NBER Working Paper 19311, August 2013.

Starting in 2006, respondents in the Health and Retirement Study were asked to submit biomarkers and notified of certain results. Respondents with very high blood pressure were notified during the interview; all were notified by mail of their BP, hemoglobin A1c, and total and HDL cholesterol alongside recommended thresholds. Average effects of biomarker collection were near zero, but notification of rare and dangerous readings triggered new diagnoses, increased pharmaceutical usage, and altered health behaviors among respondents and spouses. Very high BP or A1c readings raised new diagnosis and medication usage by 20 to 40 percentage points. Uncontrolled high BP triggered reductions in smoking and own and spouse's drinking. High A1c was associated with a 2.2 percent drop in weight and an increase in exercise among undiagnosed respondents, but with no changes among those already diagnosed, whose self-reported health and disability worsened.

Media coverage: Modern Healthcare

Unpublished working abstract, September 2014.

Numerous studies have revealed an Hispanic health paradox, in which health and mortality outcomes for U.S. Hispanics are better than one would expect based on socioeconomic status (SES). Less well understood is the SES gradient in health among Hispanics, which tends to be flatter than among non-Hispanics, may be nonexistent or even downward sloping, and is also flatter in countries of origin. Previous investigations have focused on mortality and on self-reports, and they have highlighted the role of immigrant status. In this study, I extend the analysis of the Hispanic SES gradient to the broad array of subjective and objective health outcomes and behaviors in the U.S. Health and Retirement Study, a biennial panel with retrospective data on migration and health that includes 1,700 Hispanics, half foreign born. Flatter SES gradients are the remarkably consistent finding especially in self-reported health metrics. Gradients in objective measures resemble those of non-Hispanics.

with Carl Mason

Unpublished working paper, September 2013.

We assess average net longevity benefits of bicycle commuting in the U.S. by constructing age-specific fatality rates from official fatality statistics and the 2009 National Household Travel Survey. We model the impact on the life table of switching from car to bicycle commuting. Bicycling fatality rates in the U.S. are an order of magnitude higher than in Europe. These costs follow an age pattern that punishes both young and old, while the health benefits guard against causes of mortality that rise rapidly with age. The lifetime health benefits of bicycle commuting appear to outweigh the costs, but individuals who sufficiently discount or disbelieve the health benefits may delay or avoid bicycling. Bicycling in middle age avoids much fatality risk while capturing health benefits. Significant cross-state variation in bicycling mortality risks suggest that safety improvements in the built environment might spur changes in transit mode.

with Alice Zulkarnain

Working paper, October 2012.

Length of life is a key indicator of well being, and expectations about longevity are important for life-cycle behavior. Earlier research shows that subjective survival probabilities reveal information about health and mortality; here, we also consider the age-shape of survivorship. We examine how subjective survivorship probabilities vary across age at a point in time in the U.S. Health and Retirement Study, how the level and shape of subjective survivorship compares with official life tables and changes with individual characteristics, how they evolve over time in response to new information like health shocks and parental death, and how they reflect actual mortality experiences in the panel. Beliefs can deviate from life tables but reflect relevant private information. The death of a same-sex parent appears to be more salient to individuals than physicians' diagnoses, but the latter are more predictive of future mortality and the shape of the survivorship curve.

Revise and resubmit, Education Economics

Previously NBER Working Paper 15778, February 2012.

In this paper, I explore how the timing of education across the life cycle is associated with older-age health outcomes and socioeconomic status among military retirees, a subpopulation with common levels of adolescent health but variation in educational timing. The marginal effect of education on the probability of fair or poor health around age 55 declines about half a percentage point with each decade of age at acquisition, but the effects on income and wealth are more constant. This suggests that education improves health through fostering the lifelong accumulation of healthy behaviors and habits rather than raising income or wealth.

International Journal of Economic Theory 6(2): 205-225, June 2010.

