P1: ARK/MBL/sny P2: MBL/mkv QC: MBL/abe T1: MBL June 4, 1997 19:3 Annual Reviews AR035-07 INFANT MORTALITY 149 technologic determinants has operated to decrease absolute rates while simul- taneously maintaining broad disparities in these rates (Wise & Pursley 1992). In this empirical context, the perspective of Ogburn (Ogburn 1969) suggests that innovation in technology has resulted in a broad array of sociological con- sequences, only some of which have been examined. We have accordingly structured this review to critically examine three issues: 1. The evolution of infant mortality patterns in the United States 2. The extent to which the long-documented mechanisms of socioeconomic influence on infant mortality have been altered by a technical revolution in modern health services for pregnant women and newborns 3. The impact of these altered causal pathways on the conceptual legitimacy of using infant mortality rates as a marker of social and economic inequalities While our focus is on infant mortality, the empirical and theoretical work we cite can be directly applied to other arenas of disparate health outcomes as well. We focus on the United States experience, although similar issues arise in the other industrialized countries of the world. THE EVOLUTION OF INFANT MORTALITY PATTERNS IN THE UNITED STATES One of the most enduring indictments of social and health policies in the United States is the consistently poor showing of the United States in international rank- ings of infant mortality. Although the United States is the wealthiest country in the world and expends more per capita on health and medical care than any other country, our infant mortality rates consistently rank below those of many other nations���twenty-second in 1990, below Italy, Spain, and Singapore (UNICEF 1992) (see Figure 1). While there are some reporting differences be- tween the United States and other countries that contribute to these disparities (e.g. in some countries, an infant death during the first few hours after birth is not counted as a live birth), these can account for only a part of the differences observed (Hartford 1992). The clear message derived from these data is that the United States ranks lower than all the other industrialized countries in infant survival. In addition to a relatively high overall rate of infant mortality, there are pronounced disparities within the United States among social, economic, racial, and ethnic groups and geographic areas. Some communities in the United States experience infant mortality rates resembling those of developing countries. The Annu. Rev. Sociol. 1997.23:147-170. Downloaded from www.annualreviews.org by UNIVERSITY OF WASHINGTON on 10/10/10. For personal use only.

P1: ARK/MBL/sny P2: MBL/mkv QC: MBL/abe T1: MBL June 4, 1997 19:3 Annual Reviews AR035-07 150 GORTMAKER & WISE Figure 1 #Infant mortality rates by country 1990. Figure 2 Black and white infant mortality rates in the United States 1950���1993. reasons for these disparities are complex but are generally thought to reflect the powerful effects of social and economic inequality. One important indicator of continuing disparity is provided by trends in the infant mortality rates of black (African-American) and white infants in the United States during the period 1950���1993 (see Figure 2). These data document the persistently higher rate of infant mortality among African-American infants Annu. Rev. Sociol. 1997.23:147-170. Downloaded from www.annualreviews.org by UNIVERSITY OF WASHINGTON on 10/10/10. For personal use only.

P1: ARK/MBL/sny P2: MBL/mkv QC: MBL/abe T1: MBL June 4, 1997 19:3 Annual Reviews AR035-07 INFANT MORTALITY 151 Figure 3 Infant mortality rates by household yearly income in the United States 1964���1966 and 1988. in the United States and describe a black/white disparity that has not been narrowing during the past three decades. A striking recent confirmation of the persistence of economic disparities can be seen from tabulations of data from two national infant mortality studies in the United States, the 1964���1966 National Natality and Infant Mortality Followback Surveys (NNS and NIMS) (National Center for Health Statistics 1972) and the 1988 National Maternal and Infant Health Survey (Sanderson et al 1991) (Figure 3). These data provide the only national estimates of disparities in infant mortality by household income. The income categories are only approximately similar, because categories of income were used in the survey, adjusted for changes in the Consumer Price Index (United States Bureau of the Census 1989). The 1964���1965 data precedes the implementation of Medicaid and thus pro- vides important ���baseline��� information concerning economic disparities. The 1988 data comes after more than a decade of operation of this major health care financing program that includes poor families with children. Early evaluations of the implementation of Medicaid indicate that it helped to reduce dispari- ties in the utilization of health services between poor and nonpoor families (Butler & Scotch 1978), while more recent studies have documented problems in access and declining physician participation in Medicaid (Yudkowsky et al 1990). The data displayed in Figure 3 indicate no attenuation in the relative risk of infant mortality experienced by women with lower, compared with higher, Annu. Rev. Sociol. 1997.23:147-170. Downloaded from www.annualreviews.org by UNIVERSITY OF WASHINGTON on 10/10/10. For personal use only.