Alternative wealth indices and health estimates in India

Sanjay K. Mohanty


This paper examines the differentials in estimates of reproductive and child health services by two alternative wealth indices in India. Data from the National Family and Health Survey 3 (NFHS 3, 2005-06) have been used in the analyses. Two wealth indices, namely, a single wealth index for the country without considering the rural and urban specification of economic proxies (Index 1) and a wealth index derived from rural and urban wealth indices (Index 2), are compared. Results indicate that the agreement between two alternate wealth indices is weak. Only 40 percent of the households remained in the same wealth quintile in both the indices and the kappa statistic was 0.19. A comparison of the differentials in estimates of reproductive and child health services (antenatal care, medical assistance at delivery, contraceptive use and child immunization) by wealth indices found that these estimates are higher using Index 1 compared to Index 2 for rural areas. Similarly, the estimates of underweight of children and unmet need for contraception are lower in Index 1 compared to Index 2. The reverse is true for urban areas. It suggests that constructing and using a  wealth index specific to rural and urban areas can help in understanding the true economic differentials of health estimates in population based surveys.


Sanjay K. MOHANTY, Ph.D, Reader, Department of Fertility Studies, International Institute for Population Sciences, Govandi Station Road ,Deonar, Mumbai-400088. E-mail:





wealth index; internal coherence; robustness; rural and urban; reproductive and child health services

Full Text



ISSN: 2035-5556

Reg. Tribunale di Roma n. 3321/54