Abstract
This paper uses hazard regression models to assess the impact of experienced infant and child mortality on the risk of subsequent conceptions in Ethiopia. The purpose of this paper is to test for the presence of a fertility response to an infant or child death, net of the effects of truncated breastfeeding on fecundity. Using retrospective birth history data from a national survey in Ethiopia, we find a significantly higher risk of a conception in the months following the death of an index child, even after controlling for postpartum amenorrhoea and breastfeeding status. The fertility response is strongest after the death of the fourth or fifth child, which is when most women in Ethiopia are at or near their desired family size. However, we find no evidence of a fertility response to the death of a nonindex child. We attribute the higher risk of a conception following an index child’s death to the intentional efforts of couples to reduce the waiting time to a next birth and thereby replace the deceased child. However, absent evidence of replacement fertility in response to the death of older nonindex children, we interpret the response to the death of an index child as an emotional response to child loss rather than a conscious strategy to meet a fertility target.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
The models with heterogeneity estimated in this paper use a nonparametric correction term introduced by Heckman and Singer (1984). The advantage of the nonparametric specification is that frequently one has no prior knowledge about the expected parametric form of the unobserved variables. Studies also have shown that model results can be sensitive to the parametric form of heterogeneity chosen (Heckman & Singer, 1982; Trussell & Richards, 1985, p. 247). The nonparametric approach developed by Heckman and Singer uses a discrete probability distribution with a step function to describe the mixing distribution of the unobserved effects. The distribution works with a finite number k of support points at locations θ j , and is confined to the interval [0,1] by setting θ 1 = 0 and θ k = 1. The k support points correspond to latent groups in the population. Associated with each support point is a probability mass p j that can be interpreted as the proportion of the population found in a particular latent group. θ is a 1 × k vector of location parameters, and c i is a factor loading. Each c i θ j defines a contrast to the constant term that indicates the shift in the baseline hazard associated with a particular latent group. In the model without heterogeneity, one support point is assumed (θ = 0), with all the probability mass located on the constant term. When two points of support are specified (θ 1 = 0 and θ 2 = 1), the parameters p and c i are estimated. When three points of support are specified (θ 1 = 0 and θ 3 = 1), θ 2 , p 1, p 2, and c i are estimated (see Heckman & Walker, 1987; Trussell & Richards, 1985). In general, (k-2) + (k-1) + 1 parameters are estimated for k points of support. To estimate the hazard of conception, we used the computer program CTM (for continuous time models). Although CTM permits time-varying regressors, the correction for unobserved heterogeneity is appropriate only for heterogeneity attributable to constant unobserved effects.
We estimated multistate models that distinguished nonindex children deaths by the deceased child’s birth order, but found no significant effects.
Current and ever use of contraception was asked for married women. Retrospective information on contraceptive use was not collected.
Not all women who reported complete information on the last two births reported complete birth histories, hence more women were included in the analysis of the last two births than in the analysis of the birth histories.
In hazard regression models of the resumption of postpartum menses, Berhanu and Hogan (1998) included a dummy variable set equal to one if the duration ended in a multiple of six. The estimated coefficient for heaping was negative and significant, indicating a lower than expected hazard, controlling for other factors, and thus a longer than expected duration.
Rounding up durations of postpartum amenorrhoea increases the likelihood that conceptions which occur after the resumption of menses are incorrectly assigned to the postpartum amenorrheic period and not assigned to the period following the resumption of menses. The net effect of this misreporting is to underestimate both the negative effect of postpartum amenorrhoea on the hazard of conception (because conceptions that did not occur during this period are treated as if they did) and the positive effect of death of an index child on the hazard of conception, net of postpartum amenorrhoea (because conceptions that occurred after the resumption of menses are treated as is if they did not occur during this period).
From the generalized form of the conditional hazard, the Weibull, Gompertz, and exponential regression models can be specified by changing the form of the duration dependence term.
