Aggregation Bias and Socioeconomic Gradients in Waiting Time for Hospital Admissions

Abstract

Waiting time is a rationing mechanism that is used in publicly funded healthcare systems as a mean to ensure equal access for equal need. However, several studies suggest that individuals with higher socioeconomic status wait less. These studies typically measure patients' socioeconomic status as an aggregate measure from patients' residential area and the results are hence vulnerable for aggregation biases. We shed light on the magnitude of the aggregation bias by analyzing socioeconomic gradients in waiting times when education and income are measured on three different levels: the individual level, the population cell level, and the municipal level. Our individual level socioeconomic gradient is modest compared with the literature. When socioeconomic status is measured on an aggregate level, we observe stronger associations with socioeconomic variables and less accurate estimates. A researcher who only has access to the aggregate data runs the risk of overstating the magnitude of the socioeconomic gradients.