This paper examines optimal portfolio choice when health can change the shape of the utility function. If adverse health shocks threaten to increase the marginal utility of consumption, that is, if they are Edgeworth-Pareto substitutes, risky health prompts individuals to lower their risky portfolio shares. Health naturally becomes more uncertain with age, so this theory may help explain why aging investors gradually decrease their risk exposure even when asset returns display no mean reversion and relative risk aversion is constant.

Journal of Business and Economic Statistics 26(4): 472-485, October 2008.

This paper investigates the role of self-perceived risky health in explaining continued reductions in financial risk taking after retirement. If future ad- verse health shocks threaten to increase the marginal utility of consumption, either by absorbing wealth or by changing the utility function, then health risk should prompt individuals to lower their exposure to financial risk. I examine individual-level data from the Study of Assets and Health Dynamics Among the Oldest Old (AHEAD), which reveal that risky health prompts safer investment. Elderly singles respond the most to health risk, consistent with a negative cross partial deriving from health shocks that impede home production. Spouses and planned bequests provide some degree of hedging. Risky health may explain 20% of the age-related decline in financial risk taking after retirement.

Preventive Medicine 46(1):14-21, January 2008.

This paper assesses the potential benefits of increased walking and reduced obesity associated with taking public transit in terms of dollars of medical costs saved and disability avoided. I conduct new analysis of a nationally representative U.S. transportation survey to gauge the net increase in walking associated with public transit usage. I translate minutes spent walking into en- ergy expenditures and reductions in obesity prevalence, estimating the present value of costs and disability that may be avoided. Taking public transit is associated with walking 8.3 more minutes per day on average, or an additional 25.7-39.0 kcal. Hill et al. (2003) estimate an increase in net expenditure of 100 kcal/day can stop the increase in obesity in 90% of the population. Additional walking associated with public transit could save $5,500 per person in present value by reducing obesity-related medical costs. Savings in quality-adjusted life years could be even higher. While no silver bullet, walking associated with public transit can have a substantial impact on obesity, costs, and well-being. Further research is warranted on the net impact of transit usage on all behaviors, including caloric intake and other types of exercise, and on whether policies can promote transit usage at acceptable cost.

Journal of Population Economics 26(4): 1485-1522, October 2013.

A considerable amount of uncertainty surrounds the length of human life. The standard deviation in adult life span is about 15 years in the U.S., and theory and evidence suggest it is costly. I calibrate a utility-theoretic model of preferences over length of life and show that one fewer year in standard deviation is worth about half a mean life year. Differences in the standard deviation exacerbate cross-sectional differences in life expectancy between the U.S. and other industrialized countries, be- tween rich and poor countries, and among poor countries. Accounting for the cost of life-span variance also appears to amplify recently discovered patterns of convergence in world average human well-being. This is partly for methodological reasons and partly because unconditional variance in human length of life, primarily the component due to infant mortality, has exhibited even more convergence than life expectancy.

Population and Development Review 37(3): 499-528, September 2011.

Previous research has revealed much global convergence over the past several decades in life expectancy at birth and in infant mortality, which are closely linked. But trends in the variance of length of life, and in the variance of length of adult life in particular, are less well understood. I examine life-span inequality in a broad, balanced panel of 180 rich and poor countries observed in 1970 and 2000. Convergence in infant mortality has unambiguously reduced world inequality in total length of life starting from birth, but world inequality in length of adult life has remained stagnant. Underlying both of these trends is a growing share of total inequality that is attributable to between-country variation. Especially among developed countries, the absolute level of between-country inequality has risen over time. The sources of widening inequality in length of life between countries remain unclear, but signs point away from trends in income, leaving patterns of knowledge diffusion as a potential candidate.

with Shripad Tuljapurkar

Demographic Research 24(21): 497-526, March 2011.

