September 1 4 2013 Chulalongkorn University Bangkok Thailand

September 1 – 4, 2013 Chulalongkorn University, Bangkok, Thailand Targeting poverty in the Lao PDR: how well do current approaches reach the poor? Michael Epprecht Centre for Development and Environment CDE University of Bern, Switzerland

Background of the SAE in the Lao PDR ► Results of the Lao poverty mapping analysis ► Implications for targeting OUTLINE ►

Mapping Poverty in the Lao PDR Small-area estimation analysis using Lao Expenditure and Consumption Survey 2003 (LECS III) Ø 8092 households (20% urban) in 540 villages (5%) Ø Population and Housing Census 2005 Systematic sample of 75% of households: Ø 712, 900 households Ø 4, 123, 988 individuals BACKGROUND Ø

SAE poverty mapping model applied in Lao PDR Household size squared Age composition of household Female-headed household. Education of head Education of spouse Occupation of head Ethnicity. Type of floor Type of wall Type of roof Area of house. Type of water source Type of toilet Type of cooking energy Village cooking energy. Village floor type Village ethnicity Agro-ecological region Urban X X . . X X X X BACKGROUND • • • • • Rural X X X

► Ø Reliability of results Ø Where are the poor? Ø Who are the poor? RESULTS Lao poverty mapping analysis

► Poverty estimates and confidence intervals RESULTS - RELIABILITY

Size of Lao villages (75% sample) 0 500 Frequency RESULTS - RELIABILITY 1000 ► 0 200 400 600 800 # of sample households per village

► Spatial patterns of poverty in the Lao PDR Rural poverty is low in • Xayaburi Province: high value agric. for export • Mekong corridor: irrigated rice and trade with Thailand • Boloven plateau: fertile basaltic soils coffee, tea, & cardamom RESULTS WHERE ARE THE POOR? Poverty rates are highest in the South East along the mountainous border with Vietnam

► Spatial patterns of poverty in the Lao PDR Poverty incidence and poverty density

► Ø Where are the poor? Ø Who are the poor? RESULTS WHO ARE THE POOR? Lao poverty mapping analysis

Poverty rates among male- and female-headed households:

► Poverty among ethnic groups RESULTS WHO ARE THE POOR?

With the poor a bit of everywhere: IMPLICATIONS FOR TARGETING

With the poor a bit of everywhere: What are the implications… ? …for targeting of poverty alleviation efforts? IMPLICATIONS FOR TARGETING ►

Where to reach the poor? ► IMPLICATIONS FOR TARGETING

► MDG 1: Reducing the number of poor people by 50 % in 2015 1/3 of the poor live in poor priority districts IMPLICATIONS FOR TARGETING 50% of the poor live in districts classified as not poor

► MDG 1: Reducing Laos’ number of poor by 50% IMPLICATIONS FOR TARGETING Highest poverty density Poorest areas

► Most people are poor in remote, and sparsely populated areas high ‘access’ costs person little potential ‘leakage’ to non-poor ► Most poor people live in comparatively densely populated areas low ‘physical access’ costs person, but potentially higher targeting costs cheaper service provision high ‘leakage’ to non-poor IMPLICATIONS FOR TARGETING costly service provision

► ► Considerations on type of assistance in targeting (household vs. community/ geographic) Data needs, and updatability of key data Regional development approach in sparsely populated poor remote areas Household or population segment targeting in densely populated areas IMPLICATIONS FOR TARGETING ► Balancing targeting (and implementation) costs vs. ‘leakage’