Printer Friendly

Poverty and health: asking the right questions.

Gertrude Stein, confidante of the leading writers, artists, and intellectuals of her time, lay dying. Her closest friend and lifetime companion, Alice B. Toklas, leaned forward and said, "Gertrude, what's the answer?" Gertrude looked up and with her last breath said, "Alice, what's the question?"

We have been brought together in this conference to discuss "medical care and the health of the poor." But what is the question? Or, more appropriately, what are the questions? Opinions will no doubt differ, and some may wonder why the issue is raised at all. At most conferences participants arrive prepared with a ready supply of answers. They seek to advance this or that position or program, and, like a popular TV game show, the only questions that matter are those for which answers are already in hand.

But this is a different game. I must ask your indulgence while I engage in what some may regard as an "academic exercise." After all, of what possible use is an economist if not to discuss whether something that works in practice also works in theory? But, what if we observe something that doesn't work in practice--such as health care for the poor? The purpose of this paper is to raise theoretical questions about poverty and health in order to elicit answers that might improve public policy.

Who are the poor?

A logical place to begin is by asking what we mean by poverty; that is, who are the poor? This question has a long history within economics, and even from the perspective of that single discipline gives rise to considerable controversy over definition and measurement. The question becomes even more important, however, when poverty is discussed in relation to health. As an economic concept, there is general agreement that poverty refers to some measure of income (or wealth) which indicates "inadequate" command over material resources. At health conferences, however, the concept often gets transformed into an amorphous set of "socioeconomic conditions," or an ill-defined "culture of poverty."

Let us try to avoid such confusion. This is not to deny that people can be "poor" in ways other than economic. They can be "spiritually impoverished," "morally bankrupt," in "poor health," and so on. But, to the extent possible, let us strive for clarity. If we mean low income, let's say low income. If we mean education, let's say education. And if we mean alcoholism, cigarette smoking, crime, drug abuse, fragmented families, hazardous occupations, sexual promiscuity, slum housing, social alienation, or unhealthy diets, let's say so explicitly. If we constantly redefine poverty to include anything and everything that contributes to poor health, we will make little progress either in theory or practice.

Even when poverty is defined in terms of income, there are numerous questions such as adjustment for size and composition of household, but we can leave them to the specialists.(1) There is one conceptual issue, however, which is so important as to require explicit discussion. Should poverty be defined according to some fixed standard (absolute income), or by position in the income distribution (relative income)? In my judgment, we need to combine both approaches. If we cling only to a fixed standard, economic growth gradually raises almost everyone out of poverty so defined, but the problems we usually associate with poverty persist. So-called "subsistence" budgets are adjusted to new social norms. On the other hand, to define poverty in terms of the bottom 10 or 20 percent of the income distribution does not help us get to the heart of the problem either. In a society with little inequality of income, being at the lower end need not have the same negative implications as when the distribution is very unequal.

People usually think of themselves as poor (and are regarded as poor) when their command over material resources is much less than others. Poverty, as an economic concept, is largely a matter of economic distance. Thus, in 1965, I proposed a poverty threshold of one-half of median income.(2) The choice of one-half was somewhat arbitrary, but the basic idea would not change if a level of four-tenths or six-tenths were chosen instead. There is considerable resistance to such a definition because a reduction in poverty so defined requires a change in the distribution of income--always a difficult task for political economy. But I believe it is the only realistic way to think about poverty. In this respect, as in so many others, Adam Smith had a clear view of the matter over 200 years ago. He wrote, "By necessaries I understand not only the commodities which are indispensably necessary for the support of life but whatever the custom of the country renders it indecent for creditable people even of the lowest order to be without."(3)

What is the relation between poverty and health?

Once we have identified the poor, the next question concerns their health relative to the rest of the population. We know in general the answer to this question--on average those with low income have worse health. There are, however, several aspects of the question that deserve further exploration. How does the relation vary with different measures of health, such as morbidity, disability, or mortality? Is the relation different for different diseases? Is it different at different stages of the life cycle? Is the relation stronger in some countries than in others? If any of these questions are answered in the affirmative (and they surely will be), the next step would be to seek to determine the reasons for the variation. Such inquiries could provide valuable inputs into the next stage of analysis when we seek to make inferences about causality.

Is low income the cause of poor health?

