Discussion: (0 comments)
There are no comments available.
| Journal of Law, Medicine and Ethics
View related content: Health Care
The definitive version of this article is available at www.blackwell-synergy.com.
I was initially assigned the working title, “Pursuing Equality in Health Care for the Elderly Is Futile.” I prefer to think of that particular dead end of health policy as one of listening to the wrong music for too long. Hence, this article reprises and revises the title song of the early 1980s movie, Urban Cowboy, but with Johnny Lee’s original lyrics adapted as “Looking for better health [rather than either "love" or "love of equality"] in all the wrong places.”1 The better goal is to achieve more progress in improving health for more people, including (but not limited to) the elderly. It need not be as futile as the pursuit of the elusive abstraction of “equality” for all – but only if we first move away from a path-dependent approach of recent times that remains too narrowly focused on statistical disparities in health care services received by particular groups.
A Thin Evidence Base
In examining this topic, one finds that the actual evidence base for measures of health inequality among the elderly in the United States remains rather thin. We need to reconsider what, and how, we measure in searching for apparent differences in health across and among different parts of the population. Are we even asking the right questions? Once we begin to do so, we may discover a much more complex set of causes, correlations, and limiting factors facing both researchers and policymakers. They suggest that we proceed with greater caution and humility in setting more feasible priorities and targets of intervention. We should refrain from continuing to search for the lost keys to better health only where the political light appears to be brightest – under a well-funded lamp post called “health disparities” – rather than where those keys actually might be located.
“The main emphasis of the annual report is on health disparities related to racial and ethnic minorities, and low-income groups.” – Thomas Miller
The most common starting point for a purportedly comprehensive and evidence-based review of the health inequality issue in the United States is supposed to be the annual National Healthcare Disparities Report (NHDR), produced by the Agency for Healthcare Research and Quality.2 The most recent3 6th annual report, for 2008, defines health care disparities as differences or gaps in care experienced by one population compared with another population.”4 Its broad mission is to measure differences, and changes over time, in health care quality and access to care for various health conditions in different populations. Note that the resulting findings not-too-subtly appear to sound more intentional and unfair when they are termed “disparities” than when they are called “differences.” The main emphasis of the annual report is on health disparities related to racial and ethnic minorities, and low-income groups. However, one of the seven “priority” populations it addresses involves older adults – age 65 and over. (Other priority populations include women, children, residents of rural areas, and individuals with disabilities or special health care needs).5
The highlights of the latest NHDR include the finding that health disparities persist for all populations. Of course, that appears to be somewhat inevitable, because the report starts from the contestable premise that all population groups should receive an “equally high” quality of care. Other published research, along with a similar annual National Healthcare Quality Report, has reported that Americans too often fail to receive the recommended care they need or receive care that causes harm.6 However, the NHDR tries to focus on those groups that receive even worse care than others (either between different population cohorts, or within them), or perhaps just do not get enough of it, despite its poor quality.
Although the NHDR uses 220 measures to assess health care quality, the 2008 report focuses on 45 core measures. Those core measures generally remain limited to assessing health care inputs and processes, rather than broader health outcomes. Overall, the report’s list of core measures for 2008 includes 27 for processes and only 18 for health outcomes.7 The receipts of medical care are easier to measure than its results, even though patients (and hopefully most providers) should care much more about the latter.
In general, the report’s methodology – while acknowledging that many factors contribute to disparities in quality and access. including cultural attitudes and health literacy – also places its greatest emphasis on increasing access to care through increasing the portion of the population with health insurance. It measures health care disparities primarily by race, ethnicity, and socioeconomic variables (emphasizing income variables much more than education factors, for the latter). The NHDR approach appears either to lack the data sources, or the interest, to pursue very much bivariate or multivariate analyses to control for multiple factors, which would pinpoint better the extent to which each one in particular affects a health outcome.
For older adults (above age 65), the sixth annual NHDR discusses just four measures – involving influenza vaccination, vision screening, delayed care due to cost, and health literacy, respectively. It notes progress in increasing the percentage of Medicare beneficiaries with an influenza vaccination from 1998 to 2004 (up from 68.7% to 71.7%) but points to a significant gap based on income (ranging from 61.3% for poor Medicare beneficiaries to 77.6% for high-income ones). For vision screening, the gap between blacks and whites decreased from 1998 to 2004 with no change in the gap between poor and high-income Medicare beneficiaries. There also were no significant changes based on gaps in income for delayed care due to cost, from 1998 to 2005, although Hispanic beneficiaries were more likely than non-Hispanic whites to delay care for this reason (7.8% versus 4.4%). Finally, health literacy was a broad problem for adults age 65 and over, and it was particularly poor for adults over age 75 (more than two-thirds of those Medicare beneficiaries had below basic or basic health literacy).8
The broader health picture for Americans over age 65 is that they represent a growing share of the population and their life expectancy has increased significantly. They naturally face greater health care concerns than do younger populations, but they also experience lower rates of poverty. Based on the limited evidence in the latest NHDR for health disparities among the elderly, one might be tempted to ask, “Is that all there is?”
