1 Thus, the increase in the number of births compensates for the decrease in the infant mortality gap in generating the number of "excess" deaths. In discussions of. gaps in labor market outcomes in the US intergroup health disparities, infant mortality is a frequently cited barometer. In section 2, we discuss the interaction of race and infant mortality in the context of a simple model of demand for and supply of broadly defined health care.
Moreover, southern blacks were relatively concentrated in rural areas, especially early in the period under study. Available evidence suggests that Southerners, particularly blacks, lagged far behind the rest of the country in health care consumption during most of the period under study. And fifth, the industrial employment opportunities open to blacks, especially early in the period under study, were sometimes physically poor (see Maloney and Whatley 1995).
Founded in 1924, the Duke Endowment has been instrumental in funding hospitals and health care for the poor in the Carolinas. Improvements in health care have had clear consequences for overall infant mortality rates, but the consequences for racial disparities in mortality rates are less clear and may depend on the extent of differential access to high-quality care, a topic we address later in the paper. Later in the paper, we re-explore potential differences between races in the impact of new technologies.
State-Level Data
15 See Steckel (1988) and Ferrie (2001) for micro-level studies of wealth and mortality in the nineteenth century. We have constructed targets for income per population by race, state, and year (YW, it and YNW, it), which ultimately depend on estimates of a few variables in the following two equations. We estimate Z for each state and census year using the detailed occupational information for individuals reported in the Integrated Public Use Microdata Series (IPUMS, Ruggles, and Sobek 1997).16.
We estimated years of education for the years before 1940 using individual-level information contained in the 1940 IPUMS sample. 19 We did not attempt to adjust the data for differences in the quality of education across races or states. He argues that educated blacks in the late nineteenth century had less schooling than whites than the 1940 census data suggested.
First, there may be errors in the measurement of the independent variables, particularly income and education.
Accounting for the Racial Gap in Infant Mortality
26 Interestingly, adding a variable for the percentage of the population (not registered) residing in relatively large urban areas (over 100,000 population) in columns 1 and 2 returns negative but not particularly large coefficients (-0.0018 for no whites; -0.0006 for whites ). Even with controls for income, education, urban residence, and physician supply, a strong secular trend in infant mortality is manifested in the downward march of the coefficients in the time period dummies. To what extent racial gaps in independent variables may account for the gap in infant mortality.
In Figure 3a we plot the proportion of the gap that is accounted for by racial differences in the observed characteristics at each point in time. From this perspective, about 80 percent of the gap can be accounted for by changes in characteristics up to 1945, but by 1970, only about a third of the gap can be explained. For most of the period under study, the largest contributor to the racial gap in infant mortality is the gap in the educational attainment of women ages 20 to 40.
But the racial gap in years of education narrowed rapidly over time, and by 1970 the education gap may account for only about 12 percent of the gap in infant mortality. By 1945, the income gap accounted for about 25 percent of the total gap in infant mortality, but by 1970, income accounted for only 15 percent of the gap. Finally, the geographic distribution of physicians accounts for 7 percent of the gap in infant mortality at the beginning of the study period, but with a redistribution of nonwhites (and to some extent also
27 To facilitate testing the significance of changes in the coefficients, we performed regressions with data for the entire period and interactions for the variables in the period after 1945. Another possibility is that the relationship between the observed characteristics and infant mortality changed in the post-war period and that, as a result, the coefficients estimated for the entire period 1920-1970, inaccurate descriptors of the 1950-1970 experience. The pattern in Figure 3b is strikingly similar to that in Figure 3a: from 1950 to 1970, observable characteristics explained less and less of the racial difference in infant mortality.
But for both rates, the portion of the gap explained by socioeconomic characteristics decreases over time (from about 63 to 27 percent for neonates; and from about 90 to 40 percent for post-neonates). In the next section, we discuss several variants of the first broad hypothesis—that changes in unobserved determinants of child health tended to widen the racial infant mortality gap, even as indicators of socioeconomic status converged.
