Cohort studies.

The NBER's Working Group on Cohort Studies met in Cambridge on February 25. NBER Research Associate Dora L. Costa of MIT directs the group and organized the meeting.

The NBER's Working Group on Cohort Studies investigates changes in well being across generations, with a strong focus on health The premise of the working group is that each generation is marked by particular lifetime experiences that affect It throughout the life cycle and that are transmitted to their children either through social conditioning or. in the case of some health conditions. through direct biological mechanisms. To understand the present we therefore need to look not only at early life conditions but also at the experience of previous generations. This requires the creation of new longitudinal datasets that cover the entire twentieth century, because many of the dramatic changes in life expectancy, health, and leisure and the consumer goods revolution occurred during the first half of the taventieth century. Assessing the importance of these changes in turn requires the quantification of the contribution of non-market goods to the national income accounts. The group's members therefore consist of specialists in economic history, family economics, health economics, and productivity. Many of the group's members are also closely involved with Robert Fogel's P01 Early Indicators of Later Work Levels. Disease, and Death that is creating longitudinal micro-level data on the first generations to reach age 65 at the beginning of the twentieth century.

These papers were discussed at the meeting:

Douglas Almond, NBER and Columbia University; and Kenneth Y. Chay, NBER and University of California. Berkeley, "The Long-Run and Intergenerational Impact of Poor Infant Health: Evidence from Cohorts Born During the Civil Rights Era"

Trevon D. Logan, Ohio State University, "On the Nutrition-Stature Hypothesis" Hoyt Bleakley, University of California, San Diego, and Fabian Lange, Yale University, "Chronic Disease Burden and the Interaction of Education, Fertility, and Growth"

Karen Norberg, Washington University, "Is the Sex Ratio at Birth a Social Indicator?"

Joseph Burton, University of Chicago; Robert Fogel, NBER and University of Chicago; and Sven Wilson, Brigham Young University, "Declining Disparities in Infant Mortality in the United States: Preliminary Results from a Sample of Ward Level Data"

Lorens A. Helmchen, University of Illinois at Chicago, and Louis Nguyen, Harvard Medical School, "Tangible Human Capital: The Effects of Amputations on Later Life Outcomes of Union Army Veterans"

The 1960s witnessed the greatest reduction in the mortality rates of black relative to white infants of the last 50 years. Almond and Chay use these dramatic changes to evaluate the long-run effects of health conditions in early life. The annual Natality Detail fries provide information on the characteristics and health risk factors of the universe of mothers giving birth in the United States, as well as the birth outcomes of their infants. The authors link these data to the infant health conditions that prevailed in the state and year in which the mother was born. Their analysis compares differences in the health and birth outcomes of black and white mothers born in the late 1960s to those of mothers born in the early 1960s, after controlling for age of the mother and the year in which she gave birth. Almond and Chay find that black women born in 1967-9 have substantially lower risk factor rates as adults and are much less likely to give birth to an infant with low birth weight and APGAR scores than black women born in 1961-3. The between-cohort gains for white women are small to non-existent, consistent with the smaller health improvements for white infants born during the 1960s. Further, the between-cohort relative gains are significantly larger for black women born in Mississippi than for black women born in Alabama. The authors conclude that the social policies that led to the infant health improvements (for example, hospital desegregation, Medicaid, Food Stamps) may have had long run and intergenerational health benefits. Also, studies of changes in racial health disparities over the life cycle and over time could be severely biased by not accounting for the significant changes in the health of black birth cohorts.

Social scientists who study stature have advanced markedly different hypotheses concerning the relationship between income, nutrition, and stature. Economic historians hypothesize that households in the nineteenth century substituted away from carbohydrates and fiber and towards protein and fat as their incomes rose. Anthropometric historians assert that the same gains in stature can be explained by increased nutrient intake without any nutritional substitution. Logan tests these hypotheses using the 1888 Cost of Living Survey. He shows that the income elasticity of fiber is greater than or equal to the income elasticity of protein, fat, or sugar--contrary to the nutritional substitution posited by economic historians. A food modified Engel curve reveals that the shares of carbohydrates, fat, and sugar in the diet vary with household income, but the shares of protein and fiber do not. However, Logan does find that the share of protein from animal sources increases with household income. In looking at the diet in terms of calories, he rejects the hypothesis of income-elasticity equality between cereal and meat calories, and also between cereal and dairy calories. He also finds that the diets of late nineteenth century industrial workers were surprisingly balanced by modern standards.