The Role of Medicare in Reducing Health Inequality
One initial presumption is that the enactment of Medicare coverage in 1965 and its subsequent expansion must have improved the health of Americans over age 65, and reduced health disparities between the young and the old. However, health researchers Amy Finkelstein and Robin McKnight actually found that the establishment of universal health insurance for the elderly had “no discernible impact” on their mortality during the first 10 years of the Medicare program.9 Although Medicare played essentially no role in the dramatic decline in mortality rates for the elderly that began in the late 1960s, it nevertheless did help reduce the elderly’s financial exposure to out-of-pocket medical spending risk. The Medicare-induced increase in health care consumption was ineffective at least in regard to mortality, concluded Finkelstein and McKnight, because lack of legal access to care – not lack of insurance – was the primary problem that was corrected with regard to mortality post-1965. The broader effects of desegregation and new civil rights enforcement on the overall black population included improved access to hospital care for individuals with life-threatening, treatable conditions.10
Medicare’s expansion, however, had other major effects within the overall U.S. health care system. In related research, Finkelstein concluded that Medicare’s growth, along with the overall spread of health insurance between 1950 and 1990, may explain as much as half of the six-fold increase in real per capita health spending during those four decades.11 Medicare stimulated entry of new hospitals and the adoption of new medical technologies. The market-wide changes in health insurance it helped create may have fundamentally altered the character of medical care, both for individuals who experienced a change in coverage and for those who did not.12
Other potentially more positive evidence of Medicare’s effects on the health of the elderly comes from the work of University of California-Berkeley economist David Card and several colleagues. They suggest that Medicare eligibility at age 65 reduces the death rate of a relatively sick population (people admitted to hospitals through the emergency room for “non-deferrable” conditions) by 20 percent, and that this mortality improvement persists for at least two years following the initial hospital admission.13 However, those findings reflect the relatively greater generosity of services (or more timely delivery of them) for patients covered by Medicare and supplemental private insurance relative to insurance benefits held by people just under age 65. The study also did not examine whether the total costs associated with the overall Medicare program (which include large increases in the use of services by other “healthier” patients who experience small, or no, mortality effects) were justified by the total gains in health for this particular sub-population of beneficiaries.14
Health Differences among the Elderly Based on Income
A different area for investigating health inequality among the elderly involves measures of income-related disparities within that age cohort. In other words, how progressive is the Medicare program’s structure of benefits and financing for its beneficiaries. At first glance, as an age-based entitlement program, Medicare appears to offer the same core health benefits to almost all Americans age 65 and above, regardless of income. On the other side of the beneficiary ledger, however, Medicare relies on a mixed combination of revenue sources – based on payroll taxes and income taxes. This overall financing scheme has produced less consistent net distributional effects (benefits received minus taxes paid) over time than might be first assumed.
One earlier line of research, begun by Mark McClellan and Jonathan Skinner, suggested that, as of 1990, Medicare produced net transfers of funds from low-income beneficiaries to higher-income ones.15 Up to that point, Medicare relied on relatively regressive payroll-tax financing – with a capped amount of wages subject to its flat rate. Higher-income beneficiaries received greater lifetime benefits due to their longer survival times. McClellan and Skinner concluded that Medicare was regressive if measured as a cash-transfer mechanism, but acknowledged that if the “insurance value” of Medicare benefits was taken into account, then the program might instead produce “faint redistribution” from the highest-income beneficiaries to the lowest-income ones.16
“McClellan and Skinner concluded that net Medicare benefits remain slightly higher for those in the highest-income households than those in lower-income groups.” – Thomas Miller
A subsequent paper by McClellan, Skinner, and Julia Lee identified a dramatic shift in the pattern of Medicare spending between 1990 and 1995, with increased redistribution toward the lowest-income neighborhoods, where Medicare per-capita spending grew much more rapidly.17 Although the cap on wage-based earnings subject to the Medicare share of the payroll tax was removed in the early 1990s, this shift toward more progressive net transfers was primarily due to some unusual short-term distortions in more intensive Medicare services (such as home health care spending) delivered in lower-income neighborhoods. McClellan, Skinner, and Lee described those services as of more questionable value to the beneficiaries receiving them.
In a final, updated version of this work published in 2006, McClellan and Skinner moved back toward their original findings, concluding that net Medicare benefits (lifetime expenditures versus taxes paid) remained slightly higher for those in the highest-income households than those in lower-income groups.18 However, this line of analysis remained limited in several respects. It relied on income data by zip code, rather than per Medicare beneficiary. It was also hampered by the researchers’ inability to measure easily the overall value of the health consequences produced by additional health care services received due to more Medicare spending, even while referencing evidence that suggested the low marginal value of many intensive Medicare services. In fact, the 1999 study by McClellan, Skinner, and Lee had noted that the large shift in Medicare resources toward people in lower-income neighborhoods failed to improve their survival rates and might have even slightly increased disparities in mortality rates.