Exploring the “Unexplained” Changes in the Infant Mortality Gap
In the next section, we discuss several variants of the first broad hypothesis—that changes in unobserved determinants of infant health tended to widen the infant mortality gap even when indicators of socioeconomic status converged. postneonatal mortality). This was especially true in the South, where the non-white infant mortality rate apparently rose slightly between 1950 and 1960, while the white rate fell by 6.1 per 1,000 births. Although these changes in weight distribution may appear small, they may have had important implications for the racial disparity in infant mortality because neonatal mortality among low-birth-weight infants was very high: about 160 births per thousand in the 1950s (for non-whites) compared with about 10 per thousand for newborns weighing more than 2,500 grams (Chase 1972, p. 9).
In fact, table 6 shows that given the 1950 distribution of non-white neonatal mortality rates (across the ten reported weight categories), the change in the distribution of non-white birth weights from 1950 to 1960 by itself increased the non-white neonatal infant mortality rate by 5, 7 would increase per thousand live births. An immediate concern is that the apparent shift in the nonwhite birth weight distribution is a statistical artifact related to the increasing proportion of nonwhite births occurring in hospitals. It is impossible to disprove, but table 7 shows that the proportion of low birth weight non-white babies has increased everywhere, including regions (outside the South) where birth registration rates were very high in 1950 and where there have been relatively small changes in the proportion of babies born , was. in hospitals.
Although the increasing proportion of low-birth-weight infants among nonwhites is the proximate cause of the rising infant mortality in the 1950s, the ultimate causes remain. Importantly, between 1965 and 1970, the relative movements in the proportion of low-birth-weight infants were small—there were small declines for both whites and non-whites. Therefore, the observed convergence of infant mortality rates in the late 1960s appears to have little to do with changes in the relative health of infants at birth.
This period was also marked by a dramatic decline in the proportion of mothers who breastfed their infants, and again the magnitude of the change in behavior appears to have been greater for non-whites than for whites. In 1940, the nonwhite rate of illegitimate births was 10 times the rate among whites in the states reporting the variable (15.6 percent versus 1.55 percent). In particular, during the period studied, young mothers (under 20 years of age) were much more likely to give birth to low birth weight infants than other women.
For whites, the impact is again small, but for nonwhites, the impact was relatively large, reflecting apparent increases in the probability of low birth weight in each age category for nonwhite mothers. In the South, the proportion of nonwhite hospital births increased from 24 to 74 percent between 1945 and 1960, while the white proportion increased from 68 to 97 percent. Did the racial insurance gap have implications for the quality of maternal and infant care?34 In the absence of data on the treatment intensity of infants by race and insurance status, it is not possible to construct a direct test of the hypothesis.
Second, according to Cone (1985), relatively rapid technological advances in the treatment of low birth weight infants began in the mid-1960s with the introduction of modern NICUs.
Conclusion
It is clear that the rapid decline in infant mortality rates during this period benefited both whites and non-whites. Using a panel of state-level race-specific data, we found that much of the racial gap in infant mortality rates could be accounted for by differences in those characteristics, particularly between 1920 and 1945. Nearby, the neonatal gap could was strongly influenced by a leftward shift in the nonwhite birth weight distribution during the 1950s, but the postneonatal gap, which is rather insensitive to weight at birth, failed to narrow either.
A History of Neglect: Health Care for Blacks and Mill Workers in the Twentieth-Century South. Five Decades of Action for Children: A History of the Children's Bureau. In Children and Youth: Social Problems and Social Policy, edited by Robert H. An American Health Dilemma, Volume II: Race, Medicine, and Health Care in the United States 1900-2000.
Long-term trends in health, welfare, and economic growth in the United States. In Health and Welfare during Industrialization, edited by R. Variations in Infant Mortality Rates among Counties of the United States: The Roles of Public Policies and Programs.” Demographics. Each component of the "Gap Explained By" section is the product of the relevant coefficients from table 2 and the difference in the variable's mean value for whites and non-whites (in that year).
Note: The "weight" is the proportion of the relevant race category's births in that region. Notes: Column 4 reflects the effect of the change in the birth weight distribution on the neonatal infant mortality rate. Notes: Column 4 reflects the effect of changes in the age distribution on the proportion of infants with low birth weight (below 2,500 grams).
Summing column 4 and column 6 gives the overall change in the percentage of infants born with low birth weight. Notes: The table is based on a simple decomposition of the racial difference in neonatal mortality rates. Notes: The graph plots a three-year moving average of the non-white-white gap in infant mortality.
Notes: The "calculated" part of the gap is the product of the racial gap in the observed characteristics and the simple average of the regression coefficients for whites and nonwhites.
