Health insurance selection in Chile: a cross-sectional and panel analysis

Abstract

In Chile, workers are mandated to choose either public or private health insurance coverage. Although private insurance premiums depend on health risk, public insurance premiums are solely linked to income. This structure implies that individuals with higher health risks may tend to avoid private insurance, leaving the public insurance system responsible for their care. This article attempts to explore the determinants of health insurance selection (private vs public) by individuals in Chile and to test empirically whether adverse selection indeed exists. We use panel data from Chile's 'Encuesta de Proteccion Social' survey, which allows us to control for a rich set of individual observed and unobserved characteristics using both a cross-sectional analysis and fixed-effect methods. Results suggest that age, sex, job type, income quintile and self-reported health are the most important factors in explaining the type of insurance selected by individuals. Asymmetry in insurance mobility caused by restrictions on pre-existing conditions may explain why specific illnesses have an unambiguous relationship with insurance selection. Empirical evidence tends to indicate that some sorting by health risk and income levels takes place in Chile. In addition, by covering a less healthy population with higher utilization of general health consultations, the public insurance system may be incurring disproportionate expenses. Results suggest that if decreasing segmentation and unequal access to health services are important policy objectives, special emphasis should be placed on asymmetries in the premium structure and inter-system mobility within the health care system. Preliminary analysis of the impact of the 'Garantias Explicitas de Salud' plan (explicit guarantees on health care plan) on insurance selection is also considered.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 17 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.