Nevertheless, the McClellan and Skinner line of studies ultimately decided to rely on the “presumed value” of income transfers to low-income recipients in the form of generous, community-rated insurance benefits, despite evidence that those transfers also produced inefficient overconsumption of medical care by those elderly beneficiaries and more insurance than they wanted.19
Jay Bhattacharya and Darius Lakdawalla have offered a different approach to measuring the relative progressivity of Medicare. They found that the financial returns to Medicare are actually much higher for poorer groups in the population and concluded that Medicare is a highly progressive public program.20 Their study used educational attainment as a better measure of permanent income and socioeconomic status (SES), arguing that this would correct for the aggregation bias in the McClellan/Skinner measures of SES based only on geography.21 Bhattacharya and Lakdawalla account for the role of geographic mobility by the elderly, noting that those Medicare beneficiaries who move to richer ZIP codes (with presumably higher quality medical facilities) tend to increase their total medical spending, while those moving to poorer areas lower their spending level.22 They con-clouded that, at any given age, Medicare spends far more on the poor (less educated) than it does on the rich (more educated) and found that the advantage of the poor in receipt of Medicare benefits even overcame their higher death rates. This is partly related to differences in health status, because less educated people are sicker and therefore cost Medicare more.23
However, this particular finding is limited, because other adjustments for education-related longevity and Medicare-benefit growth on a lifetime basis (rather than measured at a single point in time) erode some of Medicare’s progressivity, particularly for beneficiaries living beyond age 75. Bhattacharya and Lakdawalla also note in passing that they found a positive gradient in privately financed medical expenditures (more education leads to greater spending) once one controls for health status.24 More significantly, their analysis remains wedded to measuring only differences in the dollar amounts of Medicare benefits received, instead of more significant disparities in the health outcomes that such spending on health care services may only partly help produce.
Asking a Better Question Delivers a Better Answer
A better way to measure disparities with the Medicare beneficiary population is to ask the right question. Jonathan Skinner and Weiping Zhou drew distinctions between inequality in health spending and inequality in health outcomes.25 They found that when inequality is measured by differences in Medicare spending, health care for the elderly became more equitable during the past several decades. However, this relative growth in health spending directed at low-income elderly people did not translate into relative improvement either in survival or rates of effective care.26 When Skinner and Zhou considered how several different cohorts of Medicare beneficiaries fared in survival rates between 1982 and 1992, they found that the highest income groups gained the most, both in percentage and absolute terms. Whereas the relative levels of health spending by different income groups depend on preferences, health status, and prices, their health outcomes are more strongly influenced by health behavior, diet, and past life-course events (such as past illness) that extend beyond the health care system alone. Given the association between those behavioral factors and income and SES, Skinner and Zhou emphasized that inequalities in health can reflect wider inequalities in society.27 Noting the problem in trying to infer that spending more Medicare money on lower-income groups necessarily improves their health, Skinner and Zhou suggest that policymakers might focus more effectively on what matters most – improving the delivery of “effective” care to lower-income patients.28
Even Universal Insurance Coverage under a National Health Program Does Not Ensure More Equal Health Outcomes
More cursory examinations of how fairly Medicare’s health care services are distributed among the elderly based on income may assume that essentially the same set of benefits are delivered in the same way throughout the United States. After all, a national program providing near-universal coverage to seniors age 65 and over is supposed to reduce the role of money in accessing care. However, several decades of analysis of geographic variation in health spending and patterns of medical practice suggest how any such pursuit of equality remains elusive. For example, researchers involved in producing the Dartmouth Atlas of Health Care have estimated that among groups of Medicare beneficiaries who are otherwise similar, those living in high-spending areas receive approximately 60 percent more in services than do those living in low-spending areas. Although the prices of health care services and severity of illness are important factors in explaining geographic variation in health care spending, they combine to account for, at most, less than half (and possibly much less than half ) of the geographic variation in spending.29 Income and the preferences of individuals for specific types of care also appear to explain little of the variation in Medicare spending for the elderly. Some of the variation in medical practice probably is attributable to regional differences in medical resources for so-called supply sensitive services (e.g., physician visits, specialist consultations, diagnostic tests, and hospitalizations) and the propensity to take advantage of the financial incentives provided by Medicare or other payers in developing and using those resources, according to a 2008 Congressional Budget office review of the existing literature and evidence.30
With some regions more prone to adopt low-cost, highly effective patterns of care and others more likely to adopt high-cost patterns of care and to deliver treatments that provide little benefit or are even harmful, any possible residual evidence of health inequality based on the income characteristics of beneficiaries remains hard to find. By one older estimate, the Dartmouth Atlas researchers calculated that Medicare spending would fall by 29 percent if spending in medium-and high-spending regions were the same as in their benchmark regions, defined as those with spending in the lowest decile.31
Moreover, efforts to find “unfair” differences among beneficiaries in regard to the health services they receive and/or health outcomes they experience within a purportedly “uniform” national program are further complicated by the fact that various types of supplemental insurance for the elderly produce additional differences. Medicaid coverage for dual-eligible elderly Americans, additional supplemental benefits for certain lower-income groups, and private Medigap coverage for many other beneficiaries means that even enrollees within the traditional Medicare, let alone those in more generous private Medicare Advantage plans (which experience surprisingly higher enrollment by lower-income African-American beneficiaries in many urban areas), create a much more complex, layered pattern of benefits relative to income than Medicare’s core benefit structure might suggest.
