Self-Reported Morbidity and Self-Rated Health among the Elderly in India: Revisiting the Puzzles

Both self-reported morbidity (SRM) and self-rated health (SRH) have been questioned on the grounds of reliability and validity, particularly in the context of developing countries such as India. It has been argued that indicators of self-perceived health often do not move in tandem with objective indicators of health and are not reflective of socioeconomic inequalities. This paper attempts to systematically examine the validity of SRH and SRM, analysing data for the aged population from the last two health rounds of the Indian National Sample Survey. We consider two indicators of SRM: chronic and acute, and test their association with different positional parameters such as economic condition, educational attainment and availability of healthcare facilities, using multivariable logistic regressions. We also test the association of SRH with these parameters and SRM, using ordered probit regression. In an alternative specification, we replace the general indicators of acute and chronic morbidity with the presence of specific diseases and health conditions, and examine the validity of SRH. Though SRM is typically lower for people who are disadvantaged in terms of different socioeconomic parameters, the latter are more likely to report worse SRH. Self-reported presence of specific illnesses and health conditions also significantly increases the likelihood of giving worse ratings for SRH among the aged in India. Our findings suggest that while SRH is a valid health indicator among the aged in India, one cannot be entirely dismissive of SRM, particularly with regard to its predictive validity for SRH.