1 / 24

Small Area Estimation for Monitoring the MDGs at the Subnational Level

Small Area Estimation for Monitoring the MDGs at the Subnational Level. by Candido J. Astrologo, Jr. Jessamyn O. Encarnacion Director, National Statistical Information Center Chief, Social Sectors B Division

faye
Download Presentation

Small Area Estimation for Monitoring the MDGs at the Subnational Level

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Small Area Estimation for Monitoring the MDGs at the Subnational Level by Candido J. Astrologo, Jr. Jessamyn O. Encarnacion Director, National Statistical Information Center Chief, Social Sectors B Division National Statistical Coordination Board National Statistical Coordination Board Workshop on MDG Monitoring 14-16 January 2009, Bangkok, Thailand

  2. Outline of Presentation • Introduction • Small Area Estimation (SAE) Methodology • Other Applications of SAE • Concluding remarks and recommendations

  3. I. Introduction • Philippine official poverty statistics are released every 3 years at the regional and provincial levels of disaggregation. • All official regional poverty estimates (for 2000, 2003 and 2006) are reliable (having coefficients of variation (CVs) of at most 10%). • In the case of the official provincial poverty estimates, 28 out of 84* or 33% of the provinces are reliable with CVs less than 10%, while 46% have acceptable CVs between 10 and 20 and 21% have CVs over 20%. • No official municipal or city level estimates are generated

  4. I. Introduction • In 2005, the NSCB with funding assistance from World Bank ASEM Trust Fund, conducted a poverty mapping project using small area estimation methodology as part of the Philippine Statistical System’s continuing effort to respond to the growing need for lower level disaggregation of information on the poor. 2000 Family Income and Expenditure Survey 2000 City and Municipal Level Poverty Statistics based on SAE 2000 Labor Force Survey 2000 Census of Population and Housing • 2000 poverty estimates for all the municipalities in the country were released in November 2005 by the NSCB.

  5. II.Small Area Estimation Methodology

  6. II. Small Area Estimation Methodology Aim • Produce provincial-, municipal- and city-level estimates of poverty incidence, gap and severity based on official income-based provincial poverty lines by merging information from census and surveys

  7. II. Small Area Estimation Methodology Data Requirements • Survey containing target variable (Y), independent variables (X) • Census containing X (but not Y)

  8. II. Small Area Estimation Methodology Two types of data sources: • Household surveys • - include a detailed income and/or expenditure module • - however, due to relatively small sample size, collected information is usually only representative for broad areas of the country, e.g., regions Data sources for the Philippines: 2000 Family Income and Expenditure Survey (FIES) and Labor Force Survey (LFS)

  9. II. Small Area Estimation Methodology Two types of data sources (cont’d): 2. Census data - available for all households and can provide reliable estimates at highly disaggregated levels such as cities and municipalities - however, census data do not contain income/expenditure information necessary to estimate poverty Data source for the Philippines: 2000 Census of Population and Housing (CPH)

  10. II. Small Area Estimation Methodology Main idea • Merge information from the two types of data sources to come up with small area poverty estimates • “Borrow strength” from the much more detailed coverage of the census data to supplement the direct measurements of the survey

  11. II. Small Area Estimation Methodology Basic procedure • Use the household survey data to estimate a model of per capita income (Y) as a function of variables that are common to both the household survey and the census (X’s). • Use the resulting estimated equation/model to predict per capita income for each household in the census. • The estimated household-level per capita income are then aggregated for small areas, such as cities and municipalities.

  12. II. Small Area Estimation Methodology Candidate variables (X’s) a. Common variables from FIES/LFS and CPH (18) - Household dwelling characteristics (7) - Family characteristics (11) b. Municipal-level census means (25)

  13. II. Small Area Estimation Methodology Modeling • Regression Regression models were constructed that estimated the income of households based on household level and community-level characteristics.