Even when one examines national health systems in other countries, where promised health benefits appear to be more uniform, substantial differences in the health outcomes they seemingly produce remain. The Whitehall studies are the most notable set of long-term examinations of socioeconomic differences in physical and mental illness and mortality. Led by chief investigator Sir Michael Marmot, the initial Whitehall I Study was conducted over 10 years, beginning in 1967. It examined social determinants of health in terms of cardio respiratory disease prevalence and mortality rates among British male civil servants between ages 20 and 64. The Whitehall I study demonstrated that there was a social gradient in mortality rates from a range of causes, running from the bottom to the top of the grade levels of employment within the British civil service. Although the lower the grade, the higher the risk of death, the gradient of health involved more than just a clear dividing line between poor health for the disadvantaged and good health for everyone else.32
“The social gradient in health was shaped by such factors as one’s degree of social participation and sense of control over their life, as well as their work climate and early life experiences.” – Thomas Miller
The effects of social and occupational influences on health and illness (morbidity) were investigated further in the next round of Whitehall II studies, involving a variety of diseases affecting both men and women in civil service jobs. Whitehall researchers found that work, stress, and health were inter-related. The social gradient in health was shaped by such factors as one’s degree of social participation and sense of control over their life, as well as their work climate and early life experiences.33
The important takeaway point here regarding sources of health inequality is that the Whitehall studies demonstrated that other inequalities in the broader society of the United States shaped the social determinants of health, and they were stronger than any potentially “equalizing” factors stemming from a single-payer National Health Service. Universal provision of health care was no guarantee of universal equality in the outcomes it produced.
One overly ambitious, if not utopian, policy response that might flow from examination of the many factors shaping differences in health approach is the one recently recommended by the Commission on Social Determinants of Health, set up by the World Health Organization in 2005 and chaired by Sir Marmot. It urged that “[a]ction on the social determinants of health must involve the whole of government, civil society and local communities, business, global fora, and international agencies. Policies and programs must embrace all the key sectors of society not just the health sector.”34 Although the Commission may have suffered from an overdose of mission creep in insisting on addressing inequities in “the way society is organized,” it’s more down-to-earth recommendations for starting to improve inequitable conditions of daily living centered on early childhood development and education.35
Lesson: Measure What Matters
The first broad lesson from the evidence above is that diagnosing the real causes of health inequalities and finding effective ways to reduce them (whether just limited to the elderly or to other categories of individuals within the broader population) begins with better measurement that focuses on outputs instead of inputs, and health outcomes rather than health care processes. It needs to go well beyond surface indicators of relative access to different quantities and qualities of health services. More effective measurement should capture the influence on health outcomes of a more complex mix of possible factors, such as health behavior, diet, and life-course events. The public policy objective should be to discover the reasons for differences in health, not just stop at detecting and highlighting apparent statistical differences at the group level. This alternative path would focus more on developing effective strategies to reduce health inequalities and improve health at the individual level.
Lesson: Examine the Real Causes of Health Differences
Once one begins to ask better questions regarding the true determinants of different health outcomes, different policy answers appear. At a basic level, the work of Michael McGinnis regarding the causes of early deaths (avoidable mortality) in the United States indicates that some 40 percent of them are caused by behavioral patterns that could be modified by preventive interventions, but only a much smaller proportion – perhaps 10-15 percent – could be avoided by better availability or quality of health care.36
The impacts of other domains on early deaths in the U.S. include 30 percent due to genetic predispositions, 15 percent linked to social circumstances, and 5 percent related to environmental exposures. At a minimum, this evidence suggests that if we retargeted our public resource commitments toward dealing more effectively and earlier in life with the nonmedical determinants of population health, we not only might rely somewhat less heavily on conventional medical interventions yet deliver better health outcomes; we also might even make more progress in improving the health of more vulnerable groups such as the elderly.
Consumption of health care services provides only one input into the health-production function, in the “health capital” model of Michael Grossman.37 His extensive body of research reveals that other non-medical factors – such as exercise, nutrition, health-related behaviors, and social norms – account for much more of the differences in predicting health outcomes among individuals and groups. The “upstream” determinants of future health – which begin to shape health outcomes even before one engages the health services sector more extensively and continue to operate as powerful outside forces afterward – include one’s education level, culture, family, geography, neighborhood, health literacy, and self-care capability.
Let’s focus here primarily on education, because Grossman concludes that one’s education level is the most important correlate of good health, even more powerful than other socioeconomic variables like income or occupation. Related research by James Smith finds that additional schooling more strongly predicts disease onset than does either the income component of socioeconomic status (SES) or one’s health insurance status.38 Adreana Lleras-Muney estimates that one additional year of education increases life expectancy at age 35 by 1.7 years.39 In their examination of the determinants of mortality, David Cutler, Angus Deaton, and Lleras-Muney conclude that, although the relationship between SES and health works to various degrees in both directions, the effects of the “education” component of SES on health are more consistent than those of the “income” component.40 Much of the link between income and health is the result of the latter causing the former, rather than the reverse.
Cutler, Deaton, and Lleras-Muney observe that greater levels of education are likely to provide general human capital that can be used to maintain and improve health in a wide range of circumstances.41
Education makes it easier (or cheaper) for people to obtain and process information about the causes and consequences of one’s health. Education may change a person’s time preferences in a manner that encourages deferred gratification and future-oriented behavior, which favors investments in health relative to consumption of other commodities.42 Grossman suggests that more educated people not only use the health care system more effectively; they also demand more from it.43 More broadly, those individuals are likely to feel more in control of their lives and more engaged in patient self-management, which is related to better problem-solving skills, richer social networks, and healthier life style norms.
