 9.5 7.3 001 lebaT 3.3 0.6 001 VN16. per 1.91 ..... ..... per 37.8 10.6 28.1 19.6 lebaT h . 123456789 10.2 9.3 8.0 1.61 23.9 7.1 12.2 6.4 1.31 001 h 123456789 26.1 2.8 35.0 4.7 31.4 001 eac eac 1.1 0.5 0.3 Mean* 13.5 41.4 0.5 10.3 12.6 14.7 0.2 5.0 001 3.7 Mean* 10.2 46.2 12.4 5.2 22.3 001 within 01 . 2.4 1.5 0.2 30.3 30.0 0.1 1.71 12.9 8.9 2.0 001 within 01 22.6 15.9 36.6 1.91 5.4 7.7 001 . 2.3 1.5 0.4 19.6 33.9 1.0 8.6 16.8 13.6 2.2 001 7.1 17.1 40.0 15.5 4.6 15.8 001 . . 1.5 0.5 17.2 41.1 8.8 13.7 15.3 0.2 1.6 001 4.6 17.2 39.0 17.7 5.2 16.4 001 . Deciles 1.0 0.7 16.0 42.9 0.4 7.7 18.6 10.8 0.2 1.6 001 Deciles 4.2 14.9 47.2 10.8 5.7 17.2 001 1.1 0.2 0.2 15.4 41.6 0.5 9.7 1.61 15.4 0.3 4.0 001 2.4 1.81 43.3 13.0 4.6 24.9 001 Rural Rural Expenditurea 0.4 0.2 0.7 12.6 44.2 0.8 8.9 1.21 16.5 0.2 4.3 001 Expenditurea 2.4 9.5 47.2 13.3 4.0 23.6 001 0.7 0.4 0.5 51 Capit 10.2 45.6 0.5 9.6 12.3 0.2 5.0 001 0.8 8.3 Capit 50.6 14.1 4.8 21.4 001 Per 0.2 0.2 0.4 9.9 42.5 0.3 9.1 14.3 16.9 0.5 5.6 001 Per 0.6 6.7 54.4 12.0 4.8 21.4 001 . 1.2 0.2 8.9 45.1 0.4 13.7 10.0 13.7 0.1 6.8 001 0.3 5.0 47.6 1.51 6.0 29.6 001 . . 123456789 0.7 0.4 0.4 5.0 41.0 0.7 13.6 7.3 18.1 12.8 001 deciles all 123456789 3.2 deciles 48.8 6.8 6.2 35.1 001 all sampled, sampled, water tap tap water well well pond water tank private private standpipe spring population toilet latrine water population drilled pump well water lake,, of toilet vault toilet directly type drinking total septic the toilet total Source of Inside Outside Public Deep with Hand-dug Filtered Other Rain River Container Other otalT *for ypeT Flush with Double compost Simple oiletT over Other No otalT *for Annexes 98 Urban Poverty in East Asia A review of Indonesia, the Philippines, and Vietnam (%) 1998 Mean 61.5 6.9 15.4 2.5 1.0 12.7 001 otalT 001 001 001 001 001 001 toilet . of 2.9 2.7 001 . 3.5 3.9 9.9 al/urban: Urban Poor 23.6 32.2 38.6 Urban Poor 19.2 27.8 ypeT­ urr, Status tus 65.4 7.3 13.7 2.4 1.1 10.1 001 Status 96.5 96.1 80.8 90.1 001 72.2 AM sta Non-poor Non-poor Poverty Poverty ty VIETN erv within 3.7 5.2 001 across 001 001 001 001 001 001 po Mean 10.2 46.2 12.4 22.3 otalT VN18. yb lebaT Distribution 0.6 5.3 49.7 1.11 5.5 27.7 001 Distribution 7.9 94 23.9 40.7 48.5 56.5 Rural Poor Rural Poor 6.2 14.3 43.3 13.5 4.9 17.8 001 15 92.1 76.1 59.3 51.5 43.5 Non-poor Non-poor tank tank toilet latrine water toilet latrine water of toilet vault toilet septic directly the type of toilet vault toilet toilet septic directly the type toilet ypeT Flush with Double compost Simple oiletT over Other No otalT ypeT Flush with Double compost Simple oiletT over Other No 99