Crowd out of private insurance and take up of public insurance: evidence from nearly universal health care coverage for the elderly

JUAN MARCELO VIRDIS; MARIA DOLORES JIMENEZ-RUBIO; ALESSIO GAGGERO

Sevilla

Workshop; XIII Taller EvaluAES de Investigación en Evaluación de Políticas y Servicios de Salud; 2024

Facultad de Enfermería, Fisioterapia y Podología de la Universidad de Sevilla (Sevilla)

Many countries worldwide are actively engaged in developing public health insurance as a way of improving individual health and wellbeing. However, the decision to use public or private healthcare is largely affected by the phenomenon labelled as “crowd-out”, which implies that individuals will be prompted to give up private healthcare as the quality of health services in the public sector improves. Building on the fact that the Argentinian state funded Integral Program of Health Care (PAMI, for its acronym in Spanish) is provided to eligible pensioners, we exploit the discontinuity in the legal retirement age to study its effects on coverage. More specifically, we quantify the extent of the crowd-out effect of mandatory public insurance on voluntary private insurance and explore the heterogeneity of the effects at various socioeconomic level thresholds by using pseudo panel data from the Continuous Household Panel. The empirical modelling is based on a Regression Kink Design approach which models the variation in the level and slope of mandatory health insurance coverage and voluntary insurance coverage at legal retirement age. The findings reveal a significant increase in overall insurance coverage (public and private) leading to a narrowing of coverage disparities across different socio economic subgroups after legal retirement age is met. Additionally, we find suggestive evidence for a slight crowding-out effect on the basis of educational attainment. This study sheds light on the effect of expanding universal health insurance in middle-income economies, showcasing the benefits it can bring to less affluent groups due to the heterogeneous coverage differences across socioeconomic groups.