Education differentials may best explain growing gaps in life expectancy among the elderly, according to Ellen Meara, Seth Richards, and David Cutler. They observe that all recent gains in life expectancy at age twenty-five have occurred among better-educated groups, with education-related gaps increasing by about 30 percent between the 1980s and 2000.44 Moreover, increased education differences among the elderly account for much of this growing gap in mortality and life expectancy. Meara and her colleagues note that these education-related changes “occurred during a period of increasing attention to health disparities and increased public spending designed to improve the health of less-advantaged populations.”45 One significant factor contributing to the growing differences in life expectancy appears to be the success of tobacco control policies that substantially reduced consumption of cigarettes. Meara, Richard, and Cutler point out that declines were greatest among the most-educated groups and cite this as another illustration of how prevention can widen disparities in health across education and income groups. Looking ahead to our current public health challenge regarding increased obesity, the authors conclude that better-targeted efforts to push successful health interventions into less-educated groups may be necessary to reduce socioeconomic disparities in health.46 Interestingly, they fail to suggest a more direct and potentially more effective policy approach – that is, by first improving the education level of those groups.
Although I have emphasized the crucial role of education as perhaps the most significant upstream determinant of differences in health, another “downstream” determinant involves the effectiveness and efficiency of the particular health care delivery system that a patient encounters. Despite the temptation to chase after narrowing age- and race-based health disparities at a crude aggregate level with premature inferences about provider bias, the more important role of geography needs closer consideration. Health care is local. Its potential effect on the health of a particular patient depends greatly on where the latter actually lives. For example, geography shapes such factors as the volume of procedures conducted at particular hospitals and the availability of primary and specialist care.47 Hence, learning how to deliver more effective care, measuring and reporting how well various providers perform in meeting that standard, and designing more robust methods to ensure better accountability and incentivize improved care delivery remain essential elements in reducing broader geographic variations in care, and resulting health – instead of mislabeling those variations as health inequalities based primarily on age, race, or other demographic characteristics.
Lesson: Account for Time Lags in the Causes of Future Health Outcomes
The bias of our health care system toward intervening with relatively expensive health care services during the stages of life when symptoms of poor health become more acute, then chronic and ultimately fatal reaches its peak in care for the elderly. This “just in time” mode of medical intervention, even allowing for its occasionally more efficient instances of preventive screening or even rarer instances of customized but comprehensive treatment of the “whole person” that incorporates social welfare needs along with medical ones, still tends to overlook and obscure the long latency of early developmental factors behind many chronic illnesses that manifest much later in life.
The roots of those seemingly age-related chronic conditions actually may extend back to early childhood environments (if not initial fetal programming). David Barker’s ideas on the fetal origins of adult chronic diseases (the so-called womb with a view hypothesis) highlight the key role of the prenatal environment in causing gene expression that gives rise to susceptibility to different diseases, abilities, and personality characteristics.48 As Robert Fogel observes, “[T]he severity and extent of chronic diseases at middle and late ages are, to a large extent, due to environmental insults at early ages, including in utero.”49 For example, nutritional deficiency for a developing fetus (e.g., low birth weight) will differentially compromise important functions that are operative only much later in the entire life cycle.
“Interventions to boost both health and skills development are much more effective in early childhood than are medical investments later to maintain the depreciated health capital of the elderly. Early investments ultimately cost less and deliver larger returns later.” –Thomas Miller
Because the visible onset and the consequences of many chronic conditions do not become apparent until those later ages (even until after age 65), some may tend both to understate the role that improving environmental factors increases what Fogel terms the initial stock of “physiological capital” with which a given generation, population, or individual begins life, and to overstate the degree to which the decline in mortality and morbidity during recent decades was due to improved medical technology.50 Hence, what we measure today as apparent “health inequalities” among the elderly may be largely due to environmental factors (such as nutrition) of an earlier period that are having their effects only much later. However, there is a broader trend over the past two centuries toward more equal distribution of physiological capital.
Fogel’s research reaches several important conclusions regarding the sources of, and solutions to, remaining health status inequities. Prenatal and early childhood care and environmental issues are the most important areas of intervention to enhance the robustness and capacity of vital organ systems (the initial stock of “physiological capital”) as well as to affect its rate of depreciation. The main contribution of more advanced medical treatment over the last third of the 20th century has been to slow down the rate of depreciation in the stock of physiological capital that members of particular cohorts accumulated during much earlier developmental ages. Hence, greater emphasis on lifestyle change is the key to improving health equity in rich countries like the United States.51 Fogel notes that the risky behaviors that undermine the accumulation of enhanced physiological capital are most prevalent among the poor and poorly educated. Although he also concludes that greater access to clinical care is a high priority in promoting greater health equity, he finds that we cannot simply rely upon the extension of health insurance alone to achieve it. More aggressive outreach programs and more convenient access to health care services are needed most of all.
The Fogel analysis, to be sure, does not fit neatly into more conventional approaches to health inequalities among either the overall population or its more elderly cohorts. Nor does it offer a simplified alternative health policy prescription. What it does strongly suggest is a rebalancing and broadening of our portfolio of public and private investments in improving the absolute and relative health of more vulnerable population groups, along with a greater appreciation of the long time line for measuring its results.
