scispace - formally typeset
Search or ask a question
Journal ArticleDOI

Measuring the effect of income on adult mortality using longitudinal administrative record data.

01 Jan 1986-Journal of Human Resources (JSTOR)-Vol. 21, Iss: 2, pp 238-251
TL;DR: This article examined the effect of income controlling for education on the mortality of white married men aged 35-65 and found that low income continues to have a large and significant effect on mortality risk controlling for disability and on the probability of death through its effect on disability.
Abstract: This study enhances the 1973 CPS-IRS-SSA Exact Match File with more complete Social Security mortality data for 1973-1978 and with updated Social Security earnings and disability data. It uses the resulting data set to examine the effect of income controlling for education on the mortality of white married men aged 35-65. It finds that low income continues to have a large and significant effect on mortality risk controlling for disability and on the probability of death through its effect on disability. (EXCERPT)
Citations
More filters
Journal Article
TL;DR: In this paper, a framework for causal inference in settings where assignment to a binary treatment is ignorable, but compliance with the assignment is not perfect so that the receipt of treatment is nonignorable.

4,129 citations

Journal ArticleDOI
TL;DR: It is shown that the instrumental variables (IV) estimand can be embedded within the Rubin Causal Model (RCM) and that under some simple and easily interpretable assumptions, the IV estimand is the average causal effect for a subgroup of units, the compliers.
Abstract: We outline a framework for causal inference in settings where assignment to a binary treatment is ignorable, but compliance with the assignment is not perfect so that the receipt of treatment is nonignorable. To address the problems associated with comparing subjects by the ignorable assignment—an “intention-to-treat analysis”—we make use of instrumental variables, which have long been used by economists in the context of regression models with constant treatment effects. We show that the instrumental variables (IV) estimand can be embedded within the Rubin Causal Model (RCM) and that under some simple and easily interpretable assumptions, the IV estimand is the average causal effect for a subgroup of units, the compliers. Without these assumptions, the IV estimand is simply the ratio of intention-to-treat causal estimands with no interpretation as an average causal effect. The advantages of embedding the IV approach in the RCM are that it clarifies the nature of critical assumptions needed for a...

3,978 citations

ReportDOI
TL;DR: This article examined whether education has a causal impact on health and found that it has a large and positive correlation between education and health, and that this effect is perhaps larger than has been previously estimated in the literature.
Abstract: Prior research has uncovered a large and positive correlation between education and health. This paper examines whether education has a causal impact on health. I follow synthetic cohorts using successive U.S. censuses to estimate the impact of educational attainment on mortality rates. I use compulsory education laws from 1915 to 1939 as instruments for education. The results suggest that education has a causal impact on mortality, and that this effect is perhaps larger than has been previously estimated in the literature. Copyright 2005, Wiley-Blackwell.

958 citations

Journal ArticleDOI
TL;DR: Both ordinary and IV estimates indicate that increases in income significantly improve mental and physical health but increase the prevalence of alcohol consumption.

740 citations

Journal ArticleDOI
TL;DR: The issues of low income and income instability should be addressed in population health policy and the impact of income change on mortality risk is examined.
Abstract: OBJECTIVES: The aim of this study was to examine relationships between income and mortality, focusing on the predictive utility of single-year and multiyear measures of income, the shape of the income gradient in mortality, trends in this gradient over time, the impact of income change on mortality, and the joint effects of income and age, race, and sex on mortality risk. METHODS: Data were taken from the Panel Study of Income Dynamics for the years 1968 through 1989. Fourteen 10-year panels were constructed in which predictors were measured over the first 5 years and vital status over the subsequent 5 years. The panels were pooled and logistic regression was used in the analysis. RESULTS: Income level was a strong predictor of mortality, especially for persons under the age of 65 years. Persistent low income was particularly consequential for mortality. Income instability was also important among middle-income individuals. Single-year and multiyear income measures had comparable predictive power. All eff...