Among currently married fecund women interviewed in the 1990 NFFS the mean additional number of children wanted by women with zero living children was 4.54, for women with one living child the mean was 3.70, for women with two living children the mean was 2.76, for women with three living children the mean was 2.28, for women with four living children the mean was 1.64, and for women with five living children the mean additional number of children wanted was 1.17 (CSA, 1993, p. 247).
References
Berhanu, B., & Hogan, D. (1998). Postpartum amenorrhoea in Ethiopia: The role of weaning, child death, and socioeconomic factors. Social Biology, 45(1–2), 80–95.
Bongaarts, J., & Bruce, J. (2001). Population growth and policy options in the developing world. In P. Pinstrup-Andersen, & R. Pandya-Lorch (Eds.), The unfinished agenda: Perspectives on overcoming hunger, poverty, and environmental degradation (pp. 6–66). Washington, D.C.: International Food Policy Research Institute.
Bongaarts, J., & Potter, R. (1983). Fertility, biology and behavior. New York: Academic Press.
Brass, W., & Barrett, J. C. (1978). Measurement problems in the analysis of linkages between fertility, child mortality. In S. H. Preston (Ed.) The effects of infant and child mortality on fertility. (pp. 209–233). New York: Academic Press.
Central Statistical Authority (CSA). (1993). The 1990 National Family and Fertility Survey Report. Addis Ababa, Ethiopia: Central Statistical Authority.
Chandran, P. E. (1989). Biological influence of infant death on fertility. Journal of Biosocial Science, 21, 217–221.
Chowdhury, A. K. M. A., Khan, A. R., & Chen, L. C. (1976). The effect of child mortality experience on subsequent fertility: In Pakistan and Bangladesh. Population Studies, 30(2), 249–261.
Flinn, C. J., & Heckman, J. J. (1982). New methods for analyzing individual event histories. In S. Leinhardt (Ed.), Sociological Methodology 1982 (pp. 99–140). San Francisco CA: Jossey-Bass.
Grummer-Strawn, L. M., Stupp, P. W., & Mei, Z. (1998). Effect of a child’s death on birth spacing: A cross-national analysis. In M. R. Montgomery, & B. Cohen (Eds.), From death to birth: Mortality decline and reproductive change. (pp. 39–73). Washington, DC: National Academy Press.
Guz, D., & Hobcraft, J. (1991). Breastfeeding and fertility: A comparative analysis. Population Studies 45(1), 91–108.
Hasen, A., Assefa, H., & Yacob, Z. (1994). Proximate determinants of fertility in Ethiopia: An application of the Bongaarts Model. Addis Ababa University, Ethiopia: Demographic Training and Research Center.
Heckman, J. J., & Singer, B. (1982). Population heterogeneity in demographic models. In K. C. Land, & A. Rogers (Eds), Multidimensional mathematical demography (pp. 567–599). New York: Academic Press.
Heckman, J. J., & Singer, B. (1984). A method for minimizing the impact of distributional assumptions in econometric models for duration data. Econometrica, 52, 271–320.
Heckman, J. J., & Walker, J. R. (1987). Using goodness of fit and other criteria to choose among competing duration models: A case study of Hutterite data. In C. C. Clogg (Ed.), Sociological Methodology 1987 (pp. 247–307). Washington, DC: American Sociological Association.
Heckman, J., & Walker, J. (1991). Economic models of fertility dynamics: A study of Swedish fertility. Research in Population Economics, 7, 3–91.
Jensen, A. (1997). The impact of child mortality on fertility in Kenya, Zimbabwe and Botswana. Re-examining the links. Informal Session 6: A New Look at the Relationships Between Child Mortality and Fertility. Paper presented at the 23rd IUSSP General Population Conference, Beijing, China, 11–17 October 1997.
Kuate Defo, B. (1998). Fertility response to infant and child mortality in Africa with special reference to Cameroon. In M. Montgomery, & B. Cohen (Eds), From death to birth: Mortality decline and reproductive change. (pp. 254–315). Washington, DC: National Academy Press.
