Cohort Studies Meeting
April 13-14, 2012
Frank R. Lichtenberg, Columbia University and NBER
Lichtenberg uses patient-level data to analyze the effect of technological change embodied in pharmaceuticals on the longevity of elderly Americans. Previous patient-level studies could not control for important patient attributes such as education, income, and race; they did not provide estimates of the effect of using newer drugs on life expectancy, or of the overall cost effectiveness of new drugs relative to old drugs; and they were not based on nationally representative samples of individuals. The data here, primarily derived from the Medical Expenditure Panel Survey and the Linked Mortality Public-use Files, helps to overcome those limitations. Lichtenberg investigates the effect of the vintage (year of FDA approval) of the prescription drugs used by an individual on his or her survival and medical expenditure, controlling for a number of demographic characteristics and indicators and determinants of health status. When he only controls for age, sex, and interview year, he estimates that a one-year increase in drug vintage increases life expectancy by 0.52 percent. Controlling for a much more extensive set of other attributes (the mean year the person started taking his or her medications, and dummy variables for activity limitations, race, education, family income as a percent of the poverty line, insurance coverage, Census region, BMI, smoking and over 100 medical conditions) has virtually no effect on the estimate of the effect of drug vintage on life expectancy. Between 1996 and 2003, the mean vintage of prescription drugs increased by 6.6 years. This is estimated to have increased life expectancy of elderly Americans by 0.38 years. This suggests that 63 percent of the 0.6-year increase in the life expectancy of elderly Americans during 1996-2003 was due to the increase in drug vintage. The 1996-2003 increase in drug vintage is also estimated to have increased annual drug expenditure per elderly American by $194, and annual total medical expenditure per elderly American by $286. This implies that the incremental cost-effectiveness ratio (cost per life-year gained) of pharmaceutical innovation was about $15,000. This estimate of the cost per life-year gained from the use of newer drugs is a small fraction of leading economists' estimates of the value of (willingness to pay for) an additional year of life. It is also consistent with estimates from clinical trials.
Steven Lehrer, Queen's University and NBER; Nicholas Christakis, Harvard University; and James Rosenquist, Massachusetts General Hospital
A growing body of literature documents the relative importance of cohorts' early life conditions as compared to later period environments on a variety of health and socioeconomic outcomes. If genetic factors interact with environmental changes, then these biological channels may explain some of the heterogeneity in cohort effect estimates. Christakis, Lehrer, and Rosenquist analyze data from the Framingham Heart Study, a unique dataset that follows adults born across four decades for over 30 years, and which contains direct information on molecular genetic variation at the individual level of specific markers -- in the behavioral genetic literature, these markers have been strongly linked with BMI (FTO, MC4R) and smoking intensity (CHRNA3-5). The researchers estimate age-period-cohort models and include a rich set of interactions with genetic variants in an effort to understand the relative importance of the timing of gene-environment effects. They show that the effect of all three of the genetic variants have strong and statistically significant interactions with birth cohort indicators on the respective risky behaviors. In contrast, genetic interactions with contemporaneous environmental and age indicators do not have a significant relationship. The results are robust to the inclusion of family fixed effects to capture dynastic effects and are consistent with a growing literature in evolutionary biology that proposes there is an "adaptive disadvantage" affecting differential risk for disease states. They conclude by discussing the implications of these findings for research in economics, behavioral genetics, and public policy.
Costas Meghir, Yale University; Marten Palme, Stockholm School of Economics; and Emilia Simeonova, Princeton University and NBER
Meghir, Palme, and Simeonova study the effect of a compulsory education reform in Sweden on adult health and mortality. The reform was implemented by municipalities between 1949 and 1962 as a social experiment and implied an extension of compulsory schooling from seven or eight years depending on municipality to nine years nationally. The authors use detailed individual data on education, hospitalizations, labor force participation and mortality for Swedes born between 1946 and 1957. The individual-level data allow them to study the effect of the education reform on three main groups of outcomes: 1) mortality until age 60 for different causes of death; 2) hospitalization by cause; and 3) exit from the labor force primarily through the disability insurance program. Their results show reduced male mortality up to age 50 for those assigned to the reform, but these gains were erased by increased mortality later on. They find similar patterns in the probability of being hospitalized and the average costs of inpatient care. Men who acquired more education due to the reform are less likely to retire early.