380 citations

References
More filters
ReportDOI
TL;DR: This article developed a methodological framework that can be used to introduce and discuss alternative explanations of the correlation between health and schooling and test these explanations empirically in order to select the most relevant ones and obtain quantitative estimates of different effects.
Abstract: This paper has two purposes. The first is to develop a methodological framework that can be used to introduce and discuss alternative explanations of the correlation between health and schooling. The second is to test these explanations empirically in order to select the most relevant ones and obtain quantitative estimates of different effects. The empirical work is limited to one rather unique body of data and uses two measures of health that are far from ideal. The methodological framework can, however, serve as a point of departure for future research when longitudinal samples with more refined measures of current and past health and background characteristics become available.

354 citations

Posted Content
TL;DR: The relationship of mortality of whites to both medical care and environmental variables is examined in a regression analysis across states in 1960 as discussed by the authors, where medical care is alternatively measured by expenditures and by the output of a Cobb-Douglas production function combining the services of physicians, paramedical personnel, capital and drugs.
Abstract: The relationship of mortality of whites to both medical care and environmental variables is examined in a regression analysis across states in 1960. Medical care is alternatively measured by expenditures and by the output of a Cobb-Douglas production function combining the services of physicians, paramedical personnel, capital, and drugs. Simultaneous equation bias resulting from the influence of factor supply curves and demand for medical care is dealt with by estimating a more complete model. Both two-stage least squares and ordinary least squares estimates are presented. The elasticity of the age-adjusted death rate with respect to medical services is about -0.1. Environmental variables are far more important than medical care. High education is associated with relatively low death rates. High income, however, is associated with high mortality when medical care and education are controlled for. This may reflect unfavorable diets, lack of exercise, psychological tensions, etc. The positive association of mortality with income may explain the failure of death rates to decline rapidly in recent years. Adverse factors associated with the growth of income may be nullifying beneficial effects of increases in the quantity and quality of care. If so, the view that we have reached a biological limit to the death rate is not valid.

335 citations

Posted Content
TL;DR: This paper found that the negative relation between schooling and smoking observed at age 24 is accounted for by differences in smoking behavior present at age 17, when all subjects were still in approximately the same grade.
Abstract: Numerous studies by economists during the past decade have revealed a large, statistically significant correlation between health and years of schooling after controlling for differences in income and other variables. Cigarette smoking is a likely intervening variable because of the strong effect of smoking on morbidity and mortality, and because there is a strong negative correlation between smoking and years of schooling -- at least at high school levels and above. This paper tests the hypothesis that schooling causes differences in smoking behavior. We use retrospective smoking histories of 1,183 white, non-Hispanic men and women who had completed 12 to 18 years of schooling. The data were collected in 1979 by the Stanford University Heart Disease Prevention Program from randomly selected households in four small California cities. The most striking result is that the negative relation between schooling and smoking observed at age 24 is accounted for by differences in smoking behavior present at age 17, when all subjects were still in approximately the same grade. We conclude that additional years of schooling cannot be the cause of differential smoking behavior; one or more "third variables" must cause changes in both smoking and schooling. Analysis of smoking by cohort reveals that the schooling-smoking correlation developed only after the health consequences of smoking became widely known; it has remained strong even in the most recent cohorts. This implies that the mechanism behind the schooling-smoking correlation may also give rise to the schooling-health correlation.

290 citations

Posted Content
TL;DR: The authors developed a methodological framework that can be used to introduce and discuss alternative explanations of the correlation between health and schooling and test these explanations empirically in order to select the most relevant ones and obtain quantitative estimates of different effects.
Abstract: This paper has two purposes. The first is to develop a methodological framework that can be used to introduce and discuss alternative explanations of the correlation between health and schooling. The second is to test these explanations empirically in order to select the most relevant ones and obtain quantitative estimates of different effects. The empirical work is limited to one rather unique body of data and uses two measures of health that are far from ideal. The methodological framework can, however, serve as a point of departure for future research when longitudinal samples with more refined measures of current and past health and background characteristics become available.

289 citations