65th ISI World Statistics Congress 2025

65th ISI World Statistics Congress 2025

Identifying and Profiling the Socioeconomic Classes of Philippine Households using the Family Income and Expenditures Survey

Conference

65th ISI World Statistics Congress 2025

Format: CPS Abstract - WSC 2025

Keywords: classification, socioeconomic

Session: CPS 45 - Socioeconomic Factors and Statistical Modelling of Household Expenditure and Consumption

Monday 6 October 5:10 p.m. - 6:10 p.m. (Europe/Amsterdam)

Abstract

Socioeconomic Classification (SEC) is a broad term that often refers to categories describing an individual or household’s income, purchasing power, or relative position in society regarding wealth, and may be defined in several ways. The Philippines has no official SEC, but the Unified Socioeconomic Classification or 1SEC is a system of household classification that was introduced and launched in 2010 as a product of collaboration between professional agencies (Marketing and Opinion Research Society of the Philippines or MORES), the academe (University of the Philippines School of Statistics or UPSS), and the then National Statistics Office (NSO), now Philippine Statistics Authority (PSA). Bersales, et al (2013) were able to obtain 9 socioeconomic clusters (referred to here as 1SEC 2009) based on the 2009 Family Income and Expenditure Survey (FIES) and identified determinants of the socioeconomic clusters using ordered logistic regression models.
With technical and funding assistance to PSA’s Statistical Methodology Unit (PSA SMU) from the World Bank in 2016, the SEC using the 2015 FIES was reviewed with UPSS and MORES collaborating with PSA. With a modified procedure, a 9-tier classification (referred to here as 1SEC 2015) was constructed using a hierarchical clustering algorithm, with the least-spending household assigned to Cluster 1, and the highest-spending household assigned to Cluster 9 (See Figure 1). Several models (ordered logistic regression models and classification trees) were considered in identifying important predictors of 1SEC. The middle clusters were still difficult to classify, and there are observed differences in the sets of predictors identified based on 2009 FIES and 2015 FIES. In particular, expenses incurred in internet use, buying new clothes, and trips abroad were new predictors.
With the 2021 FIES now available, an update will be done by the authors to identify patterns of expenditures related to SEC of Philippine Households.

Figures/Tables

Figure 1