This paper focused on estimating the Foster-Greer-Thorbecke (FGT) poverty
measures (poverty incidence, poverty gap, severity of poverty) at the provincial level
in the Philippines. Design-based and model-based estimation techniques were used
and the estimates obtained were compared in terms of some measures of accuracy,
precision and reliability. The sources of data include 2000 Family Income and
Expenditure Survey (FIES) and 2000 Census of Population and Housing (CPH). For
estimates of provincial poverty incidence, the model-based estimation technique
produced reliable estimates in 71 provinces. The linear model used has an adjusted
R
2
of 63.55% with 4 predictors; namely, provincial proportion of the total barangays
having (1) street pattern; (2) access to the national highway; (3) surroundi ngs as
waste disposal system; and (4) provincial average number of male persons per
household. For provincial poverty gap, more than 75% of the provinces have reliable
estimates based on model-based estimation technique. The predicting model has an
adjusted R2
of 62.67% and the identified predictors include provincial average
number of members aged between 7 and 14; the provincial proportion of the number
of housing units with floor area at most 10 sq. m.; and the provincial number of
housing units of single type. In estimating provincial poverty severity index, the
predicting model has an adjusted R2 of 51.08% and the predictors include the
provincial proportion out of the total count of (1) housing units with floor area at most
10 sq. m.; (2) households with members aged less than 7; and (3) barangays with the
surroundings as waste disposal system. Thus, the model-based estimation technique
produced reliable estimates of the provincial FGT poverty measures in the
Philippines.