The slope and curvature of the survivorship function reflect the considerable amount of variance in length of life found in any human population. Part is due to the well- known variation in life expectancy between groups: large differences according to race, sex, socioeconomic status, or other covariates. But within-group variance is large even in narrowly defined groups, and changes substantially and inversely with the group average length of life. We show that variance in length of life is inversely related to the Gompertz slope of log mortality through age, and we reveal its relationship to variance in a multiplicative frailty index. Our findings bear a variety of implications for modeling and forecasting mortality. In particular, we examine how the assumption of proportional hazards fails to account adequately for differences in subgroup variance, and we discuss how several common forecasting models treat the variance along the temporal dimension.

with Shripad Tuljapurkar

Population and Development Review 31(4): 645-675, December 2005.

The second half of the twentieth century witnessed much convergence in life expectancy around the world. We examine differences in the age pattern of mortality between countries over time to show that inequality in adult life spans, which we measure with the standard deviation of life table ages at death above age 10, S₁₀, is increasingly responsible for the remaining divergence in mortality. We report striking differences in level and trend of S₁₀ across industrialized countries since 1960, which cannot be explained by aggregate socioeconomic inequality or differential external-cause mortality. Rather, S₁₀ reflects both within and between-group inequalities in life spans and conveys new information about their combined magnitudes and trends. These findings suggest that the challenge for health policies in this century is to reduce inequality, not just lengthen life.

Unpublished working paper prepared for session 114 of the 2011 Annual Meeting of the Population Association of America, April 2011.

By the end of 2009, the Great Recession that began in December 2007 had doubled the unemployment rate, reduced stock prices by 25 percent, and lowered housing prices by 10 percent. The advent of the monthly American Time Use Survey (ATUS) in 2003 facilitates a detailed examination of trends in time allocations following these large macroeconomic shocks to the prices of time and assets. Previous research has explored the effects of being unemployed on time use and well-being, while here I explore the broader impacts of macroeconomic fluctuations on time use by all consumers using the 2003-2009 waves of the ATUS. When unemployment is high, consumers report spending more time sleeping, preparing food and eating or drinking, socializing and relaxing, and using the telephone, while they spend less time working and traveling for work. High unemployment also increases time spent by some subgroups on caring for children and adults outside the household. These effects are largely independent of labor force status, suggesting they are broad-based and reflective of incremental changes in the price of time rather than large jumps associated with the onset of unemployment. Wealth effects associated with lower housing prices shift time use toward home production and away from leisure. These results shed new light on the channels through which macroeconomic fluctuations affect health and well-being.

Unpublished working paper prepared for session 101 of the 2010 Annual Meeting of the Population Association of America, April 2010.

Do reductions in mortality lead to increases or decreases in income per capita? Recent research has called into question the emergent view from the late 1990s of health improvements as an important contributor to labor productivity and thus income per capita. While improved health certainly raises the quality of labor, the technique of development accounting typically recovers a relatively small direct effect of health on growth through this channel. Other studies have revealed evidence that improvements in health produce increases in population size and thus capital shallowing, which reduces income per capita through a "Solow effect." In this paper, I use a new dataset on the average and variance of adult length of life around the world to reexamine the role of the mortality transition in promoting economic growth. I find that longer adult life is associated with higher levels of capital accumulation, as life-cycle theory suggests, and with higher technology, even when controlling for the effects of institutions. But institutions appear to be relatively more important than the mortality environment for human capital accumulation.

The B.E. Journal of Macroeconomics: Contributions 9(1): Article 7, March 2009.

Sustained growth in both incomes and life spans are the hallmarks of modern development. Fluctuations around trend in the former, or business cycles, have been a traditional focus in macroeconomics, while similar cyclical patterns in mortality are also interesting and are now increasingly studied. In this paper, I assess the welfare implica- tions of cyclical fluctuations in mortality using a standard model of intertemporal preferences. Mirroring the classic result of Lucas (1987) regarding business cycles, my findings suggest that short-term fluctuations in mortality are not very costly. Secular improvements in life expectancy and gains against static health inequalities appear to be much more important.

Social Science & Medicine 67(12):2051-8, December 2008.

