Self-rated health and endogenous selection into primary care.

This study assesses the causal effects of primary care utilization on subjective health status in Turkey using individual-level data from the 2012 Health Research Survey. Employing recursive bivariate ordered models that take into account the possibility that selection into healthcare might be correlated with the respondent's self-reported health status, we find that selection into primary care is endogenously determined and that the utilization of primary care significantly improves self-rated health after controlling for sociodemographics, socioeconomic status, health behaviors and risk factors, and access to healthcare. We show that the causal association between healthcare utilization and health status is robust to the use of objective measures of health and specific types of care, suggesting that the use of a single-item question on self-rated health and binary measures of preventive care utilization is valid.