URBAN POVERTY AND MOBILITY IN INDONESIA

URBAN POVERTY AND MOBILITY IN INDONESIA

Indonesia has undergone rapid urbanisation in the last half century, and this is expected to continue over the next two decades Millions 197 75 8 85 9 95 2 5 1 15 2f 25f 3f 35f Percentage Point Change Developing Countries El Salvador Jordan Canada (188-191) Belarus Thailand Angola Cameroon Bolivia Netherlands Lao PDR Iran, Islamic Rep. Nigeria Algeria Turkey Dominican Republic Developing EAP Indonesia Korea, Rep. Malaysia China Haiti Saudi Arabia (196-9) South Korea (196-9) 4 3 2 1 The urbanisation rate has grown from 17 to 54 percent since 197, and is expected to reach 71 percent by 235, while the rural population shrinks Indonesian Population, 197-235 8 6 4 2 Urbanisation (%) The speed of urbanisation is one of the fastest in the world 5 45 4 35 3 25 2 15 1 5 Change in Urbanisation, 1975-215* Urban Rural Urbanisation Rate 197-215 215-35f Urban CAGR* 4.4% 2.% Rural CAGR*.5% -.8% Urban % Change 37ppt 17ppt Source: WDI for 197-215, BPS for 22-35 Notes: Source: Countries with populations over 5m; selected countries with different periods WDI, World Bank (214) Urban China

Urbanisation has led to a prosperous middle class, but poverty is becoming urbanised and inequality is higher and rising faster Percent Percent There are three times more economically secure in urban areas than rural ones but poverty is becoming increasingly urban, and urban inequality is higher and rising faster Economically Secure Emerging Vulnerable Poor Urban and Rural Populations by Economic Class, 213 Urban Population (m) 54.2 4.6 27.5 1.3 Rural Population (m) Note: Poor live beneath the national poverty line, the vulnerable have greater than 1 percent chance of being poor next year (under 1.5 times the line), the emerging consumer class a greater than 1 percent chance of being vulnerable next year (under 3.5 times the line), the economically secure have a less than 1 percent chance of being poor or vulnerable next year (>3.5 times the line), while the upper class consume more than $2 PPP per person per day (see World Bank (forthcoming) Indonesia s Middle Class for more details). Source: Susenas and World Bank calculations 13. 17.7 34.1 6.5 8 4 4 8 6 5 4 3 2 1 Urban Share of Poor 215 235f Note: Forecast urban share of poor assumes same poverty reduction in both urban and rural areas from 215 to 235, but with urbanisation rate increasing from 5 percent in 21 to 67 percent in 235. Source: BPS and World Bank calculations Gini, 24-14 5 45 4 35 3 25 2 15 1 5 24 214 Source: BPS and World Bank calculations

Indonesian Urban Flagship Report Main Report Motivation & Trends Drivers of Urbanisation, Prosperity, Inclusiveness & Livability Background Papers Recent Trends in Indonesian Urbanisation The Many Dimensions of Urban Poverty Urban Economic Mobility Roles of Agglomeration Economics, Human Capital Externalities & Market Access Quantifying the Costs of Congestion in Indonesian Cities Impact of Congestion & Pollution on Productivity Rural-Urban Migration Urban-Rural Growth Spillovers Policies for Leveraging Urbanisation What Constrains Labor Mobility in Indonesia? Land Constraints, Slums, & Urban Development in Indonesia Connecting People to Jobs Where is the Gap? Assessing the Mismatch Between Supply & Demand for Skills in Indonesian Cities Key Issues in Urban Governance & Finance

RECENT TRENDS IN INDONESIAN URBANISATION

Metropolitan population growth has come in the suburban ring rather than the urban core, but economic growth is concentrated in the core Population (m) Constant 2 IDR Trillions Most metropolitan population growth has been in the suburban ring rather than the urban core but despite this, most of the economic growth has come in the core 12 1 Multi-District Metropolitan Population Growth, Core versus Suburbs, 1996-215 8 6 Change Pop. CAGR* 22m 1.7% 1,6 1,4 1,2 1, 8 6 GDP in Core and Peripheral Districts, 1993-215 4 2 1996 21 24 27 211 215 Core Periphery Note: CAGR is compound annual growth rate Source: Susenas 3m.6% 4 2 1993 27 213 Core Periphery Source: Susenas, INDO-DAPOER, World Bank calculations