Chulhee Lee, Seoul National University
Prenatal exposure to the disruptions caused by the Korean War (1950–3) negatively affected individual socioeconomic and health outcomes at older ages. Lee finds that the educational attainment and labor market performance of the subjects of the 1951 birth cohort, who were in utero during the worst time of the war, were significantly lower in 1990 and 2000. The results of difference-in-difference estimations suggest that the magnitude of the negative cohort effect is significantly larger for individuals who were more seriously traumatized by the war. As for health outcomes, the 1950 male birth cohort exhibited a significantly higher disability rate in 2005 and the women married to those men were more likely to be disabled at old age. If potential selections in pregnancy, birth, and survival are considered, the negative effects of the war may be even greater than suggested in this study. The long-term effects of in-utero circumstances differ by gender, which may be attributable to the strong population selection for the 1951 female cohort and to the influence of the husband's health status on a woman's health. Different aspects of human capital (for example, health and cognitive skills) were impaired by in-utero exposure to the war, depending on the stage of pregnancy when the negative shocks were experienced.
Martha J. Bailey, University of Michigan and NBER; and Andrew Goodman-Bacon, University of Michigan
Bailey and Goodman-Bacon use the roll out of the first Community Health Centers (CHCs) to estimate the long-term health effects of increasing access to primary care. Their results show that CHCs reduced age-adjusted mortality rates among those 50 and older by almost 2 percent within ten years. The implied 6-to-8-percent decrease in the one-year mortality risk among the treated group amounts to 18-to-24 percent of the 1966 poor-nonpoor mortality gap for this age group. Large effects for those 65 and older suggest that increased access to primary care has long-term benefits, even for populations with near universal health insurance.
Stefania Albanesi, Columbia University and NBER
Albanesi examines the impact of the decline in maternal mortality on fertility and women's human capital. Fertility theory suggests that a permanent decline in maternal mortality initially increases fertility and generates a permanent rise in women's human capital, relative to men. The resulting rise in the opportunity cost of children leads to a subsequent decline in desired fertility, generating a boom-bust response. Albanesi assesses these predictions using newly digitized data on maternal mortality for 25 advanced and emerging economies for the period 1900-2000. The empirical estimates suggest that the decline in maternal mortality contributed significantly to the baby booms and subsequent baby busts experienced by these economies in the twentieth century, and that the female-male differential in education attainment grew more in those countries that experience a sizable maternal mortality decline.
Richard Hornbeck, Harvard University and NBER; and Suresh Naidu, Columbia University and NBER
The availability of low-wage immobile labor may discourage economic development. In the American South, post-bellum economic stagnation has been attributed in part to white landowners' access to immobile low-wage black workers; indeed, subsequent Southern economic convergence was associated with substantial black out-migration. Hornbeck and Naidu estimate that the 1927 Mississippi flood caused immediate and persistent out-migration of black workers from flooded counties. Following this decline in the availability of low-wage black labor, landowners in flooded counties dramatically mechanized and modernized agricultural production relative to landowners in nearby similar non-flooded counties. The temporary displacement of black workers led to a permanent economic transition, although landowners had incentives to discourage black out-migration and to maintain a system of labor-intensive agricultural production.
Evan Roberts, University of Minnesota, and Pamela Wood, Monash University
Roberts and Wood provide evidence on the fetal origins hypothesis in an historical cohort born between 1907 and 1922 in New Zealand, one of the earliest birth cohorts ever documented. By linking maternity hospital records to military enlistment records from World War II, they have a unique historical dataset with accurate measures of health and socioeconomic status at birth and in early adulthood. Using birth weight as an indicator of early health, they find associations between it and early adult health that are slightly larger than in other studies. They estimate that an increase in birth weight of 1kg is associated with an increase in adult height of 2.5 cm, and a decrease in systolic blood pressure of 2.4mm/Hg. These results suggest that adult health was more sensitive to socio-economic differences at birth than previously estimated.