The effect of general health checks on healthcare utilization: accounting for self-selection bias

Sungwook Yoon, Duk Bin Jun, Sungho Park

Research output: Contribution to journalArticlepeer-review

9 Scopus citations


The general health check is one of the most common preventive healthcare measures in many countries. In this study, we propose an empirical approach which jointly models the decision to obtain a general health check and healthcare utilization, tackling the self-selection problem by using eligibility to obtain a health check for free as an instrumental variable. Eligibility has some exogenous variations by design and this helps us to partial out the effect of general health checks from self-selection biases. We apply the model to a large 12-year panel data set provided by the Korean National Health Insurance Service. We find that participation in the general health check increases healthcare utilization and ignored self-selection generates substantial upward bias in the estimates. We also find that the health check effect shows noteworthy heterogeneity across gender and income groups. Before health checks, healthcare utilization of males and people in low income groups is lower than those of females and people in high income groups respectively. However, these become comparable across different groups after health checks. This finding implies that general health checks can be an effective vehicle for health equity.

Original languageEnglish (US)
Pages (from-to)3-36
Number of pages34
JournalJournal of the Royal Statistical Society. Series A: Statistics in Society
Issue number1
StatePublished - Jan 1 2020


  • Counterfactual analysis
  • General health check
  • Healthcare service
  • Instrumental variables
  • Self-selection bias

ASJC Scopus subject areas

  • Statistics and Probability
  • Social Sciences (miscellaneous)
  • Economics and Econometrics
  • Statistics, Probability and Uncertainty


Dive into the research topics of 'The effect of general health checks on healthcare utilization: accounting for self-selection bias'. Together they form a unique fingerprint.

Cite this