Some Statistical Dimensions in the Generation of Philippine Poverty Statistics

This paper presents some statistical issues and possible solutions in the estimation of Philippine Poverty Statistics based on the official methodology adopted. In particular, the paper shows the effect of survey weights used in the estimation of population counts, number of households, number of poor persons among others. The empirical evidence used combined data sets from the Labor Force Survey and the Family Income and Expenditure Survey shows the extent of possible underestimation of magnitudes using the official methodology and thereby necessitating a possible revision. Ultimately, the paper recommended an alternative approach in generating household weights that would yield estimates which are consistent with population projections.