Many writers simply assert that poverty is the cause of poor health, without rigorous testing. In England, social class is often used as a proxy for poverty, but this is problematic, as illustrated in Table 1. There is a large differential in mortality between the lowest and the highest class and a large differential in income as well, but more detailed inspection reveals a complex pattern. Class II has only 5 percent greater mortality than class I, even though income is 23 percent lower. In contrast, the differential in mortality between classes IV and V is 21 percent, but the income difference is only 2 percent. It may be tempting to explain these data by asserting that the relationship between income and mortality is nonlinear. Thus, at low levels of income, i.e., classes IV and V, even a small increase in income has a strong effect on mortality, while at high levels, i.e., classes I and II, the effect is very weak. This explanation won't wash, however, once we note that the mortality differentials between classes I and V were no smaller in 1971 than in 1951. During those two decades real earnings rose by more than 50 percent for all classes; thus if nonlinearity is the explanation for the pattern shown in Table 1, there should have been an appreciable narrowing in the class mortality differentials between 1951 and 1971. No such decrease occurred. Furthermore, there was no decrease between 1971 and 1981 despite additional increases in real income.
Indexes of mortality and income in England and Wales by social
class, 1971 (class I = 100)

 Age-adjusted mortality, weekly
 men 15-64 years of age income

I. Professional 100 100
II. Managerial 105 77
III. Skilled 136 58
IV. Semi-skilled 148 51
V. Unskilled 179 50

Source: Adapted from Wilkinson, pp. 2 and 11. Richard G.
Wilkinson, "Socioeconomic Differences in Mortality:
Interpreting the Data on Their Size and Trends," in Richard G.
Wilkinson, ed., Class and Health: Research and Longitudinal
Data (London and New York: Tavistock Publications, 1986).

England is not alone in experiencing persistence of class (occupation) differentials in mortality in the face of rising real income and universal coverage for medical care. In Scandinavia, the age-standardized mortality ratio for male hotel, restaurant, and food service workers is double that of teachers and technical workers.(4) A Swedish study of age-standardized death rates among employed men ages 45-64 found substantial differences across occupations in 1966-1970 and slightly greater differentials in 1976-1980.(5) In Sweden, there is growing recognition that these differentials cannot be explained by differential access to health care. Johan Callthorp, M.D., writes, "There is no systematic evidence that the health care system is inequitable in the sense that those in greater need get less care or that there are barriers towards the lower socioeconomic groups."(6)

What explains the correlation between poverty and health?

The fact that variables A and B are correlated does not, of course, prove that A is the cause of B. Two other possibilities must be considered. First, the causality may run in the opposite direction; B may be the cause of A. The possibility that health affects social class has been explored extensively by British writers.(7) Almost all agree that there is some "selective mobility," but no consensus has emerged regarding its importance. R. G. Wilkinson concludes that "its contribution to observed class differences in health is probably always small."(8) But Roy Carr-Hill writes "There is an effect which should not be ignored: the size of the effect could be substantial, but it cannot be estimated properly without a lifelong longitudinal study."(9)

Attention must also be paid to the other logical possibility, namely that there are one or more "third variables" that are the cause both of low income and poor health. These variables could include genetic endowment as well as numerous socioeconomic factors. Among the latter, most U.S. studies have focused on schooling. There is a vast literature that explores the relation between health and education.(10) To be sure, income and education are correlated, but the correlation is not so high as to preclude attempting to sort out their separate relationships with health. In the United States, the coefficient of correlation between education and income within age-sex-race groups never reaches as much as .50 and is typically around .40. When health is regressed on both income and schooling, the latter variable always dominates the former. Indeed, in some studies, income is negatively related to health, once years of schooling is controlled for.(11)

Why is there such a strong correlation between schooling and health?

One possible answer, of course, is that schooling is the cause of good health. That is, at any given level of income, those with more education know how to use medical care more effectively, choose better diets and other health behaviors, and so on. This line of reasoning has been developed most fully by Michael Grossman.(12) But again, as a matter of logic, we must consider two other possibilities. Good health may lead to more schooling, or, there may be "third variables" that affect both schooling and health. Among the "third variables," my favorite candidates are time preference(13) and self-efficacy.(14)

Time preference is an economic concept that refers to the rate at which people discount the future relative to the present. Individuals with high rates of time preference will tend to invest less in the future: on average they will have less education, lower income, and worse health. A perfect capital market would enable those with low rates of time discount to provide funds to those with high rates until their rates were equal at the margin, but the real world bears little resemblance to this theoretical model. For one thing, low income individuals who want to borrow a great deal cannot provide effective collateral. Also, many choices about health do not involve money; thus there is no effective market in which individuals with different rates of time preference can make trades.

