Measuring Philippine poverty through cluster analysis
Date of Award
Master of Arts in Psychology, Concentration in Developmental Psychology (Non-Thesis Program)
Beja, Edsel L., Jr., Ph.D.
Statistics on poverty incidence are often employed in the development of national-level policies such as conditional cash transfer programs for the poorest of the poor. In certain instances, these statistics can strongly influence the manner in which government resources are allocated and deployed. The seemingly excessive focus on poverty incidence, however, ignores its deficiencies - particularly its unidimensional structure or its overemphasis on income. These deficiencies are made more evident through an analysis of the Capability Approach which elaborates on the multidimensional nature of poverty and, more importantly, the value to policymaking of viewing poverty as a multidimensional issue. The nuances implicit in the multidimensionality of poverty should therefore not be ignored. Instead, these must influence the manner in which poverty is viewed, measured, and addressed. As such, the development of alternative means of determining macro-level poverty measurements would prove critical in the proper formulation, calibration, and deployment of policies and interventions against poverty. This study is an attempt to reframe poverty in the Philippines within the Capability Approach. The study develops an alternative means of estimating the proportion of poor Filipinos by subjecting information from the Annual Poverty Indicators Survey to a combination of Principal Component Analysis, Multiple Correspondence Analysis, and Cluster Analysis. This study builds on existing literature by incorporating Multiple Correspondence Analysis into the Cluster Analysis Framework for poverty measurement. The results suggest that poverty incidence measures may severely underestimate poverty and obscure significant and/or growing intraregional and interregional inequality. The results also suggest that the regional minimum wages can serve as heuristic cutoffs for poverty in the Philippine setting.
(2018). Measuring Philippine poverty through cluster analysis. Ateneo de Manila University.