Estimating Official Poverty Statistics in Non-FIES Years
Author: Manuel Leonard F. Albis, Roxanne Jean L. Elumbre, Dannela Jann B. Galias & Jessa S. Lopez
With poverty eradication being the focus of countries around the world, the need for a comprehensive and timely poverty statistics has never been more pressing. Although official poverty statistics are generated by national statistical systems, existing limitations prevent the generation of yearly official poverty statistics. In this regard, this paper presents an alternative method of estimating poverty statistics even on years when household survey data is not collected. Macroeconomic indicators, along with demographic and employment information from the Labor Force Survey were considered in the model and were found to be plausible determinants of poverty incidence. Ultimately, a complete annual time series of poverty statistics was produced from an estimated Dynamic Factor Model (DFM) using the state-space form. Relatively high forecast accuracy was observed for the predicted values of poverty incidence.
Keywords:Dynamic Factor Model (DFM), factor analysis, poverty statistics, poverty incidence, state-space form, forecasting, poverty estimation