Mother's education, learning-by-doing, and child health care in rural India

Comparative Education Review

Volumn 49, Issue 4, 2005, Pages 534-551

  Lee, Sang Hyop ^a,b   Mason, Andrew ^a,b,c  

^a UNIVERSITY OF HAWAII   (United States)

^b East West Center   (United States)

^c Department of Economics ^*  (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 27744596823     PISSN: 00104086     EISSN: None     Source Type: Journal    
DOI: 10.1086/454372     Document Type: Review

References (31)

1. Previous studies include Constantijn W. A. Panis and Lee A. Lillard, "Health Inputs and Child Mortality: Malaysia," Journal of Health Economics 13, no. 4 (1994): 455-89;
2. Mark Rosenzweig and T. Paul Schultz, "Who Receives Medical Care? Income, Implicit Prices, and the Distribution of Medical Services among Pregnant Women in the United States," Journal of Human Resources 26, no. 3 (1991): 473-508;
3. Mark Rosenzweig and Kenneth I. Wolpin, "Heterogeneity, Intrafamily Distribution, and Child Health," Journal of Human Resources 23, no. 4 (1988): 437-61;
4. Michael Grossman and Theodore J. Joyce, "Unobservables, Pregnancy Resolutions, and Birth Weight Production in New York City," Journal of Political Economy 98, no. 5 (1990): 983-1007;
5. Sang-Hyop Lee, "Demand for Immunization, Parental Selection, and Child Survival," Review of Economics of the Household 3, no. 2 (2005): 171-96.
6. The six diseases are tuberculosis, diphtheria, whooping cough (pertussis), tetanus, polio, and measles. As part of the National Health Policy, the Expanded Programme on Immunization (EPI) was introduced in 1978 with the objective of providing free vaccination services to all eligible children and expectant mothers. In order to increase the pace of immunization, the Universal Immunization Programme (UIP) was introduced in 1985-86 and is being implemented through the existing network of the primary health care system, which includes primary health centers (PHCs), subcenters, and referral centers called community health centers. Refer to World Health Organization (WHO), The Expanded Programme on Immunization in South-East Asia, SEARO Regional Health Papers, vol. 12 (New Delhi: World Health Organization, 1986), for details.
7. Gary S. Becker, "A Theory of the Allocation of Time," Economic Journal 75 (September 1965): 493-517.
8. See, e.g., Dona Strobino, Virginia Keane, Elizabeth Holt, Nancy Hughart, and Bernard Guyer, "Parental Attitudes Do Not Explain Underimmunization, " Pediatrics 98, no. 6 (1996): 1076-83.
9. According to the WHO guidelines, the recommended immunization schedule is as follows: BCG (against tuberculosis) at birth; three doses of DPT (diphtheria, pertussis, tetanus) vaccine and three doses of oral polio vaccine at 6, 10, and 14 weeks; and measles immunization at 9 months. The NFHS shows that almost every child who is vaccinated met these criteria.
10. The usual truncation problem arises whenever the health care is defined to depend on the duration of life or it is dependent on the achievement of a given age. For example, immunizations given after some age is reached would be truncated by death prior to the immunization age and thus be spuriously related to life expectancy. If these children are more likely to come from poor households with mothers of lower education levels, ignoring these children may lead to a systematic overestimate of immunization coverage. On the other hand, if families are more likely to vaccinate weaker children, engaging in compensatory behavior, ignoring these children may create an underestimate of immunization coverage in the analysis. There have been a number of ways in which these problems have been addressed. See Kenneth I. Wolpin, "Determinants and Consequences of the Mortality and Health Inputs of Infants and Children," in Handbook of Population and Family Economics, ed. Mark R. Rosenzweig and Oded Stark (Amsterdam: North-Holland, 1997), 1A:535-38, for a review of several alternative ways of dealing with this problem. However, it is very difficult to assess the value of the approach in practice. Dealing with this problem is beyond the scope of our analysis.
11. See, e.g., Fred Arnold, Minja K. Choe, and T. K. Roy, "Son Preference, the Family-Building Process, and Child Mortality in India," Population Studies 52, no. 3 (1998): 301-15;
12. Fred Arnold, Sunita Kishor, and T. K. Roy, "Sex-Selective Abortions in India," Population and Development Review 28, no. 4 (2002): 759-85;
13. R. Pande, "Selective Gender Differences in Childhood Nutrition and Immunization in Rural India: The Role of Siblings," Demography 40, no. 3 (2003): 395-418;
14. Robert D. Retherford and T. K. Roy, Factors Affecting Sex-Selective Abortion in India and 17 Major States, National Family Health Survey Subject Reports no. 21 (Mumbai: Institutional Institute for Population Sciences and East-West Center, Honolulu, 2003).