There is renewed interest in understanding how fluctuations in mortality or health are related to fluctuations in economic conditions. The traditional perspective that economic recessions lower health and raise mortality has been challenged by recent findings that reveal mortality is actually procyclical. The epidemiology of the phenomenon --- traffic accidents, cardiovascular disease, and smoking and drinking --- suggests that socioeconomically vulnerable populations might be disproportionately at risk of "working themselves to death" during periods of heightened economic activity. In this paper, I examine mortality by individual characteristic during the 1980s and 1990s using the U.S. National Longitudinal Mortality Study. I find scant evidence that disadvantaged groups are significantly more exposed to procyclical mortality. Rather, working-age men with more education appear to bear a heavier burden, while those with little education experience countercyclical mortality.

Population and Development Review 34(S): 103-125, March 2008.

Traditional theories posit a line of causality running from tech- nological change into mortality decline. But that relationship could be self-sustaining if longer life spans contribute to increased knowledge production. The productivity of scientists appears to decline only gradually with age, implying that a reduction in adult mortality will stimulate knowledge production by leaving more productive scientists alive. The growth in empirical scientific knowledge in the 17th century preceded any widespread mortality declines, which occurred in the 19th century. But the vital statistics of members of the Royal Society of London indicate that the life spans of British scientists were increasing at the same time that scientific knowledge began to grow rapidly, the latter fostered by the Society itself. How significantly the emergence of these early health inequalities contributed to the massive increases in population health following industrialization is unclear and deserving of further inquiry.

National Tax Journal 57(1): 45-65, March 2004.

Changes in the monthly pattern of Earned Income Tax Credit disbursements over the past decade identify a large macroeconomic con- sumption response from EITC checks. This paper recovers a large and significant MPC out of EITC disbursements based on a comprehensive array of macroeconomic data and econometric specifications. Point estimates of the contemporaneous consumption response average 0.7 and do not attribute a disproportionate share of consumption to durable goods. Results are consistent with other empirical findings that consumption is excessively sensitive to income, and they suggest that the EITC is a much more effective fiscal stimulus tool than broad-based tax refunds.

with Ronald D. Lee and Shripad Tuljapurkar

in Shripad Tuljapurkar, Naohiro Ogawa, and Anne H. Gauthier, eds., Riding the Age Waves - Vol. 3: Aging in Advanced Industrial States, New York: Springer, 79-100, 2010.

Recent decades have seen the emergence of massive public sector transfer programs in industrial nations. Because many transfers are age related, the population age distribution is a powerful influence on government budgets, and aging will be very costly. We construct stochastic projections of the budgets for the federal and state/local governments, disaggregated by program. These are driven by stochastic population projections and by stochastic projections of productivity growth and real interest rates. The demography influences budgetary outcomes through the age specificity of seven categories of tax payment and 28 government spending programs, as well as through public goods expenditures, debt service, and provision of congestible services. Forecasts of government deficits and debt under current tax and benefit trajectories make it clear that adjustments will have to be made in the future to avoid implausibly high and unsustainable debt levels. Subsequent forecasts are conditional on an upper bound to the federal debt/GDP ratio of 0.80. There is a very slight chance (2.5%) that the overall tax burden will barely rise, but most likely it will have to increase by 62% (from 24% in 1994 to 38% of GDP in 2070), and there is a slight chance that taxes would rise by 120% (to 53%). We have not yet explored adjustment through reduction of benefits. The expected GDP shares of child related expenditures and age neutral expenditures are both flat up to 2070. The expected share of old age expenditures, however, rises from 8.5% of GDP in 1994 to 22.5% in 2070. We find that there is a strong negative correlation between rates of expenditure on child-oriented programs and on elder-oriented programs, as one would expect given the importance of fertility for both outcomes. Thus focusing exclusively on the tax burden resulting from population aging could somewhat exaggerate the increases needed in the future. We also find that the rising cost of Social Security benefits (OASDI) accounts for only 28% of the rise in old age expenditures; fixing Social Security will not in itself fix the federal budget. The rising costs of health care will contribute even more, and must be addressed as well.

