Using Australian panel data to account for unobserved factors in measuring inequities for different channels of healthcare utilization

Inequity in healthcare utilization is typically measured as the unequal distribution of services by observable non-need indicators, such as income, after controlling for observable need indicators. However, important sources of unequal healthcare utilization are often unobserved. The unobserved element may reflect need factors, such as imperfectly measured severity of illness, that would predict greater utilization across different healthcare channels, but also based on choice, such as patient preferences to use a particular healthcare channel over an alternative one, which may differ in its effect between channels. Accounting for unobserved sources of utilization may, therefore, help to understand contradictory inequalities between different healthcare channels, such as pro-poor inequalities for general practitioner use and pro-rich inequalities for specialist visits. This paper uses survey data from the Household Income and Labour Dynamics in Australia and panel data methods to investigate if seemingly contradictory inequalities between different healthcare channels are explained by latent individual-level heterogeneity. Results show that unobserved individual-level heterogeneity affects inequities across different healthcare channels, providing indications that the unobserved element may primarily represent unobserved need.