Lesson: The Long-Term Effects of Early Advantages and Disadvantages in Skills Development
A related line of research by James Heckman further examines how broader “skill development” factors strongly influence resulting differences in health status across various groups.52 He finds that gaps in health status, like gaps in abilities and skills, are shaped and show up at early ages and persist – well before formal education in school begins. Individuals’ early behavioral traits and development of their cognitive and non-cognitive skills are strongly influenced by their family, and this early foundation will go on to affect greatly the evolution of one’s health capital into adulthood.53 Heckman emphasizes that the development at early ages of non-cognitive abilities like perseverance, motivation, time preference, risk aversion, and self-control will have direct effects on future health choices and health outcomes as an adult, as well as on one’s social and economic well-being. Because those early advantages and disadvantages accumulate quickly, Heckman finds that interventions to boost health and skills development are more effective in early childhood than later in life. Although later interventions for disadvantages may be possible, they are likely to be much more costly and less effective than early remediation, he counsels. Hence, his broader policy recommendation is that building mutually reinforcing early advantages for targeted populations is much less expensive than trying to correct developmental deficits and their likely consequences several years (or decades) later.54
The broader message from the work of both Fogel and Heckman is that to improve overall population health as well as reduce longer-standing health disparities, interventions that improve environments during early childhood – particularly for more disadvantaged and vulnerable populations – would provide more bang for the buck (at least if public policymakers were less prone to the sort of hyperbolic discounting more commonly attributed by researchers to potentially empowered health care consumers). However, this change in emphasis (at least in public policy) to move away from playing medically-intensive catch up games, with diminishing returns, would appear to conflict with the current tendencies of our health care politics to encourage over-investment in public spending on more complex, costly, and intensive health care services for the elderly as a whole, rather than address much earlier the roots of the various differences in their health status later in life.
Lesson: Account for Patient Variation and Technological Innovation
Another important factor behind differences in patients’ health outcomes is how effectively they use the health care system and how it responds to their demands. This suggests that a more rapid overall diffusion of innovative health technology, but lags in its pattern of adoption, could appear to “worsen” health inequality even as it improves population health as a whole. Culter, Deaton, and Lleras-Muney describe an education-related health gradient whenever a mechanism or technology exists that more knowledgeable and educated people can use more effectively to improve their health.55
Sherry Glied and Adriana Lleras-Muney found that more educated individuals have a larger survival advantage in those diseases and health conditions where there has been more technological progress in medicine.56 That is primarily because more educated people appear to benefit from development of new health care technologies more rapidly than do less-educated people. Glied and Lleras-Muney observe that the former are better informed about medical innovation. They have a more positive view of its risks and benefits, and they do a better job in searching among providers that differ in quality and practice patterns.57
Moreover, groups with greater levels of education not only use the health care system more effectively; they also demand more from it. Because education encourages future-oriented behavior, additional investments in health care are more valuable to the better educated in terms of their time preferences and opportunity costs.58
Apart from its effect on the rate of medical technology adoption by patients, education factors can shape differences in their compliance with effective treatment regimes and ability to self-manage their care. Individuals with greater self-control and conscientiousness follow medical instructions and take better care of themselves. Michael Grossman describes more educated people as exhibiting better productive efficiency in obtaining better health outcomes, by allocating a given amount of health services inputs more effectively.59
The research of Dana Goldman and Darius Lakdawalla suggests that economy-wide growth in levels of education may encourage a particular kind of technological change, one that involves patient-intensive, “own-time” investments as opposed to simpler “time-saving” technologies.60 They imply that the latter type of medical technology change is less likely for diseases confined more to the educated or rich. Hence, Goldman and Lakdawalla expect people with chronic, but treatable, conditions to exhibit greater health disparities.61 (Any connection to health inequality among the elderly remains unstated by them, but may not be coincidental). They therefore conclude that prevention of treatable conditions is more effective than prevention of untreatable disease in reducing health inequality.62
Finally, David Cutler, Ellen Meara, and Seth Richards explain how research effort in medical technology and treatment tends to be targeted to address the most common health conditions in the population as a whole. They conclude that a byproduct of this “induced innovation,” which necessarily responds more to the medical needs of majority groups (like whites in the U.S.) is growth in mortality disparities between minority and majority groups.63
Accordingly, it should be no surprise that a health care system driven most effectively by its dominant payers and most articulate and active voters is likely to direct its technological development and deliver better results to those with more education and income.
Setting Priorities and Targets
Before we are swallowed up in a never-ending effort to reverse apparent statistical indicators of health inequality among various groups (such as the elderly), we need to appreciate both their complex chain of causes and the quite diversified portfolio of policy approaches more likely to help ameliorate the most significant health disparities. Perceived opportunities to find and address any measurable differences in the health of various population groups, or even to target the most vulnerable ones for intervention, may appear limitless. However, the availability of resources and effective tools to alleviate, if not eliminate, them are not. Doing something, but not everything, and doing it better presumes a more realistic understanding of the degree to which more heightened levels and frequencies of intervention are advisable.
This article notes the lack of deep and robust evidence of health inequality among the elderly in this country. Many supposed indicators of health inequality are based on health inputs and processes of care rather than outcome-based measures that should matter most. They fail to focus on the more powerful
causes of health differences which reach well beyond the quantity (and even the quality) of medical services delivered and received. Moreover, a humbler, but more effective approach to improving the health of vulnerable populations and reducing unnecessary differences in their health outcomes would take into account the entire time frame in which the health of the elderly is shaped and the long latency of crucial developmental factors in early life. It would also acknowledge the complicating factors of variation in patients’ characteristics and capabilities, and particularly how they demand and use rapidly improving medical technologies.