This suburban growth undermines the formation of agglomeration economies, reducing the economic benefits of urbanisation Urban Share Denser urban cores are more conducive to agglomeration economies than less dense suburban rings so growing suburban populations mean the urban share of the economy has grown more slowly than its share of the population Agglomeration economies of cities: higher urban incomes and productivity levels point to the existence of benefits from city size. But why do these benefits exist? There are three main theories: reductions in transport costs for goods, people and ideas The evidence best supports the theory that cities speed the flow of ideas: faster flow of ideas makes firms more productive; spread of ideas may increase rate of technological change; transmission of ideas increases human capital acquisition for workers and facilitates innovation Evidence is consistent with this, suggesting cities thrive because of their ability to spread knowledge Thus the growth of significantly less dense suburban areas in Indonesian cities is unlikely to result in the same economic benefits that come in denser urban cores 7% 6% 5% 4% 3% Urban Share of Population and GDP, 1993-213 2. 1.5 2%.5 1% %. 1993 1996 1999 22 24 27 211 213 Urban Share of GDP (LHS) Urban Share of Population (LHS) Ratio (RHS) Source: Susenas, INDO-DAPOER, World Bank calculations 1. Ratio

while less economic opportunity and lower population density means higher suburban poverty and less efficient and effective public services Percent Percent There are less economically secure and more poor in the suburban periphery while access to services is worse 1 9 8 7 6 5 4 3 2 1 Consumption Distribution, 214 Core Periphery Smaller Cities Substandard Housing of Urban Areas, 22-14 7 6 5 4 3 2 1 22 214 22 214 Core Periphery Extreme Poor Poor Vulnerable Econmically Secure Sanitation Water Overcrowding Source: Susenas and World Bank calculations Source: Susenas and World Bank calculations

Background Paper 1: Recent Trends in Urbanization Research Questions Location of Urban Growth: How has the urban population grown across the country? In larger and smaller cities, on and off Java? In the urban core and the suburban periphery? Has it been in high-risk areas (e.g., prone to flooding)? Drivers of Urban Growth: How much of urban population growth is natural growth, rural urban migration and rural reclassification? Nature of Urban Growth: What are the patterns across and within cities of urban population growth, density, economic growth and poverty? Nature of Urban Employment: What sectors and occupations have seen the most job growth? The most wage growth? What is the balance between formality and informality across different urban sectors?

THE MANY DIMENSIONS OF URBAN POVERTY

Monetary and non-monetary poverty are less correlated in urban areas, and monetary well-being does not mean non-monetary well-being Percent Non-monetary poverty tends to be co-located in poor rural children, but there is much less correlation in urban areas while 4 percent of the economically secure are deprived on at least one dimension of housing, water and sanitation Poor Health Access 4% 6% 5% Rural Children 9% 2% 5% 6% Poor Education Access 41% Poor Health Access 8% Urban Children 3% 2% 3% 1% 2% 2% Poor Education Access 7% 1 9 8 7 6 5 4 3 Economically and non-monetarily secure Economically secure, nonmonetary poor Emerging 22% 2 Vulnerable 18% Poor Transportation 5% Poor Transportation 26% 1 Urban Poor

Urban housing, while better than in rural areas, is generally substandard. However, different urban areas experience different problems Percent Note: XXXX living condition definitions Source: Susenas and World Bank calculations Most urbanites have at least one substandard living condition. Outer regions suffer most from poor water and sanitation, while overcrowding and poor roofing are bigger issues in the urban core 1 9 8 7 6 5 4 3 2 1 Substandard Housing of Urban Areas, 22-14 22 214 22 214 22 214 22 214 Core Periphery Single District Cities Rural Overall Sanitation Water Overcrowding Roof Wall Floor In all types of urban areas, at least 6 percent of residents lack at least one of good quality housing materials, clean water or proper sanitation Poor quality roofing is increasing in all areas, but is a greater problem in urban cores or single district cities than in the suburban ring Poor quality housing materials are the main driver of substandard living conditions in single district cities Access to water and sanitation have been improving over time but remain a significant challenge in the suburban sprawl On the other hand, overcrowding is low in the periphery and declining, but higher and rising in the urban core where land is more scarce