Self-efficacy is a psychological term that describes people's beliefs in their capability to exercise control over their own behavior and their environment. Differences among individuals in self-efficacy are probably correlated across several domains, such as health and education, thus helping to explain the close relationship between these variables.

How does low income affect health?

Let us return to the line of inquiry that has poverty as a cause of poor health. Within that framework the central question concerns the mechanism through which low income translates into bad health. To what extent does the health of the poor suffer because they have inadequate access to medical care? To what extent is their poor health the result of deficiencies in other health-producing goods and services such as good food, good housing, or a safe environment? If poor health is attributable to inadequate medical care, are the barriers faced by the poor simply a matter of purchasing power, or are there other impediments?

What are the most important health problems facing the poor?

In addressing this question it is important to distinguish between relative risk and absolute risk, a distinction that is often obscured in the media and even in policy discussions. For example, infant mortality may be twice as high among the poor as the non-poor (a relative risk of 2 to 1), while the differential in mortality from heart disease may be only 50 percent (relative risk 1.5 to 1). The absolute level of risk of infant mortality, however, may be very low relative to heart disease mortality; thus, the poor might benefit more from efforts devoted to heart disease rather than to infant mortality.

To illustrate this point, consider the tremendous attention given by the media (and many health policy experts) to black-white differences in infant mortality and the relative neglect of other black-white health differentials. It is true that the black infant death rate is double the white rate, while the difference in overall life expectancy is only 9 percent (75.9 years versus 69.7 years in the United States in 1989). But if the black infant mortality rate were reduced to the white level (and all other age-specific rates remained unchanged), black life expectancy would only rise by six-tenths of a year. Over 90 percent of the black-white difference in life expectancy would remain. Isn't there a danger that undue emphasis on attention-grabbing headlines results in a misallocation of health care resources from the perspective of those whose health problems are being addressed?

Which health problems of the poor are most amenable to solution?

In order to make rational allocations of resources to alleviate the health problems of the poor it is necessary but not sufficient to know the relative importance of the problems. It is also necessary to know how readily the problems can be solved or alleviated. Unfortunately, the bulk of health policy research dwells on documenting the problems of the poor, while neglecting the more difficult task of assessing the efficacy of alternative interventions. Policymakers and the public need to know both the costs and the benefits of such alternatives. For example, treatment for infectious diseases may be very efficacious, while treatment for cancer may not be. Some prevention programs such as immunizations may provide a great deal of benefit for little cost, but others, such as mass screening of cholesterol levels, may use a vast amount of resources for limited benefits.

Are there reasons for providing medical care to the poor other than improving health outcomes?

Suppose the contribution of medical care to health at the margin is quite small. Is that sufficient reason to ignore the provision of care to the poor? Not necessarily. In his critique of the Oregon plan for rationing medical care to the poor, Bruce Vladeck writes, "We expect the health system to take care of sick people whether or not they are going to get better."(15)

Medical care may be valued by the poor (as it is by the nonpoor) for the caring and validation services that it provides. If this is the case, serious questions arise concerning the kind of care provided to the poor. In particular, is "high-tech" overemphasized at the expense of simpler, more valuable services? The fact that medical care has value apart from improving health outcomes provides no grounds for rejecting a cost-benefit approach to resource allocation. But it does highlight the need to incorporate the value of all services in such analyses.

What policy instruments are available to help the poor?

A sociologist tried to explain poverty to a colleague in economics. "You know, the poor are different from you and me." "Yes," replied the economist, "they have less money." This apocryphal exchange highlights a continuing controversy over the best way to help the poor with respect to health or anything else. If more resources are to be allocated to the poor, is it better to provide cash and allow the poor to decide how to spend it, or should the transfers be tied to particular goods and services? The arguments for tied transfers usually derive from a paternalistic assumption that the poor, left to their own devices, will not spend the money "wisely," i.e., they will buy cake when those making the transfers think they should buy bread. A more sophisticated version of this argument invokes "externalities." It may be the case that forcing the poor to spend their additional resources on immunizations rather than alcohol helps the nonpoor because the former creates positive externalities, while the latter creates negative ones.(16) But the same is true of expenditures by the nonpoor.