Bleakley and Lange provide new evidence on the appropriate model of the economic and demographic transition. They analyze the eradication of hookworm disease in the American South (c. 1910). In previous work, Bleakley showed that the eradication of hookworm disease led to a significant increase in school attendance and literacy. This study shows that the increase in human capital investment was accompanied by a decrease in fertility that was both economically and statistically significant. A decline in the hookworm infection rate from 40 to 20 percent is associated with a decline in fertility that amounts to 40 percent of the entire fertility decline observed in the American South between 1910 and 1920. The relative change in fertility and schooling caused by hookworm eradication is approximately equal to aggregate co-movements during the period considered. The authors argue that this is consistent with models of the fertility transition emphasizing economic incentives rather than changing cultural attitudes and birth control technologies. Furthermore, the data support models emphasizing intergenerational altruism. Variables affecting children's economic prospects affect parental fertility decisions. One consequence of this finding is that changes in the economic opportunities faced by parents directly are not necessary to explain the economic and demographic transition.

Could the sex ratio at birth be influenced by the family structures into which children are being born? All over the world, there are slightly more boys born than girls, but among African-Americans the proportion of male births is lower than among whites--and the reason for this difference is unknown. Norberg investigates the empirical and theoretical questions raised by a new finding: that parents' partnership status at the time of a child's conception predicts offspring's sex. First, in a sample of 86,436 human births pooled from five U.S. population-based surveys, she finds 51.5 percent male births reported by respondents who were living with a spouse or partner before the child's conception or birth, and 49.9 percent male births reported by respondents who were not. Because household type was observed before conception, the effect could not be explained by paternal bias against daughters; it also was unexplained by parental age, education, income, ethnicity, and year of observation, and was larger when comparisons were made between siblings. This effect was of similar magnitude in births occurring before and after legalization of abortion in the United States. Norberg further investigates the possible role of sex-selective abortion, by examining the effects of the legalization of abortion on ethnicity, state, and year-specific aggregate counts of male and female births. Here, she finds that legalization of abortion immediately reduced the proportion of male births by about 1.5 births in 1000, among both whites and blacks, and among married and unmarried mothers; in other words, the effect of legalization of abortion is about 10 percent of the magnitude of the effect of partnership status, and seems most consistent with the use of abortion to avert the birth of children at risk for sex-linked genetic disorders. Furthermore, legalization of abortion did not change the estimated effects of ethnicity or partnership status. In both the individual-level and aggregate data, Norberg finds that household type explains at least half of the black/white difference in the sex ratio at birth. The effect is also of the right magnitude to explain many reported associations between child sex and family composition after birth.

Burton, Fogel, and Wilson take up the case of the infant mortality rate (IMR) during the first half of the twentieth century in the United States. They present first results from an analysis of ward-level data in the 24 largest U.S. cities in 1900. Their initial findings show that: 1) absolute variation in IMR across wards was strikingly high at the turn of the century in all the U.S. cities for which they have data; 2) the decline in intra-city variation was both profound and rapid; and 3) the year-to-year variation in IMR within wards was often substantial. They illustrate these findings with a case study of Milwaukee. In 1900, IMR by ward ranged from 104 per thousand in the lowest ward to 350 per thousand in the highest. Even though the city average of 175 was significantly higher than the national average of about 125, four wards actually had lower than average IMR. By 1918, the range in Milwaukee had fallen from 19 to 108. This illustrates how the use of relative disparity ratios tells a very incomplete and potentially misleading story of the trend in disparities. The absolute variation in Milwaukee fell from 246 in 1900 to 89 in 1900, but the ratio between the worst and best wards actually increased from 3.4 to 5.7. The authors show that relative disparity measures can frequently show that disparities are worsening when the trends in absolute differences between groups show significant improvement.

Apart from the loss of life, the loss of limbs was arguably the most immediate, and enduring, human cost of the U.S. Civil War. Helmchen and Nguyen use a representative sample of more than 30,000 Union Army veterans to study the long-term medical, social, economic, and psychological effects of amputations in this population. The large size and the representative nature of the sample, combined with the diverse set of outcome measures and the long time horizon that they cover, allow the authors to estimate the cost that an amputation imposed on a veteran. By exploiting variation in anatomic level and function that was lost, they analyze how amputees compensated for their new but irreversible disability in the manual-labor-intensive work environment of the United States in the second half of the nineteenth century. This work also has implications for contemporary societies where amputations resulting from landmines are still a concern.