Background Paper 2: The Many Dimensions of Urban Poverty Research Questions Who are the urban poor? E.g., education, employment sector and occupation, migrant status, gender, age, etc. What is the role of employment in urban poverty? Poverty by self- versus wage-employment, formality versus informality, industry/sector Within these categories, what does earnings dispersion look like? Is the urban self-employment sector twotiered? Does this intra-sector heterogeneity explain workers poverty status? How much do sources of income explain changes in urban poverty? Does under-employment play a role in explaining poverty, and is this disguising unemployment? What hours do the urban poor work? What are the patterns of non-economic poverty? How does access to health and education, water and sanitation, etc. differ by urban location (larger vs. smaller cities, core vs. periphery) and across different sub-groups (e.g., consumption deciles, gender, age and ethnicity, etc.)? How do commute times and travel modes vary by population size, density and area? How does this vary by sub-group? How does air quality vary by population size, density and area? How frequent are different crimes in different urban areas? What are the causes of urban crime? Special focus 1: Marginalised and vulnerable groups Who are they, what is the cause of their marginalisation, what policies are needed to improve their welfare? Special focus 2: Slums Who lives in slums, what access do they have to employment and services, what mobility do they experience?

URBAN ECONOMIC MOBILITY

Economic mobility is considerably lower in urban Indonesia compared to rural areas, as well as to many other developing countries Bolivia Belarus El Salavador Ecuador Moldova Paraguay Peru Indonesia (Rural) Colombia Kyrgyzstan Panama Russia Argentina Kazakhstan Brazil Mexico Armenia Honduras China (Rural) Indonesia (Urban) Costa Rica Thailand Turkey Georgia Ukraine Romania Chile Poland Philippines China (Urban) A mobility parameter has been developed which shows that mobility in urban Indonesia is lower than in rural Indonesia and than in many other developing countries y it is the log income of individual i at time t, generated by the autoregressive process: y it = ρy it1 + λ i + ε it λ i is an individual effect β t-1 captures: income persistence (ρ); and degree to which dispersion of λ i affects the dispersion of y it-1 1 - β t-1 is the mobility parameter.6.5.4.3.2.1. Mobility by Country, Time Periods Vary Note: Mobility parameter is 1 β t-1, where β t-1 summarises the co-movement of an individual s income over time, containing both the auto-regressive persistence of incomes and the degree to which the dispersion of individual incomes in any time period is due to a fixed individual effect. It is taken as the midpoint of the upper and lower bound estimates. Source: Calculated from Kraay and Van der Weide (217)

Many poor urban households move up the distribution over time, but others fall back: what are the channels and barriers to mobility? 27 Per Capita Consumption Quintile Urban Rural Average consumption growth for the poor and vulnerable is higher than typically measured but many fail to rise up the distribution and others fall back 8% 7% 6% Annualised Per Capita Consumption Growth of Bottom 4, 23-13 Urban and Rural Household Transition Matrix, 27-14 214 Per Capita Consumption Quintile (percent of 27 quintile) Q1 Q2 Q3 Q4 Q5 Q1 44 25 18 9 4 5% Q2 25 27 23 17 8 4% Q3 13 19 26 24 19 3% Q4 7 13 23 3 27 2% 1% % Indonesia (Rural) Non-anonymous Anonymous Indonesia (Urban) Q5 4 8 1 24 53 Q1 46 27 14 9 4 Q2 35 24 19 17 5 Q3 2 25 23 22 9 Q4 13 2 25 26 17 Q5 8 15 17 25 35 Note: Non-anonymous mean growth of the bottom 4 percent shown as mid-point of the upper and lower bound Source: Kraay and Van der Weide (217) Note: Real per capita household and individual consumption quintiles from households matched over both waves Source: IFLS and World Bank calculations