15. Although we assume that mothers do not know the sex of their children before the first prenatal visit, this assumption may not be correct in the Indian setting. Sex determination services may not provide prenatal care because sex determination in India is sometimes done through clandestine means, largely in the unregulated private sector. Thus, it is theoretically possible that the decision to go for prenatal care in formal settings is based on prior knowledge of the sex of the fetus.
16. Rosenzweig and Wolpin, "Heterogeneity, Intrafamily Distribution, and Child Health."
17. Identification of the learning-by-doing effect in this article purely relies on distributional and functional assumptions. Since different assumptions on the distribution can lead to different results, this identification strategy is less satisfactory than relying on exclusion restrictions. Thus our results should be interpreted with caution. Unfortunately, it is very difficult to have exclusion restrictions in this type model, because all the variables affecting prenatal care will also affect child immunization. Guido W. Imbens, "Sensitivity to Exogeneity Assumptions in Program Evaluation," AEA Papers and Proceedings 93, no. 2 (2003): 126-32, discusses the issue of the sensitivity of the estimated effect to the distributional assumptions on the error terms. Following his methodology would allow for a general distribution of the error terms that affect prenatal care and immunization and find bounds on the estimate of the effect of prenatal care on immunization as we vary the distributional assumptions. Although this is an interesting approach, we do not pursue the methodology here.
18. Anne R. Pebley, Noreen Goldman, and German Rodriguez, "Prenatal and Delivery Care and Childhood Immunization in Guatemala: Do Family and Community Matter?" Demography 33, no. 2 (1996): 231-47.
19. International Institute for Population Sciences, National Family Health Survey (Bombay: MCH & Family Planning, 1995).
20. The information based on mother's memory does not include the date of vaccination, and it may include a higher percentage of children who did not meet the criteria recommended by WHO. However, Noreen Goldman and Anne R. Pebley demonstrate that inclusion of maternal recall data improves the accuracy of estimates of immunization coverage, even though it is subject to recall error. See Noreen Goldman and Anne R. Pebley. "Health Cards, Maternal Reports, and the Measurement of Immunization Coverage in Guatemala," Social Science and Medicine 38, no. 8 (1994): 1075-89.
21. Our model predicting per capita household expenditure used 24 state dummy variables and the father's occupation (seven categories), education (eight categories), age, and age squared. Only the father's characteristics are included in the estimation, because the mother's labor force participation may be endogenous in child health investment decisions. Unfortunately, the NSS does not contain reliable information on income. See John Strauss and Duncan Thomas, "Human Resources: Empirical Modeling of Household and Family Decisions," in Handbook of Development Economics, ed. J. R. Behrman and T. N. Srinivasan, vol. 3 (Amsterdam: North-Holland, 1995) for pros and cons of using different measures of income.
22. Also see Angus Deaton, The Analysis of Household Surveys (Baltimore: Johns Hopkins University Press, 1997), chap. 1, for the difficulties of measuring income as well as consumption in developing countries. Because we are predicting per capita expenditure using a different data set, the estimated coefficients of per capita expenditure may not be efficient.
23. Pande, "Selective Gender Differences";
24. Vinod Mishra, T. K. Roy, and Robert D. Retherford, "Sex Differentials in Childhood Feeding, Health Care, and Nutritional Status in India," Population and Development Review 30, no. 2 (2004): 269-95.
25. Scheduled castes and tribes are those castes and tribes identified by the Government of India as socially and economically disadvantaged and in need of protection from exploitation.
26. See, e.g., Sekhar Bonu, Manju Rani, and Timothy D. Baker, "The Impact of the National Polio Immunization Campaign on Levels and Equity in Immunization Coverage: Evidence from Rural North India," Social Science and Medicine 57, no. 10 (2003): 1807-19.
27. Note, however, that the estimates presented here assume that no correlation exists between the community variables and unobserved components in the outcome. Because immunization programs may be placed using criteria that are related to the outcomes being studied (i.e., nonrandom program placement), this condition is often violated. As a crude test, we estimate whether the distance to clinics is negatively related to the immunization rate by regressing the variable measuring the availability of a clinic on the full immunization rate of the village. No statistical evidence of nonrandom program placement was found.
28. Pande, "Selective Gender Differences";
29. Mishra et al., "Sex Differentials in Childhood Feeding."
30. Mishra et al., "Sex Differentials in Childhood Feeding."
31. World Health Organization, "Report of the Technical Review Group Meeting," June 7-8, 1998, Vaccine Research and Development, WHO/VRD/GEN/98.02, World Health Organization, Geneva.