  14. II. Small Area Estimation Methodology Production of small area estimates 2000 poverty estimates for each city/municipality, province (urban and rural): - poverty incidence - poverty gap - severity of poverty

  15. II. Small Area Estimation Methodology HOW to update these city and municipal level estimates? 2003 SAE 2000 SAE 2003 Family Income and Expenditure Survey 2000 Family Income and Expenditure Survey Time-invariant (i.e., variables that may be considered “stable” over time) 2003 Labor Force Survey 2000 Labor Force Survey 2000 Census of Population and Housing 2000 Census of Population and Housing

  16. II. Small Area Estimation Methodology Features of the 2000 and 2003 SAE methodologies used

  17. IV. Other Applications of SAE Other indicators where SAE technique was applied in the Philippines • Proportion of households not meeting energy adequacy at the provincial level • Provincial prevalence of underweight among 6-10 year old children • District (or barangay) level estimation of the proportion of underweight Filipino children aged 0-5 years • Proportion of stunted 0-5 year-old children at the provincial level • Provincial prevalence of hypertension among adults • Labor and employment statistics at the provincial level

  18. IV. Other Applications of SAE Relevance/Actual policy use of 2000 SAPE • The 2000 small area poverty estimates released by NSCB in 2005 were already used by DSWD in their Pantawid Pamilyang Pilipino Program. • The Department of Social Welfare and Development (DSWD) used the municipal poverty incidences in identifying priority municipalities for KALAHI-CIDSS (e.g., Samar) • NNC and DSWD used the data in December 2007 to identify priority households for the Pamaskong Handog of GMA. • The SAE were used by the Department of Agriculture (DA) as one criterion in the identification of target sites of the Cordillera Highland Agricultural Resources Management Project (CHARMP II).

  19. IV. Other Applications of SAE Relevance/Actual policy use of 2000 SAPE • Regional KALAHI Convergence Group (RKCG) used the estimates to serve as one of the bases in identifying its convergence municipalities throughout the region (e.g., MIMAROPA). • National Nutrition Council (NNC) Region VIII used the SAE in assessing the nutritional situation of municipalities in the region in October 2007. • Results were used as input to determine target enrolment for health insurance sponsored programs of PhilHealth in 2007. • Leyte: SAE results were used to determine priority municipalities in Leyte in May 2007 for: (i) sponsorship program for schooling of indigent children; and (ii) for micro-enterprise development (MED) projects. • The Department of Energy also expressed interest in the SAE results as a possible reference for the installation of bio-diesel.

  20. IV. Other Applications of SAE Relevance/Actual policy use of 2003 SAPE • The 2003 intercensal small area poverty estimates was also used by the DSWD as basis for prioritizing target households for their proposed National Household Targeting System for Poverty Reduction (NHTSPR) • The Department of Energy also expressed interest in the SAE results as a possible reference for the installation of bio-diesel.

  21. IV. Concluding remarks and recommendations • Small area estimation techniques can be used successfully to produce poverty estimates at the provincial and municipal levels. • The estimates at provincial level were in general consistent with, but more precise than the direct estimates obtained from the survey data alone (official methodology), with an average SE (CV) of less than 2% (5%) • The precision of the municipal level estimates was more or less similar to that of the official provincial level estimates

  22. IV. Concluding remarks and recommendations • Possible generation of SAE for other indicators critical in decision- and policy-making such as: • a) Unemployment – not available for city/municipal levels • b) Infant and maternal health • c) Post-census populations (alternative pop’n. projections) • d) Non-income component indicators of the HDI (i.e., life expectancy, functional literacy, and basic education participation rate)

  23. IV. Concluding remarks and recommendations • Generation of poverty statistics for basic sectors – not available for: 1) city/municipal levels and 2) some sectors, where direct estimation of poverty statistics is not possible due to data constraints. • Generation of poverty maps at lower levels of disaggregation - poverty estimates overlaid and/or combined with information on education, health, access to infrastructure, environment, crime, among others.

  24. Maraming Salamat po! URL: http://www.nscb.gov.ph e-mail: info@nscb.gov.ph

More Related