National Transfer Accounts Working Paper WP10-01, February 2010.

Government finances depend on a vast array of variables, but there is a clear hierarchy among them, and forecasting requires a parsimonious treatment of complicated budget rules and behavior. The number of people paying taxes or receiving benefits and the intensity by which they do so are the most fundamental of these inputs. In this paper, we describe our approach to forecasting federal and state and local government finances using forecasts of the U.S. population by age and sex and measures of our assumptions about the intensity of program utilization by age and sex, or “age profiles” in common usage.

with Ronald D. Lee and Timothy Miller

a special report of the California Policy Research Center, 2003

This report provides a composite demographic profile of California's aging population over the next 50 years. Estimates of future demographic characteristics utilize a "probabilistic" projection method that simulates 10,000 possible futures, in an effort to overcome the built-in biases of most population projections and inconsistencies in standard demographic assessments of "most likely" futures. This method enables us to explore future needs while recognizing the degree of uncertainty associated with our projections.

with Ronald D. Lee

in J.M. Poterba, ed., Tax Policy and the Economy 16: 141-181, 2002.

Population aging, accelerating as the Baby Boom generations age, will have important fiscal consequences because expenditures on social security, Medicare, and institutional Medicaid make up more than a third of the federal budget. However, the projected fiscal pressures are far in the future, and long-term projections are very unreliable. Our analysis here has two goals: to examine the fiscal impact of population aging, and to do this in a probabilistic setting. We find that the old age dependency ratio is virtually certain to rise by more than 50% through the 2030s, and will probably continue to increase after 2050, possibly by a great deal. Under current program structures, population aging would be virtually certain to increase the costliness of Federal programs as a share of GDP by 35 percent (+/- 2 percent) by the 2030s, and by 60 percent (+/- 15 percent) in the second half of the century. We project Federal expenditures (excluding interest payments and pre-funded programs) to rise from 16 percent of GDP in 2000 to 30 percent in 2075, almost doubling, while state and local expenditures rise only modestly relative to GDP. Almost all of this increase is for programs going primarily to the elderly, which rise from 8 percent of GDP in 1999 to 21 percent of GDP in 2075, due mainly to costs of health care for the elderly, with pensions a distant second. We expect that governments will respond to these aging-induced cost changes by altering program structures, so that these conditional projections will not be realized. Looking at social security, we find that raising the payroll tax rate by 1.89 percent would have relatively little effect on the probabilities of early exhaustion, raising the 2.5 percent bound for the exhaustion date from 2024 to 2036, but raising the median date of exhaustion from 2036 to 2070, and with a 55 percent chance of insolvency within the 75 year horizon. Looking at Medicare, which now costs 2.2 percent of GDP, we project a median share in 2075 of 11 percent, five times as great, with a 95 percent probability interval at 5 percent to 26 percent of GDP. Thus there is a 97.5 percent chance that the ratio will at least double, and a 2.5 percent chance that it will increase at least twelve-fold. Although the future is highly uncertain in many respects, unforeseen demographic or economic change will almost certainly not avert the long-run fiscal crunch. Changing demographic realities will require some combination of substantial tax increases or substantial benefit cuts, or other forms of restructuring.

with Ronald D. Lee

in J.S. Little and R.K. Triest, eds., Seismic Shifts: The Economic Impact of Demographic Change, Federal Reserve Bank of Boston Conference Series No. 46, 2001.

Population aging is a natural and inevitable stage in the process of the demographic transition, a transition that has been unfolding in the US over the past 200 years. Because the US has, and may continue to have, fertility close to the replacement level of two children per woman, as compared to the substantially lower fertility throughout much of the industrial world, population aging in the US is expected to be relatively mild. Nonetheless, the ratio of those aged 65 and above to those aged 20 to 64 is expected to double, and could rise higher, over the next 75 years. Longer life and smaller families are fiscally costly. Major changes in taxes, benefits, or program funding structures are necessary. Policy makers and the public must be educated to these new realities, and difficult decisions must be made.

with Jennifer Roff

Unpublished working paper, January 2013.