One important starting point is to re-focus on achieving absolute, rather than relative, improvement in health outcomes. We should obsess less in politics and policy over relative inequalities for different large subgroups of the population in accessing more quantities of health services and target instead efforts to achieve absolute improvements in their respective health outcomes.
Rather than pour more resources into just-in-time medical interventions to reduce slightly the predictable findings of some health differences among the elderly that took a half-dozen or more decades to produce, we should consider investing earlier, if not more often, in children from disadvantaged environments. Interventions to boost both health and skills development (the mutually reinforcing components of health and human “capital”) are much more effective in early childhood than are medical investments later to maintain the depreciated health capital of the elderly. Early investments ultimately cost less and deliver larger returns later.
The implications for intervention priorities are to aim for more far-reaching improvement in the health of future generations of elderly Americans – through more focus on prenatal and early childhood care, encouragement of appropriate behavior by pregnant women, and reformation of destructive lifestyle practices that “are more frequently practiced among the poor and the poorly educated than among the rich and the well-educated.”64
The physiological-capital strategy of early intervention also implies that improving outreach, mentoring, and education; and providing more convenient access to and better delivery of health services represent more important factors than expanding the level and scope of health insurance coverage per se. It still needs to be complemented and bolstered by a health-capital strategy that helps more vulnerable people effectively “produce” better health. Important components of the latter include improving their educational opportunities, deregulating the delivery and financing of medical services to provide those patients with more choice and control, and increasing competition in health care markets. This strategy also would expand counseling support to encourage more future-oriented behavior, offer more assistance for consumers navigating a complex health system, and improve access to more actionable consumer health information.
To be sure, there will remain millions of chronically or acutely ill older patients in the meantime who need access to more effective and efficient medical treatment – no matter how much progress might be made in providing better incentives and tools for improving the installed base of health capital for younger and future generations. In determining which heightened levels and frequencies of medical intervention are advisable for the most vulnerable groups among the elderly, we should consider the degree to which they address more persistent and treatable, but less avoidable, conditions. Unfortunately, targeting resources on the most medically significant problems among the elderly may conflict with the greater political marketability of higher visibility (and broader-based) interventions. In weighing the political factors between addressing the greatest needs versus claiming the greatest credit (serving fewer people, but the most vulnerable ones, better – or simply serving more at the medical benefits buffet line), the temptation for officeholders and public program administrators remains to slice the salami so thin that all the elderly get is more baloney.
1. J. Lee, Looking for Love in All the Wrong Places with Lyrics, available at (last visited January 17, 2012).
2. Agency for Healthcare Research and Quality, National Healthcare Disparities Report 2008, AHRQ Publication No. 09-0002, available at (last visited January 16, 2012).
3. NHDR 2008 was the most recent report available shortly after the date of symposium presentation from which this article was adapted.
4. Id., at 1
5. Id., at 172
6. E. A. McGlynn, S. M. Asch, J. Adams, J. Keesey, J. Hicks, A. DeCristofaro, and E. A. Kerr, “The Quality of Health Care Delivered to Adults in the United States,” New England Journal of Medicine 348, no. 26 (2003): 2635-2645, at 2643; S. M. Asch, E. A. Kerr, J. Keesey, J. L. Adams, C. M. Setodji, S. Malik, and E. A. McGlynn, “Who Is at Greatest Risk for Receiving Poor-Quality Health Care?” New England Journal of Medicine 354, no. 11 (2006): 1147-1156.
7. See National Healthcare Disparities Report 2008, supra note 2, at 19-21 (Table 1.2), and 287-289.
8. Id., at 248-255.
9. A. Finkelstein and R. McKnight, “What Did Medicare Do? The Initial Impact of Medicare on Mortality and Out-of-Pocket Medical Spending,” Journal of Public Economics 92, no. 7 (2008): 1644-1668, at 1650, 1653, 1665.
10. Id., at 1650-1653.
11. A. Finkelstein, “The Aggregate Effects of Health Insurance: Evidence from the Introduction of Medicare,” Quarterly Journal of Economics 122, no. 1 (2007): 1-37, at 29-33.
12. Id., at 33.
13. D. Card, C. Dobkin, and N. Maestas, “Does Medicare Save Lives?” Quarterly Journal of Economics 124, no. 2 (2009): 597- 636, at 598.
14. Id., at 632-633.
15. M. McClellan and J. Skinner, “The Incidence of Medicare,” National Bureau of Economic Research, NBER working paper no. 6013 (1997): 1-58 at 3-4, 47.
17. J. Lee, M. McClellan, and J. Skinner, “The Distributional Effects of Medicare,” in J. M. Poterba, ed., Tax Policy and Economics, vol. 13 (Cambridge, MA: MIT Press, 1999): 85-108, at 86-87.
18. M. McClellan and J. Skinner, “The Incidence of Medicare,” Journal of Public Economics 90 (2006): 257-276, at 258.
19. Id., at 267, 272.
20. J. Bhattacharya and D. Lakdawalla, “Does Medicare Benefit the Poor?” Journal of Public Economics 90 (2006): 277-292, at 277-278, 290.
