Presentation on theme: "The Spatial Distribution of Urban Poverty: Metropolitan Areas versus Small Towns Peter Lanjouw (World Bank) Urbanization and Poverty and Poverty Reduction:"— Presentation transcript:
The Spatial Distribution of Urban Poverty: Metropolitan Areas versus Small Towns Peter Lanjouw (World Bank) Urbanization and Poverty and Poverty Reduction: Bridging Urban and Rural Perspectives World Bank, May 13-14, 2013
Unpacking urban poverty in a selection of developing countries Why do we know so little about spatial distribution of urban poverty? Small area estimation of poverty Urban poverty across city size in 9 countries 4 Robustness checks: – Cost of living differences – Non-monetary indicators of poverty – Subjective welfare – Artifact of SAE methodology? Why do we observed a poverty-city size gradient? – Limited access to infrastructure services in small towns. Is Africa Different? Preliminary Evidence
Heterogeneity of Urban Poverty: Why do we know so little? Main source of information on distributional outcomes - household surveys - permit only limited disaggregation. Very large data sources (e.g. census) typically collect very limited information on welfare outcomes. Our solution: Small Area Estimation (ELL, 2002; 2003) – Impute a measure of welfare from household survey into census, using statistical prediction methods.
International Experience Poverty Maps are completed or underway in around 50+ countries – Including Brazil, China, India, Indonesia In some countries poverty maps available for multiple periods in time – Ecuador, Panama, Paraguay, Morocco – RIMISP project has produced panel poverty maps across LAC PovMap2 Software development by WB – Available at:
Small Area Estimates of Urban Poverty: 8 countries
Box-Plot of Poverty Rate by City Size: Mexico and Thailand
Box-Plot of Poverty and City Size: Vietnam
Urban Poverty: Summary CountryReference YearIncidence of Poverty (%) Share of small town in urban population (%) Share of small town poor in urban poor (%) Smallest town size category Largest city size category Brazil Mexico India2004/ Sri Lanka Albania Kazakhstan Thailand Vietnam Morocco
Robustness Are our findings of a poverty-city size gradient robust? – cost of living variation – non-income dimensions of wellbeing – Subjective welfare – Statistical artifact?
Spatial Price Variation Household surveys rarely permit construction of price indices that distinguish between city size Evidence from Brazil indicates that adjusting for price variation does attenuate the poverty city size gradient But overall gradient remains – Ferre et al (2012)
Non-Income Dimensions It is sometimes suggested that gradient disappears for non-income dimensions such as child-health Evidence from Mexico indicates presence of a similar gradient based on SAE estimates of child malnutrition (anthropometrics) – Lanjouw and Rascon (2011)
Subjective welfare Evidence from Vietnam indicates that after controlling for access to services, average welfare, as well as households’ real expenditure levels, households in large cities have lower subjective wellbeing than households in small town. – Lanjouw and Marra (2013) Consistent with concept of relative welfare, and also with presence of externalities such as pollution and congestion in large towns.
Does SAE drive finding? Evidence in India, based on NSS sample survey data, is consistent with SAE results. – World Bank (2010) Difficult to test elsewhere due to lack of appropriate survey data.
Why a poverty gradient? Internal dynamism of large cities – Agglomeration externalities “new” economic geography – Scarcity of competent city planners and administrators Neglect of small towns? – Skewed distribution of infrastructure availability Ferre et al (2012) document lower per-capita service availability in small towns
Service delivery and town size in India
Urban Poverty in Africa CountryReference YearIncidence of Poverty (%) Share of small town in urban population (%) Share of small town poor in urban poor (%) Smallest town category Largest city size category Kenya Mali Swaziland Togo Malawi Senegal Gabon Niger Guinea Cote d’Ivoire Central African Republic Mauritania Sierra Leone
Should policymakers focus on small towns? How best to intervene? – Target delivery of improved services? – But possibly higher per capita cost of providing access no guarantee that improved service delivery will result in economic growth of towns. One additional argument: – Small town development may stimulate rural non- farm development