This paper employs recently developed marital matching models to examine empirically the role played by marital sorting in observed measures of marital production. Using the US Collaborative Perinatal Project (CPP), a large-scale study from the 1960s, we find that marital surplus is strongly correlated with indexes of child quality, as measured by cognitive test scores, and with the durability of the marital union. At ages beyond infancy, correlations between neurocognitive outcomes and marital surplus are independent of the parental characteristics that generate the match, suggesting that they may represent effects of the match itself. They are also robust to controlling for household income and number of siblings. High marital surplus is associated with assortative mating on education and age, suggesting complementarity in parental inputs in child production and a joint effect of parental education and age on child outcomes that exceeds the linear sum of the parts. To the extent that marital surplus can be considered a proxy for the subjective well-being of the couple, our results suggest that parental happiness is an important input for child quality above and beyond its indirect effects on marital stability and earnings.

PLoS ONE 5(9): e12157, 1-9, September 2010.

Recent findings suggest advanced paternal age may be associated with impaired child outcomes, in particular neurocognitive skills. Such patterns are worrisome given relatively universal trends in advanced countries toward delayed nuptiality and fertility. But nature and nurture are both important for child outcomes, and it is important to control for both when drawing inferences about either pathway. We examined cross-sectional patterns in six developmental outcome measures among children in the US Collaborative Perinatal Project (n = 31,346). Many of these outcomes at 8 mo, 4 y, and 7 y of age (Bayley scales, Stanford Binet Intelligence Scale, Graham-Ernhart Block Sort Test, Wechsler Intelligence Scale for Children, Wide Range Achievement Test) are negatively correlated with paternal age when important family characteristics such as maternal education and number of siblings are not included as covariates. But controlling for family characteristics in general and mother's education in particular renders the effect of paternal age statistically insignificant for most developmental measures. Assortative mating produces interesting relationships between maternal and paternal characteristics that can inject spurious correlation into observational studies via omitted variable bias. Controlling for both nature and nurture reveals little residual evidence of a link between child neurocognitive outcomes and paternal age in these data. Results suggest that benefits associated with the upward trend in maternal education may offset any negative effects of advancing paternal age.

NBER Working Paper 18227, July 2012, updated November 2014. Forthcoming, Journal of Human Capital .

Recent engagements in Iraq (OIF) and Afghanistan (OEF) raise questions about impacts on service members of overseas deployment. The 2010 National Survey of Veterans asked a broad cross section of veterans about deployment to OEF/OIF and combat exposure. Analysis of these data suggests that impacts of deployment and combat on the current socioeconomic well-being of OEF/OIF veterans may be small, but combat appears to reduce self-reported health and other nonpecuniary indicators. Among older cohorts, with clearer sorting into treatment and control groups, patterns are similar, consistent with a system that compensates for lost earnings but not necessarily other trauma.

with Alair MacLean

Unpublished working abstract , September 2014. Accepted for presentation at the 2015 Annual Meeting of the Population Association of America.

Veterans of the Vietnam War, the last large-scale U.S. engagement prior to the All-Volunteer Force era, have now largely reached retirement age, with record high rates of service-connected disability compared with veterans of earlier wars. Understanding the determinants of healthy aging among this cohort is important for assessing current and future needs of veterans, especially now that policies and events have generated another large wartime cohort, and for gaining insights into health dynamics over the life cycle. In this paper, we compare objective and subjective metrics of health across male veterans and nonveterans in a population- based panel survey of Americans over age 50, the Health and Retirement Study, which recently began collecting biomarkers. We revisit earlier results that suggest subjective self-reports by veterans may be overstated, a "soldiering on" effect, relative to objective measures of health. Our findings speak to the lifelong influences of earlier-life conditions and of the lingering challenges posed by exposure to combat.