Paternalism aside, there is the practical question of whether tied transfers can alter consumption patterns. If a family that previously spent $250 per month on food receives $100 worth of food stamps, there is no reason to expect their spending on food to rise to $350. Indeed, food expenditures are not likely to increase by any more than if they were given $100 in cash. The relative price of food, at the margin, is no different after the transfer than before. The only way to assure a disproportionate increase in food consumption would be to provide food stamps greater in amount than what the family would voluntarily spend on food, given its income plus the cash value of the food stamps.

In devising programs for the poor, physicians usually advocate more medical care, educators more schooling, the construction industry more housing, and so on. But what area(s) would the poor give highest priority? This question may be beyond the scope of this conference, but it cries out for attention from policy analysts in some setting.

In choosing between in-kind and cash programs, policymakers should also consider the pecuniary effects of alternative transfers to the poor.(17) One result of Medicare and Medicaid, for example, was higher incomes for physicians--surely not a goal of the Great Society. These programs also led to an increase in the price of medical care for the general public, including many low income persons who did not qualify for Medicaid. If, instead of Medicare and Medicaid, the government had transferred to the elderly and the poor an equivalent amount of cash, some of it would have been used for medical care, but much of it would have been used for other goods and services, including food, clothing, consumer durables, and the like. The income and price effects would probably have been very different from those of Medicare and Medicaid, and possibly more egalitarian.

Why are Americans less willing than others to subsidize medical care for the poor?

The health policy literature abounds with papers that describe and decry the difficulty faced by poor Americans in obtaining health care. But these papers are typically silent as to why the United States is the only major industrialized country that does not have national health insurance. In 1976, I proposed several answers to this question: distrust of government, the heterogeneity of the population, the weakness of "noblesse oblige," and a robust voluntary sector.(18) In a recent paper I reappraise these explanations in the light of subsequent political, social, and economic developments.(19) I have a healthy respect for my opinion, but it would be useful to hear other views on this question.

What is the most efficient way to provide medical care for the poor?

The debate on this issue is clear cut. On the one hand are those who want to provide the poor with health insurance and leave it to them to obtain the care they need. The contrary view advocates special programs directly aimed at providing care for the poor. Inasmuch as both approaches have been tried in the United States and abroad, it should be possible to make some judgments about their relative costs and benefits.

Is it acceptable to provide highly cost-effective care for the poor although the care is different from that available to the nonpoor? A good example is prenatal care and delivery of babies. The Maternity Center Association can provide high quality midwifery service in their childbearing center for less than half of what Medicaid pays for in-hospital normal childbirth.(20) At present, some poor women get the high-cost care and some get little or no care.

The question of efficient provision of care to the poor is complicated by the fact that there may be gross inefficiencies in care provided to the nonpoor--overtesting, inappropriate surgery, and so on. Should programs for the poor aim at reproducing these misallocations of resources?

What is "two-tier" medical care?

Discussions of medical care for the poor frequently invoke the phrase "two-tier" medicine. For strict egalitarians this is a deplorable concept. But others have argued that an explicit two-tier system would serve the American poor better than does the present jumble of services that range from no care (e.g., prenatal) to the most sophisticated (e.g., neonatal intensive). In thinking about this issue it may be useful to notice that two-tier systems can vary greatly, as shown in Figure 1. In both systems, the people in the first tier receive more and better service than those in the second. But in version A most of the population is in the first tier and only the poor are in tier two. In version B the proportions are reversed; most of the population is in the second tier and only the affluent and/or well-connected are in tier one.

Version A provides a "safety net"; version B provides an "escape valve." Most Americans tend to associate two-tier medicine with version A; most other countries have opted for version B. Several interesting questions may be posed about these alternative approaches. Do the two versions have different consequences for cost, access, and quality? For example, consider cost.

Suppose expenditures per capita in tier 1 are identical in the two systems and the same is true for tier 2 except that in each country they are 50 percent less than tier 1. Suppose that in system A 80 percent of the population are in tier 1 and 20 percent in tier 2, and that the proportions are reversed in system B. In that case, the average expenditure per person in system A will be 50 percent greater than in system B.

What are the political, social, and economic factors that lead a country to adopt one version or the other? It would seem that individuals who were certain that they would be in tier 2 under either system would prefer B. Similarly, individuals who were certain that they would be in tier 1 under either system might also prefer B. Supporters of A are likely to be individuals who think they would be in tier 1 under A, but in tier 2 under B. Many Americans probably fit that category.

