Locations and Neighborhoods in Urbanizing Africa Simon Franklin - - PowerPoint PPT Presentation

Locations and Neighborhoods in Urbanizing Africa

Simon Franklin

London School of Economics

UNU-WIDER Conference

Helsinki September 2015

Outline for the talk

Three points of motivation:

1

Rural-urban migration.

2

Role of neighborhoods for economic outcomes.

3

Welfare and housing in cities. Questions for urban research Two examples of randomized experiments from Addis Ababa

1

Lower transport costs.

2

Subsidized housing and relocation.

Agenda Migration

(1) Migration will drive urbanization

Birth rates in African cities are falling. Migration flows to cities are enormous. Addis Ababa (1997-2007): Growth due to birth rate: < 3% (versus 40% in rural areas). Gained 25% of its population in migrants. Many migrants are young people,

• ften moving separately from their families, and looking for jobs or

to pursue education.

Agenda Neighborhood Effects

(2) The role of neighborhoods

Chetty et al (2015): Differences in income-mobility across different areas of the US. Moving to Opportunity Experiment confirms this.

◮ And evidence that moving (geographically) makes a big

difference.

(Conclusion is not that we should knock down bad neighborhoods!) Large literature on the importance of neighborhoods for: access to services, social interactions, schooling, access to jobs. Know very little about this in the context of developing countries...

Agenda Neighborhood Effects

Spatial inequality in Africa?

We do know that people who live, or move to, cities... ... have higher earnings (Young, 2013) and higher living standards (Gollin et al, 2015). ... and that they face barriers to those movements because moving to a city is risky (Bryan et al, 2015).

Agenda Neighborhood Effects

Spatial inequality in Africa?

We do know that people who live, or move to, cities... ... have higher earnings (Young, 2013) and higher living standards (Gollin et al, 2015). ... and that they face barriers to those movements because moving to a city is risky (Bryan et al, 2015). We don’t know what happens within cities: Where new migrants are living and how that influences economic outcomes. What makes urban life risky, and how those risks can be mitigated. About the role of networks, neighborhoods, public services in urban areas.

Agenda Neighborhood Effects

Spatial inequality in Africa?

Growing body of evidence on slums (Marx, Suri, Stoker; 2013) ...upgrading and tenure security. New slums, where migrants live are, increasing going to be located far away from city centres. Long term residents of slums face rent pressures as cities grow. But mobility is important, and requires functioning rent markets.

Agenda Urban transfers

(3) Welfare and social services in Urban Africa

Governments and donors increasingly spending on transfers in urban areas. Will these be appropriate?

Agenda Urban transfers

(3) Welfare and social services in Urban Africa

Governments and donors increasingly spending on transfers in urban areas. Will these be appropriate? Social safety nets.

◮ Mitigate risk/vulnerability for migrants? Impact on labour supply?

Housing.

◮ Built in the right places? Foster good neighborhoods? ◮ Allow for mobility? How are new migrants housed?

Agenda Urban transfers

Urban research agenda

What are the impacts of housing policies and welfare ... on urban labour markets, and neighborhoods? ... migration rates, and who migrates? Alot of evidence from rural areas, not cities.

Agenda Urban transfers

Urban research agenda

What are the impacts of housing policies and welfare ... on urban labour markets, and neighborhoods? ... migration rates, and who migrates? Alot of evidence from rural areas, not cities. Because research is hard to do in urban areas, especially at the neighborhood level

1

fewer natural experiments.

2

geographic disaggregation of data.

3

RCTs harder to perform because of spillovers & GE effects.

Experiments in Addis Ababa

Research in urban Ethiopia

Does it matter where people live?

1

The impacts of reduced transport costs on job search in Addis Ababa.

2

The impacts of highly subsidized government condominium housing, located far from the city, on many economic outcomes.

Experiments in Addis Ababa

Addis Ababa: Urban Density

Experiments in Addis Ababa

Population 4 million. Projection- 8 million by 2025

Experiments in Addis Ababa

Firms are clustered in the centre

Experiments in Addis Ababa

Jobs are in the center...

Experiments in Addis Ababa

...Especially skilled and professional jobs

Experiments in Addis Ababa

Information about jobs is also located in the center

Experiments in Addis Ababa

Getting to the jobs is expensive

Transport and job search

Are individuals living on the periphery prevented from accessing jobs because of where they live?

I run a randomized control trial distributing non-fungible transport subsidies for 12 weeks, conducted in Addis Ababa.

Transport and job search

Are individuals living on the periphery prevented from accessing jobs because of where they live?

I run a randomized control trial distributing non-fungible transport subsidies for 12 weeks, conducted in Addis Ababa. I show that lowering search costs caused job seekers to: Search more (using a high-frequency panel). Find better jobs.

Transport and job search

Are individuals living on the periphery prevented from accessing jobs because of where they live?