Background Paper 3: Urban Economic Mobility Research Questions What is the degree of economic mobility? Intra- and inter-generational What are the determinants of economic mobility (and immobility)? E.g., education, employment sector and occupation, migrant status, gender, age, etc. How do the sources of income change over time for the mobile and non-mobile? Economic and geographical mobility within cities Is economic mobility related to geographical mobility? Do people move out of slums into other parts of the city when they can afford to? That is, are slums poverty traps or stepping stones into formality? Are physical moves welfare enhancing even if they remain informal? Role of labor market segmentation With segmented markets, people may be poor due to a lack of access to formal sectors, wage work or certain industries, rather than a lack of pre-market skills: What are the raw earnings differences according to formality, wage- versus self-employment, industry How much can be explained by observable factors? Can time-invariant individual heterogeneity explain differentials between portions of the labor market?

RURAL-URBAN MIGRATION

A quarter of Indonesians are migrants, mostly long-term. The majority are from rural areas and migrated to urban or majority urban areas Percentage of Population Percent Around 24 percent of working age Indonesians are internal migrants, mostly long-term migrants 59 percent of migrants are from a mostly rural district. 62 percent go to a mostly or completely urban district 1 8 Indonesian Population by Migrant Category, 215 24% Recent Migrant Long-term Migrant 1 8 6 Working Age Migrants by Origin and Destination, 215 6 4 76% Non-migrant 4 2 2 Origin Destination Note: Long-term migrants live in a different district to their birth and migrated over five years ago; recent migrants live in a different district to the one they lived in five years ago. Source: Susenas and World Bank calculations Urban <3% Rural 3-5% Rural >5% Rural Note: Districts are treated as urban (kotamadya), or classified based on their percentage rural (for kabupaten). R3 is <3% rural, R2 is 3-5% rural, R1 is >5% rural. Source: Susenas and World Bank calculations

Internal migration has been an important channel for some households to exit poverty or for upwards mobility 27 Economic Status 57 percent of recent poor rural-urban migrants escaped poverty, and 41 percent of the vulnerable became non-poor Rural-Urban Migrant Transition Matrix, 27-14 214 Economic Status (percent of 27 status) Poor Vulnerable Non-poor Poor 43 37 2 Vulnerable 26 33 41 Non-poor 7 21 72 Migration can bring long-term economic benefits for households Using the 28 Rural-Urban Migration in Indonesia (RUMiI) survey, Resosudarmo et al. (29) find that the household income of migrants is slightly higher than non-migrants in Medan, Tangerang, Samarinda, and Makassar but what are the socio-economic consequences for: Originating communities Receiving communities Migrants and their families Note: Classes are those below the national poverty line (poor), between the poverty line and the vulnerability line (vulnerable, under 1.5 times the poverty line) and above the vulnerability line (non-poor) Source: IFLS and World Bank calculations

Background Paper 4: Rural Urban Migration Research Questions Macro-level trends in rural urban migration Historical and contemporary trends in rural urban migration Trends in seasonal and temporary migration versus permanent migration Comparison of migration rates in Indonesia to other countries in the region Extent to which population growth in urban areas is due to migration (versus administrative reclassification and natural urban growth) Who migrates and who stays? Socio-demographic characteristics of migrants Differences in seasonal / temporary migrants compared to long-term / permanent migrants Geographic distribution of migrants (across cities, within cities and slums) and over time (do they enter slums and then graduate out?) Drivers of and barriers to rural urban migration Effects of migration On originating communities: remittances, poverty, burden for the elderly, human capital, including skills transmission and networks On receiving communities: attitudes towards migrants On migrants themselves: economic mobility including their relative wages, their children's health and education outcomes, access to social protection How well do migrants do relative to other workers with similar characteristics? Among migrants, what are the pathways out of poverty? Which migrants are able to take advantage of these and which are not? How do the economic effects differ for first-, secondand later-generation migrants?