This paper investigates the determinants of total household income in selected rice-based farming
villages in the Philippines. A quantile regression approach was applied on cross-section data
obtained from 656 farming households across four provinces. Determinants of household income were
examined using an ordinary quantile regression approach, which, unlike conditional mean regression,
allows parameter variation across income quantiles. The quantile regression approach also enables
the analysis of income determinants for extreme categories such as low-income households. Results
indicate that coefficients estimated through ordinary least squares (OLS) could be misleading. The
quantile estimates preserved their signs in most cases but their magnitude varied across quantiles.
The paper particularly emphasizes the determinants of income for poor households. The quantile
estimations show that education of the male head and the existence of migrant workers in households
are the most important determinants of income for poor households.