I run a randomized control trial distributing non-fungible transport subsidies for 12 weeks, conducted in Addis Ababa. I show that lowering search costs caused job seekers to: Search more (using a high-frequency panel). Find better jobs. The subsidies only bring those on the periphery to the centre. Evidence that distance from jobs generates frictions and inequality in labour markets.

Transport and job search

Vulnerability and Volatility

Study makes use of weekly phone call data to track outcomes.

◮ Young people’s outcomes are extremely volatile: work, search,

expenditure.

◮ Movements between formal/informal, temporary work and casual

labour.

◮ Construction & domestic work.

Transport subsidies reduce the rate at which job seekers become discouraged. Transport subsidies reduce participation in temporary work.

Transport and job search

Percentage of Individuals Searching over time

.5 .6 .7 .8 .9 1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Week

Treatment Control

Transport and job search

Targetting and Take-up

Study population made up of two groups:

1

Found and interviewed at the central jobs in central Addis Ababa.

2

Found and interviewed at home in neighborhoods of the city. Though I find impact in both groups, subsidies more useful for the highly motivated individuals.

Transport and job search

Targetting and Take-up

Study population made up of two groups:

1

Found and interviewed at the central jobs in central Addis Ababa.

2

Found and interviewed at home in neighborhoods of the city. Though I find impact in both groups, subsidies more useful for the highly motivated individuals. Take-up: The subsidies were non-fungible, so only taken up by individuals who had use for them, or were motivated to search for work.

Transport and job search

Implications of Transport Subsidy Study

Evidence that reducing transport costs, all else equal, leads to better employment outcomes. Urban planning: Make central housing available and affordable. Reduce transport costs. Cash constraints and large distances from jobs could be locking some individuals out of labour markets. Role for well targeted welfare programs for migrants and other vulnerable youth in transition.

Government housing and neighborhoods

Government housing and neighborhoods Housing Challenge

Housing Challenges in Africa & Ethiopia

Upgrading or Relocation?

Government housing and neighborhoods Housing Challenge

Housing Challenges in Africa & Ethiopia

Upgrading or Relocation? Integrated Housing Development Program 4 or 5 story condominium houses. Ownership transferred: subsidized mortgages. 50, 000+ units per year. Lottery for opportunity to move into the housing.

Government housing and neighborhoods Housing Challenge

Where will the new housing be built?

Government housing and neighborhoods Housing Challenge

Where will the new housing be built?

Government housing and neighborhoods Housing Challenge

Moving to or away from Opportunity?

Government housing and neighborhoods Study Design

Baseline survey of 2000 households in June 2015

Sample of 1000 households who won March lottery... Will be moving into new housing within three months, in 200 corridors in 9 housing sites, 4 different neighborhoods.

Government housing and neighborhoods Study Design

Baseline survey of 2000 households in June 2015

Sample of 1000 households who won March lottery... Will be moving into new housing within three months, in 200 corridors in 9 housing sites, 4 different neighborhoods. ... and 1000 households still waiting for housing, who were eligible for March lottery Matched 1-1 to winners on observables. Stand to win in the next lottery, in new sites, in the next year. Ex ante 50% chance of winning in March lottery.

Government housing and neighborhoods Study Design

Mortgages are large, so are asset values.

Land is “free”, construction costs ($180/m2) covered by mortgages. Down-payments for 1/2 Bedroom flats are $8, 00 − $4, 000. Seems well beyond the means of the urban poor. Value of land transfer ($6000+).

Government housing and neighborhoods Study Design

Mortgages are large, so are asset values.

Land is “free”, construction costs ($180/m2) covered by mortgages. Down-payments for 1/2 Bedroom flats are $8, 00 − $4, 000. Seems well beyond the means of the urban poor. Value of land transfer ($6000+). So will the poor benefit? Most applicants have made the down-payments. Will they move in to the housing or will they rent them out?

Government housing and neighborhoods Study Design

Research Questions

Impacts of physical housing Measure the impacts on health, savings, community, risk-sharing. Long run impacts on children.

Government housing and neighborhoods Study Design

Research Questions

Impacts of physical housing Measure the impacts on health, savings, community, risk-sharing. Long run impacts on children. Impact of location Impact of where households move on outcomes. Impact on how far households move (home and work). ... on labour market outcomes. ... on access to services.

Government housing and neighborhoods Study Design

Research Questions

Impacts of physical housing Measure the impacts on health, savings, community, risk-sharing. Long run impacts on children. Impact of location Impact of where households move on outcomes. Impact on how far households move (home and work). ... on labour market outcomes. ... on access to services.

Government housing and neighborhoods Study Design

What is the impact of neighbors?

Random composition of neighborhoods in blocks and corridors. Impact of neighbor composition (wealth, employment, language diversity). How do new community networks develop after disruption. How to attitudes to strangers, prejudice, public goods change?

Government housing and neighborhoods Study Design

Thank you!