Where are the Poor

1
Where are the Poor? M A P P I N
G P O V E R T Y in
Kenya and Uganda
Presenter Patti Kristjanson
(ILRI) Presentation to USAID, SAKSS Workshop,
Entebbe, August 2003
2
Research and Policy Advisory teams ILRI Patti
Kristjanson, Andrew Odero, Russ Kruska, Philip
Thornton CBS Godfrey Ndenge, Collins Opiyo,
George Kamula, David Nalo, John Kirimi PEC
Peter Ondieki UBOS Tom Emwanu Makerere Paul
Okwi Institute of Economics Dr. Ddumba
Ssentamu Ministry of Finance Margaret
Kakande WB Johan Mistiaen, Hans Hoogeveen,
Peter Lanjouw, Berk Ozler, Makhtar Diop, Fred
Kilby Financial technical support Rockefeller
Foundation, WB, WRI, DFID, SLP Consortium for
Spatial Information (CGIAR), Food Insecurity and
Vulnerability Information and Mapping Systems
(FIVIMS, FAO), US-CG Linkage Funds (USAID)
3
East Africa Regional Poverty Map
Grey poverty line Blue 40-60 Red 60
4
How to obtain a higher resolution poverty map?

• Welfare Monitoring Surveys can reliably estimate
  household expenditures, but have relatively few
  observations (e.g. Kenya 1,048 rural clusters,
  210 urban 12 households per cluster surveyed)
• Household expenditures (food non-food), from
  these household surveys are used to calculate
  more specific poverty and inequality measures
  linked to a poverty line
• e.g. In Kenya,
• Rural poverty line KShs 1,239/mo/adult
• Urban poverty line KShs 2,648/mo/adult

Different consumption baskets and prices
5
The Approach

• Population and housing censuses do not collect
  information on household expenditures, but, they
• provide complete coverage of a country
• can be aggregated to small statistical or
  administrative areas (communities)
• So, we combined the detailed information obtained
  in the WMS household surveys with the more
  extensive coverage of the censuses to develop
  detailed geographic poverty estimates based on a
  consumption welfare indicator

6
Kenya Percent of Rural Poor PopulationLocation
Level

Source CBS, ILRI, WB forthcoming
7
Kenya Density of Rural Poor PopulationLocation
Level

Source CBS, ILRI, WB forthcoming
8
Kenya Gap for Rural Poor Population to reach
Poverty LineLocation Level


Source CBS, ILRI, WB forthcoming
9
Percent of Rural Population below Poverty Line
Central Province
Poorest Divisions 40-50 poverty
Kiambu District (least poor in Kenya old
District 25) Division 17- 31 Location
11- 41
Poorest Locations 50 poverty
Source CBS, ILRI, WB forthcoming
10
Nyanza Province Percent of Rural Population
below Poverty Line
Source CBS, ILRI, WB forthcoming
11
Nairobi Percent of Urban Population below
Poverty Line
Sub-Location level
Muthaiga
Korogocho
Ngundu
ILRI Uthiru
City Centre
Viwanda
Kibera
Karen
Nairobi West
Poverty rates and inequality are as high in urban
areas as they are in many rural
areas Sub-location-level poverty rates range from
6 to 78 within Nairobi
Source CBS, ILRI, WB forthcoming
12
Uganda poverty maps 1991 and 2000
First, a 1991 map by linking 1992 integrated
household survey with 1991 census District and
County-level poverty estimates Then, by using
the panel element in the 1999 national household
survey with that from 1991, District-level area
welfare estimates for 1999 were also calculated
13
District-Level Rural Poverty Incidence in Uganda
1991 and 1999
1991
1999
Percent of population below rural poverty line
Source UBOS, ILRI, WB, forthcoming
14
County-Level Rural Poverty Incidence in Uganda
1991
Percent of population below rural poverty line
Source UBOS, ILRI, WB, forthcoming
15
County-Level Rural Poverty Incidence in Uganda
1999
Percent of population below rural poverty line
Source UBOS, ILRI, WB, forthcoming
16
Changes in Poverty 1991 to 1999

• Some good news!
• Poverty reduced in most regions, but stagnation
  in northeast
• Drop in poverty is highest in the central region
• Quite a few Counties, e.g. in southwest Uganda
  eastern Uganda, went from over 50 poverty to
  less than 20

17
Uses of this information

• Geographic distribution of well-being varies
  considerably across Kenya and Uganda within
  Districts
• This info alone does not tell us WHY
• First step in an important process this needs to
  feed into PRSP processes and get used by analysts
  decision makers at ALL levels!
• e.g. Min of Health or Education combines this
  info with location of health facilities could
  estimate no. of poor people with good/poor access
  to different types of health or education services

18
Example Poverty and distance to government
health facilities in Kajiado District
Source ILRI
19
Technology Recommendation Domains e.g.
Probability of adoption of Napier in Kiambu
District, Kenya A function of PPE, pop density,
average distance to nearest 2 urban centres,
total distance on all roads to Nairobi (Staal et
al., 2003)
Human population
2000 2050 High prob
258,000 438,000 Kiambu 567,000
962,000
Source ILRI
20
Beyond Overlays Spatial econometric analysis of
factors influencing poverty, and
poverty-environment linkages
Analysis to understand the interactions between
poverty, technological change, market
development, and investment in natural resource
management Exploring specific livestock-poverty
reduction linkages with geo-referenced hh surveys
focusing on livestock assets and management
strategies
21
Map of spatial prediction of probability of dairy
cattle adoption, based on parameter estimates of
GIS-derived variables Kenya Highlands
Source Staal et al, 2002
22
Appropriate investment strategies Example
Pattern of milk surplus or deficit in Uganda -
Parish level (kgs of milk/year/square kilometre)
Source Staal and Kaguongo, 2003
23
Targeting Interventions Example Comparison of
access to veterinary services and level of cattle
density
Source Staal and Kaguongo, 2003. Estimates
based on national community surveys, 1999/2000,
and government cattle population statistics
24
Proportion of communities with reported access to
extension services, by district
Source Staal and Kaguongo, 2003. Estimates from
national community surveys, 1999/2000
25
ILRI/CG Lessons relevant SAKSS

• You dont need to start from scratch
• Data methods Tremendous opportunities for
  leveraging 10 years of experience in the CG -
  investments in spatial and other databases, tools
  methods, etc
• Human capacity, e.g. GIS technicians, database
  management build on existing capacity and
  partnerships already developed
• Infrastructure matters!
• Build on existing spatial analyses experience
  ILRI, IFPRI partners on the forefront

26
ILRI-KENYA_at_cgiar.org ILRI is a Future Harvest
Centre