Estimating subjective poverty lines with discrete information

This paper proposes a novel methodology for the estimation of subjective poverty lines (SPLs) using a discrete information approach that obviates the potential discomfort of asking respondents directly about the value of their individual SPL. To estimate income SPLs, we utilize the Youden index. Using a simulated data-set, we first show that the level of bias between the estimated SPLs and their corresponding actual values is low. Next, we demonstrate an application of the proposed methodology to the EU-SILC 2018 microdata, utilizing the “ability to make ends meet” indicator as the classification variable. We show that in Western EU countries the values of SPLs with discrete information are typically greater than estimates based on a standard minimum income question. The opposite is true for Eastern EU countries.