Isolation and poverty: the relationship between spatially - - PowerPoint PPT Presentation

Isolation and poverty: the relationship between spatially differentiated access to goods and services and poverty

Kate Bird, Andy McKay & Isaac Shinyekwa

Understanding and Addressing Spatial Poverty Traps. Spier Estate, Stellenbosch, 29 March 2007

Overview

• What are spatial poverty traps?
• How do they drive and maintain poverty

and chronic poverty?

• Why develop an index of isolation?
• The components of our index
• Results: the index applied to Uganda

What are spatial poverty traps?

• Spatial poverty traps are where ‘geographic

capital’ is low & poverty is high

– Geographic capital – natural, physical, political, social and human capital of an area

• Conceptual framework (typology)

– Remote rural areas (frictional distance, locational disadvantage) – Low potential or marginal areas (ecologically disadvantaged) – Less favoured areas (politically disadvantaged) – Weakly integrated areas (poorly linked and economically disadvantaged)

Why are spatial poverty traps important?

• Rural poverty is three times higher (incidence)

than urban poverty in SSA, E & SE Asia and Latin America (IFAD, 2001)

• Approx 1.8 billion people, most of them poor, live

in less-favoured or low potential areas

• Multi-dimensional poverty & destitution are

strongly concentrated - in spatial poverty traps

• Poverty persists in spatial poverty traps even

where a country has experienced economic growth and aggregate reductions in the poverty headcount

What drives spatial poverty traps: market & state failures

• Market failure

– Under-investment – Economic activities which extract resources but fail to deliver pro-poor growth

• State failure, inadequate provision of

– Institutional, political & governance failures – Inadequate provision of infrastructure – Poor security – Limited attention to developing an enabling environment – Poor basic services (esp. education & health) – Limited social protection

What drives spatial poverty traps: agro-ecology, stigma and exclusion

• Agro-ecology

– High risk (covariate overlaying idiosyncratic) – Drudgery intense livelihoods – Few opportunities for diversification into higher return activities – Migration - seasonal, circular and permanent

• Stigma & exclusion

– Marginalisation/ minority groups – ethnicity, language, religion, culture, livelihood group (e.g. pastoralists), habits – Discrimination – Blamed for their own poverty – Poorly connected to elites, poorly represented in national discourse

What drives spatial poverty traps: physical isolation & inadequate infrastructure

• Physical isolation & inadequate infrastructure

– Why is provision poor?

• Less favoured area – socially & politically excluded (weak lobby, stigmatised

group)

• Can be technically difficult to deliver (remote, rugged terrain)
• Can be expensive per head (where population densities are low – they aren’t

always) – Implications?

• Increases ‘frictional distance’
• Isolation, weak access to markets, goods and services
• Bangladesh, significant and substantial impact on living standards;
• Peru – spatially differentiated household expenditure and income;
• Tanzania – people within 100m of a gravel road earn 1/3 more than the rural

average;

• Ethiopia – more than 8km from markets, people more likely to withdraw from

markets

• Nepal – frictional distance (journey time from village to market) significantly

reduced subjective well-being

– Key infrastructure depends on binding constraint (roads, electricity)

What drives spatial poverty traps: communication, media & ICTs

• Communication, media & ICTs

– National Radio, national TV stations, national newspapers, landline telephone, cell-phone connectivity, internet access – Why is provision poor?

• Role of the state as a provider/ enabler/ regulator
• Private sector (limited effective demand?)

– Implications?

• Isolation from mainstream society, new ideas and technical transfer
• Impact for shared values, shared narratives, national unity vs

marginalisation

• Impact on enterprise and markets (information asymmetries,

transaction costs, market fragmentation)

What drives spatial poverty traps: crime & insecurity

• Greater problem in remote and isolated areas in many developing

countries (e.g. Madagascar – Fafchamps & Moser, 2004)

• Why?