Lehrer, E. (1984). The impact of child mortality on spacing by parity: A Cox-Regression analysis. Demography, 21(3), 323–337.
Lindstrom, D. P., & Berhanu, B. (1999). The impact of war, famine, and economic decline on marital fertility in Ethiopia. Demography, 36(2), 247–261.
Lindstrom, D. P., & Berhanu, B. (2000). The effects of breastfeeding and birth spacing on infant and early childhood mortality in Ethiopia. Social Biology, 47(1–2), 1–17. .
Lindstrom, D. P., & Woubalem, Z. (2003). The demographic components of fertility decline in Addis Ababa, Ethiopia: A decomposition analysis. Genus LIX, (3–4), 147–158.
Lloyd, C., & Ivanov, S. (1988). The effect of improved child survival on family planning practice and fertility. Studies in Family Planning, 19(3), 141–161.
Mensch, B. S. (1985). The effect of child mortality on contraceptive use and fertility in Colombia, Costa Rica and Korea. Population Studies, 39, 309–327.
Nur, O. E. M. (1985). An analysis of the child survival hypothesis in Jordan. Studies in Family Planning, 16(4), 211–218.
Palloni, A., & Rafalimanana, H. (1999). The effects of infant mortality on fertility revisited: New evidence from Latin America. Demography, 36(1), 41–58.
Park, C. B., Ataharul, M., Chakraborty, N., & Kantner, A. (1998). Partitioning the effect of infant and child death on subsequent fertility: An exploration in Bangladesh. Population Studies, 52, 345–356.
Park, C. B., Han, S. H., & Choe, M. K. (1979). The effect of infant death on subsequent fertility in Korea and the role of family planning. American Journal of Public Health, 69(6), 557–565.
Preston, S. H. (Ed.). (1978). The effects of infant and child mortality on fertility. New York: Academic Press.
Rutstein, S., & Medica, V. (1978). The Latin American experience. In S. H. Preston (Ed.), The effects of child mortality on fertility. (pp. 93–112). New York: Academic Press.
Santow, G., & Bracher, M. D. (1984). Child death and time to the next birth in Central Java. Population Studies, 38, 241–253.
Sibanda, A., Woubalem, Z., Hogan, D. P., & Lindstrom, D. P. (2003). The proximate determinants of the decline to below replacement fertility in Addis Ababa, Ethiopia. Studies in Family Planning, 34(1), 1–7.
Singh, K. K., Suchindran, C. M., & Singh, K. (1993). Effects of breast feeding after resumption of menstruation on waiting time to next conception. Human Biology, 65(1), 71–86.
Taylor, C. E., Newman, J. S., & Kelly, N. U. (1976). The child survival hypothesis. Population Studies, 30(2), 263–278.
Trussell, J., & Richards, T. (1985). Correcting for unmeasured heterogeneity in hazard models using the Heckman-Singer procedure. In N. B. Tuma (Ed.), Sociological methodology, 1985. (pp. 242–276). San Francisco CA: Jossey-Bass.
Yamaguchi, K. (1991). Event history analysis. Newbury Park CA: Sage.
Yamaguchi, K. (1992). Accelerated failure-time regression models with a regression model of surviving fraction: An application to the analysis of “permanent employment” in Japan. Journal of the American Statistical Association, 87(418), 284–292.
Acknowledgments
Work on this study was supported by a grant from the David and Lucile Packard Foundation. The authors would like to thank the Central Statistical Authority of Ethiopia for providing access to the 1990 National Family and Fertility Survey. An earlier version of this paper was presented at the 2001 Annual Meeting of the Population Association of America, Washington, D.C.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Lindstrom, D.P., Kiros, GE. The impact of infant and child death on subsequent fertility in Ethiopia. Popul Res Policy Rev 26, 31–49 (2007). https://doi.org/10.1007/s11113-006-9018-1
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11113-006-9018-1