Journal of Public Economics 113: 54-66, May 2014.

Previously NBER Working Paper 16108, June 2010.

Military spending, fatalities, and the destruction of capital, all of which are immediately felt and are often large, are the most overt costs of war. They are also relatively short-lived. But the costs of war borne by combatants and their caretakers, which includes families, communities, and the modern welfare state, tend instead to be lifelong. In this paper I show that a significant component of the budgetary costs associated with U.S. wars is long-lived. One third to one half of the total present value of historical war costs are benefits distributed over the remaining life spans of veterans and their dependents. Even thirty years after the end of hostilities, typically half of all benefits remain to be paid. Estimates of the costs of injuries and deaths suggest that the private burden of war borne by survivors, namely the uncompensated costs of service-related injuries, are also large and long-lived.

with James Geiling and Joseph M. Rosen

Military Medicine 177(11): 1235-1244, November 2012.

This paper focuses on the long-term challenges that families, communities and governments will face in caring for veterans of the wars in Iraq (OIF) and Afghanistan (OEF). It briefly reviews the medical impacts of war --- examining new patterns of injury and illness in the current conflicts for which the government, military, Veterans Affairs (VA) and related services will need to provide medical and psychological care to wounded servicemen and women, especially as they reach middle age in 2035. This paper specifically discusses the future, long-term burdens of veterans' current medical maladies --- a topic not often discussed --- and suggests proactive health programs that may help contain costs and ensure the provision of quality care. Major medical conditions among OEF/OIF veterans are Post-Traumatic Stress Disorder (PTSD), Traumatic Brain Injury (TBI), amputations/polytrauma, and the complications that develop either from or together with these disorders. There are short-term medical expenses that include transportation, surgery, and hospitalization; and long-term medical costs, which include mental health care, rehabilitation, and disability. Communities or families adapting to care for wounded veterans will absorb costs that include private medical care, informal at-home care, lost job productivity, and the effects of family stress. Developing appropriate mitigating interventions for our wounded service personnel in the near term --- such as smoking cessation, alcohol-abuse counseling, weight control, resilience programs, and physical therapy --- will not only better serve these veterans, but may also reduce the potentially enormous cost burden of their future health care.

with Alair MacLean

Unpublished working paper prepared for session 139 of the 2011 Annual Meeting of the Population Association of America, April 2011.

Military service has traditionally been the domain of healthy, robust males, but service can also reflect risk preference and socioeconomic status. Service also raises the probability of exposure to violence through combat, a significant stressor, and it may represent other types of treatments as well, both positive and negative. We might expect to find an ambiguous relationship between military service and later-life health, and several recent studies support this. In this paper, we explore the relationship between combat exposure and health past age 50 in the Health and Retirement Study, a rich longitudinal panel including many male veterans that now asks about combat exposure in its core survey. Using regression analysis and an instrumental variables approach, we show that combat exposure harms mental health and emotional well-being and raises a biomarker of stress at older ages, but it appears often to have negligible effects on a wide array of physical health metrics.

with Alair MacLean

Armed Forces & Society 36(5): 765-785, October 2010.

The following paper tests the hypothesis that veterans have better health if they were officers when they were in the US military than if they served in the enlisted ranks. It examines this hypothesis by presenting results from logistic regressions that are based on four surveys: the National Survey of Veterans, the Survey of Retired Military, the Panel Study of Income Dynamics, and the Wisconsin Longitudinal Study. In all four of these surveys, the evidence is consistent with the hypothesis that military rank is associated with health, particularly among veterans who served longer. It also suggests that the health gradient by rank is independent of similar gradients by education and income, as well as health differences by race. These findings indicate that health may be influenced not just be differences in civilian society, but also by those in the military.

NBER Working Paper 16163, July 2010.

