How accurately are household surveys measuring the LGBT population in Colombia? Evidence from a list experiment

Abstract

This paper studies whether household surveys precisely identify the LGBT population and are suitable to measure labor market discrimination in Colombia. We first quantify the size of the LGBT population and estimate labor market inequalities from these data, highlighting potential pitfalls from using this approach. We then present findings from a list experiment in the capital city of Bogotá. Results show that household surveys underestimate the size of the LGBT population and may yield biased estimates of labor market inequalities. While survey estimates range between 1-4%, we find that LGBT individuals represent around 12-22% of the total population. We find heterogeneous reporting by sex at birth, age groups, educational attainment, and marital status. Our findings suggest that while current measurement practices are a step forward for LGBTQ+ visibility, further steps are required before household surveys may be used to consistently estimate discrimination and guide policy responses to protect their welfare.