– Social fragmentation & limited livelihood options mean that it is a real

• ption for youth

– Few leisure options/ high alcohol consumption – contributes to drunken brawls/ rape – Ineffective policing (police more likely to harass & brutalise local population than solve crime) – Banditry/ armed insurgents/ terrorist groups

• Impact?

– Constrained movement (esp. women & girls) – Negative impact on well-being and on livelihoods – Low levels of trust – Risk averse behaviour/ Low levels of investment

Poverty in Uganda

• Poverty incidence - 56% (1992), 44% (1997/98), 34%

(1999/2000), 38% (2002/03)

• Post-conflict bounce-back & coffee boom in early 1990s
• Then, broadly spread economic growth, UPE, road-building,

decentralisation

• Now, increasing inequality, growth only benefiting top 20%,

poor may be getting poorer

• North and West much poorer than Central and Eastern areas

(Centre well connected, East – long-run benefits from coffee boom?)

• 2 wave panel - Integrated household survey (1992) & Uganda

National Household Survey (1999/2000)

– nearly 40% of the 1398 panel households experienced transitory poverty (29.6% moved out of poverty, 10.3% moved into poverty) – 18.9% were chronically poor

The components of our index

• Extensive fieldwork in Uganda – collecting different dimensions of

isolation at the District level

– Ministry of Works (density of feeder roads) – UNDP (data on health and safe water) – Uganda Electricity Transmission Company Ltd (availability of electricity at the District level) – Ministry of Information (data on radio and television stations) – Monitor & New Vision (circulation figures)

• Uganda National Household Survey data

– Average distances to primary and secondary schools – Average distances to the municipality HQ – Average distances to Kampala

• Many new Districts. We used the 47 (1998-99) Districts which

formed the strata for the Uganda National Household Survey

• BUT some indicators could not be constructed for all Districts – data

not present for some districts due to conflict

Constructing the index (1)

• Want to look at the relationship between isolation and

poverty

• Many dimensions of isolation important in their own right

– relate to different dimensions of the problem

• BUT highly desirable to focus on a limited number of

measures of isolation (otherwise will have a ‘woods for trees problem’ and won’t see patterns)

• We used factor analysis – widely used in the analysis of

Demographic and Health Survey data to define asset quintiles (combining a diverse range of assets which a household may or may not own)

Constructing the index (2)

• Constructed 2 indices

– Average distance in the District to key amenities and locations (e.g. roads, main town of the District) – Availability of key facilities and amenities within the District (schools, health centres etc)

• Focus mainly on the first index
• Values of these indicators used to classify

the 47 Districts into quartiles

Results: the index applied to Uganda (1)

• North & West = more isolated
• Centre & East = least isolated
• We knew this already. How does this add value?
• Significant heterogeneity within regions

– Mubende & Nakasonsolo (Central – overall, least isolated) top quartile of isolation – Bushenyi (Western) lowest quartile of isolation

• Broad similarities for both indices – but some interesting

differences

– One Northern District in the top quartile (best provided) for facilities – One Central District in the bottom quartile (least well-provided) for facilities

Results: the index applied to Uganda (2) Isolation & Poverty Dynamics

• Applied the index to households in the Uganda 2

wave panel to see whether isolated households are more likely to be chronically poor

• We found a strong association between isolation

& incidence of consumption poverty

– poverty incidence increases with isolation quartile – poverty incidence in the most isolated quartile is more than twice that of the least remote quartile

• Similar findings for depth of poverty

Results: the index applied to Uganda (3) Isolation & Poverty Dynamics

• People in isolated rural areas more likely to be

chronically poor

– In the most isolated quartile, twice as many households are chronically poor – In the least isolated quartile, hhs are much less likely to have been poor in either of the two periods – Likelihood of escaping poverty decreases systematically with isolation – Likelihood of school age children not being in school increases with isolation (much higher in most isolated quartile) – Use of protected drinking water falls with isolation – Access to electricity falls with isolation – Likelihood of being involved in non-agricultural activity falls with isolation (strong and systematic pattern) – important implication for exit from poverty