As of this writing, the wars in Iraq and Afghanistan are in their eighth and tenth years, having accrued nearly a trillion dollars in direct military costs. I review the history of cost forecasts for these ongoing engagements, highlighting the differences across them in scope and accuracy, assessing the methods and practice of cost forecasting, and exploring the implications of the war costs themselves. Besides the mostly unanticipated length and breadth of the military conflicts themselves, a related and equally important component of costs is the life cycle of costs associated with caring for veterans. The forecasts we have of such costs imply high levels of public spending per veteran and very high levels of costs associated with pain and suffering per veteran, as high as 10-25 percent of lifetime wealth. I also discuss the methods and motivations associated with war cost forecasts by comparing them with other types of aggregate forecasts, which are prone to similar types of errors. The history of war cost forecasts calls for a concerted effort to improve them by increasing their frequency and transparency.

Social Science & Medicine 67(11): 1657-1668, December 2008.

I explore trends in mortality among U.S. military retirees using a new dataset of payroll records that include pay grade. Trends in mortality by pay grade reveal that health inequalities steadily widened between 1974 and 2004. Additive differentials in mortality rates remained stable, but since mortality declined exponentially, by a factor of about one third, proportional differentials in mortality and thus additive differentials in life expectancy have widened. The advantage in life expectancy enjoyed by retired officers grew roughly from 3 to 4 years. The sources of these trends remain unclear and are beyond the ability of the data to inform, but the results bear implications for trends in inequality and for policy.

Unpublished working paper, June 2013.

The building at 2232 Piedmont Avenue in Berkeley, California, which has housed the Department of Demography since 1988, was designed by renowned architect Julia Morgan and occupied by one family prior to its sale to the University in 1958. Like other genealogies, the family history of the Kelloggs is traceable through ancestry.com via ProQuest and other public databases. In particular, U.S. census records from 1940 and earlier provide snapshots of the family that occupied 2232 Piedmont from its construction in 1909. An overview of these records offers some ethnographic insights into the earliest occupants of the building, their way of life, and into the building's configuration. It also provides a primer for the building's current occupants on using tools familiar to genealogists for demographic research. Finally, these records are interesting because of what they reveal about the objective quality in this particular instance of a type of historical population data that demographers frequently use. The substantial age difference between the head of household and spouse, 14 years and 7 months, was inaccurately recorded in 4 of 5 post-marital decennial censuses, with inconsistencies affecting the records of both spouses.

Exhibits are available upon request. They are large image files drawn from the ancestry.com database that are apparently public but also potentially sensitive.

Unpublished working paper, June 2008.

Research on socioeconomic inequalities in health often examines inequality in mortality rates or life expectancy. The choice of inequality measure is important because it should be consistent with the dominant temporal trend in the characteristic of interest. The author argues that proportional measures of inequality in mortality rates and additive measures of inequality in life expectancy are preferable to the alternatives because the dominant temporal pattern in mortality has been one of proportional decline in mortality rates concomitant with additive increase in life expectancy. Proportional differentials in mortality rates are mathematically consistent with additive differentials in life expectancy, and both are stable measures of inequality given the temporal trends.

Unpublished working paper prepared for session 74 of the 2006 Annual Meeting of the Population Association of America, March 2006.

Demographers are well aware of the vast decreases in mortality among industrialized countries during the past century, and they generally expect mortality decline to continue into the future. But how do individuals perceive the chances for their future survival? Survivorship expectations should influence a wide array of economic decisions and thus well-being, from the amount of education to obtain to the amount of money to save for retirement. Previous research exploring survivorship expectations in panel data has shown that they are correlated with current health status, behaviors, and with future mortality, and that they approximate life table quantities. In this paper, I take a closer look at the accuracy of survivorship expectations along two main dimensions. I examine systematic biases in self-reported survivorship expectations by racial group, and I assess the updating of survivorship expectations over time within a cohort and between cohorts. I find an interesting mix of irrational optimism and pessimism among groups identified by race, sex, and age, but I also find evidence that over time, individuals appear to understand that mortality decline is occurring.