在印度,過盛的女性孩童死亡率一直是個無法紓解的特有現象,尤其是在北印度的旁遮普省與哈里亞納省。本文欲運用三份1992-2006期間之印度國家家庭衛生調查(National Family Health Survey, NFHS)統計資料,全面性的探討孩童死亡率性別差距之基本真相。更詳細的說,本文欲探討以下三個問題:(一)旁遮普省與哈里亞納省的孩童死亡率之性別差距在過去二十年間是否有減少?(二)糧食、預防與治療照護之分配不均,是否能解釋既有的孩童死亡率之性別差距?(三)有哪些因素能夠解釋孩童死亡率之性別差距?本文使用卜瓦松迴歸(Poisson Regression)研究孩童死亡率性別差距之決定因素。考克斯迴歸(Cox Regression)研究結果所呈現的存活曲線,顯示兩個重要的發現:初生女嬰的存活率較男嬰高;而在嬰兒後期與孩童時期則有相反現象。
Excess female mortality during childhood is a distinct and unrelieved phenomenon in India, particularly in the two northern Indian states- Punjab and Haryana. This paper investigates the basic facts of sex differentials in child mortality in a very comprehensive manner using the three rounds of NFHS data sets from 1992 to 2006. More specifically, this paper examines the following three basic questions: (1) Have the sex differentials in child mortality in Punjab and Haryana narrowed down during the past two decades? (2) Does discrimination in food as well as preventive and curative care explain the existing sex differentials in child mortality? And (3) What are factors that explain the sex differentials in child mortality? The determinants of childhood mortality have been studied through Poisson regression. The survival curves resulted from Cox regression reveal two important findings: survival chances for females during neonatal period is higher, while an opposite phenomenon appears in the post-neonatal and childhood ages.