21. Id., at 277-279, 282, 288-289.
22. Id., at 289.
23. Id., at 282.
25. J. Skinner and W. Zhou, “The Measurement and Evolution of Health Inequality: Evidence from the U.S. Medicare Population,” in A. J. Auerbach, D. E. Card, and J. M. Quigley, eds., Public Policy and the Income Distribution (New York: Russell Sage Foundation, 2006): 288-319, at 296, 306.
26. Id., at 301-307.
27. Id., at 293.
28. Id., at 292, 308.
29. Congressional Budget Office, Geographic Variation in Health Care Spending, February 2008, at 1.
30. Id., at 2.
31. J. E. Wennberg. E. S. Fisher, and J. S. Skinner, “Geography and the Debate over Medicare Reform,” Health Affairs Web exclusive (February 13, 2002): W96-W114, at 104.
32. M. G. Marmot, “Understanding Social Inequalities in Health,” Perspectives in Biology and Medicine 46, no. 3 (2003): S9-S23.
33. J. E. Ferrie, ed., Work, Stress and Health: The Whitehall II Study, Public and Commercial Services Union on behalf of the Council of Civil Service Unions/Cabinet Office, 2004, available at (last visited January 17, 2012).
34. World Health Organization Commission on Social Determinants of Health, Closing the Gap in a Generation, 2008, at 1, available at (last visited January 17, 2012).
35. Id., at 2-3.
36. J. M. McGinnis, P. Williams-Russo, and J. Knickman, “The Case for More Active Policy Attention to Health Promotion,” Health Affairs 21, no. 2 (2002): 78-93.
37. M. Grossman, “On the Concept of Health Capital and the Demand for Health,” Journal of Political Economy 80, no. 2 (1972): 223-252.
38. J. P. Smith, “Consequences and Predictors of New Health Events,” in D. A. Wise, ed., Analyses in the Economics of Aging (Chicago: University of Chicago Press, 2005): 213-240, at 230-231.
39. A. Lleras-Muney, “The Relationship between Education and Adult Mortality in the United States,” Review of Economic Studies 72, no. 1 (2005): 189-221, at 215.
40. D. M. Cutler, A. S. Deaton, and A. Lleras-Muney, “The Determinants of Mortality,” Journal of Economic Perspectives 20, no. 3 (2006): 97-120, at 114.
41. Id., at 115.
42. M. Grossman, “The Human Capital Model,” in A. J. Culyer and J. P. Newhouse, eds., Handbook of Health Economics, vol. 1A (Amsterdam, The Netherlands: Elsevier, 2000): 347-408, at 397-404; M. Grossman and R. Kaestner, “Effects of Education on Health,” in J. R. Behrman and N. Stacey, eds., The Social Benefits of Education (Ann Arbor, Michigan: The University of Michigan Press, 1997): 1-273, at 69.
43. Id.; see Grossman, supra note 41, at 397-404.
44. E. R. Meara, S. Richards, and D. M. Cutler, “The Gap Gets Bigger: Changes In Mortality and Life Expectancy, by Education, 1981-2000,” Health Affairs 27, no. 2 (2008): 350-360, at 350, 353.
45. Id., at 356-357.
46. Id., at 358-359.
47. J. Klick and S. Satel, The Health Disparities Myth (Washington, D.C: AEI Press, 2006): 1-80, at 28-32, 38-40.
48. D. J. P. Barker, “Fetal Origins of Coronary Disease,” BMJ 311, no. 6998 (July 15, 1995): 171-174; D. J. P. Barker “The Fetal and Infant Origins of Adult Disease,” BMJ 301, no. 6761 (November 17, 1990): 1111.
49. R. W. Fogel, “Changes in the Physiology of Aging during the Twentieth Century,” National Bureau of Economic Research, NBER working paper no. 11233 (2005): 1-12, at 7, available at (last visited January 19, 2012).
50. R. W. Fogel, “Secular Trends in Physiological Capital: Implications for Equity in Health Care,” Perspectives in Biology and Medicine 46, no. 3 (2003): S24-S38, at S34.
51. Id., at S34-S35.
52. J. J. Heckman, “The Economics, Technology, and Neuroscience of Human Capability Formation,” Proceeding of the National Academy of Sciences 104, no. 33 (2007): 13250-13255.
53. Id., at 13250.
54. Id., at 13251-13252.
55. See Cutler, Deaton, and Lleras-Muney, supra note 40, at 115.
56. S. Glied and A. Lieras-Muney, “Health Inequality, Education, and Medical Innovation” National Bureau of Economic Research, NBER working paper no. 9738 (November 2003): 1-56.
57. Id., at 22-23.
58. See Grossman and Kaestner, supra note 40, at 69; Grossman, supra note 40, at 347, 396.
60. D. Goldman and D. Lakdawalla, “Understanding Health Disparities across Education Groups,” National Bureau of Economic Research, NBER working paper no. 8328 (2001): 1-47, at 5.
61. Id., at 37.
62. Id., at 36-37.
63. D. M. Cutler, E. Meara, and S. Richards, “Induced Innovation and Social Inequality: Evidence from Infant Medical Care,” National Bureau of Economic Research, NBER working paper no. 15316 (September 2009): 1-48.
64. See Fogel, supra note 49, at S35.
There are no comments available.
1150 17th Street, N.W. Washington, D.C. 20036
© 2014 American Enterprise Institute for Public Policy Research