What is basic medical care?

A frequent conclusion of health policy discussions in the United States is that everyone should have access to "basic" medical care. Many observers believe that the nonpoor would be more willing to subsidize a "basic" package than they would complete equality of care. Problems arise, however, in trying to define the contents of that package. Moreover, no matter how it is defined at any point in time, no one should imagine that the contents can remain fixed over time. In a world of changing technology and rising real income, a fixed approach to basic care will prove no more satisfactory than will a fixed poverty standard based on some notion of subsistence. The basic care package will constantly have to change in order to include "whatever the custom of the country renders it indecent for creditable people, even of the lowest order, to be without."


In summary, there are numerous questions about poverty and health that need to be addressed, both at this conference and subsequently. Many of them concern the relation between poverty and health: its extent, pattern, and explanations. Other questions revolve around possible confounding variables such as education, which is correlated with income and health. Still other questions focus on medical care: its efficacy in improving health, its value to the poor, the best way to provide it. In pursuing these questions we need to find a middle road between a mindless optimism that ignores reality and a constricting pessimism that denies the possibility of creating a more efficient and more just society.


1. See John L. Palmer, Timothy Smeeding, and Christopher Jenks, "The Uses and Limits of Income Comparisons," in The Vulnerable, eds. John L. Palmer, Timothy Smeeding, and Barbara Boyle Torrey (Washington, D.C.: Urban Institute Press, 1988), pp. 9-27.

2. Victor R. Fuchs, "Toward a Theory of Poverty," in The Concept of Poverty, Task Force on Economic Growth and Opportunity, First Report (Washington, D.C.: Chamber of Commerce of the United States, 1965), pp. 71-91.

3. Adam Smith, The Wealth of Nations (New York: Random House, Modern Library edition, 1937), p. 821.

4. O. Andersen, "Occupational Impacts on Mortality Declines in the Nordic Countries," in Future Demographic Trends in Europe and North America, ed. W. Lutz, International Institute for Applied Systems Analysis, Laxenburg, Austria (New York: Academic Press, Harcourt-Brace-Jovanovich, 1991), p. 46.

5. J. Callthorp, "The 'Swedish Model' under Pressure -- How to Maintain Equity and Develop Quality?" Quality Assurance in Health Care 1 (1) (1989): 11-22.

6. Callthorp, "The 'Swedish Model'," op. cit.

7. See, for example, A. J. Fox, Social Class and Occupational Mobility Shortly Before Men Become Fathers, OPCS Series LS No. 2, London HMSO, 1984; J. Stern, "Social Mobility and the Interpretation of Social Class Mortality Differentials," Journal of Social Policy 12 (1983): 27-49; and M.E.J. Wadsworth, "Serious Illness in Childhood and Its Association with Later Life Achievement," in Class and Health: Research and Longitudinal Data, ed. Richard G. Wilkinson (London and New York: Tavistock Publications, 1986).

8. Richard G. Wilkinson, "Socioeconomic Differences in Mortality: Interpreting the Data on Their Size and Trends," in Class and Health: Research and Longitudinal Data, ed. Richard G. Wilkinson (London and New York: Tavistock Publications, 1986), p. 10.

9. Roy Carr-Hill, "The Inequalities in Health Debate: A Critical Review of the Issues," Journal of Social Policy 16 (1987): 509-42 (p. 527).

10. See, for example, Mark C. Berger and J. Paul Leigh, "Schooling, Self-Selection, and Health," Journal of Human Resources 24 (1989): 435-55; Phillip Farrell and Victor R. Fuchs, "Schooling and Health: The Cigarette Connection," Journal of Health Economics 1 (1982): 217-30; Michael Grossman, "The Correlation between Health and Schooling," in Household Production and Consumption, ed. Nestor E. Terleckyj (New York: Columbia University Press (for NBER), 1976), and Donald S. Kenkel, "Health Behavior, Health Knowledge, and Schooling," Journal of Political Economy 99(1991): 287-304.

11. Richard Auster, Irving Leveson, and Deborah Sarachek, "The Production of Health, an Exploratory Study," Journal of Human Resources 4 (1969): 412-36.

12. Michael Grossman, "The Correlation between Health and Schooling," op. cit.

13. Victor R. Fuchs, "Time Preference and Health: An Exploratory Study," in Economic Aspects of Health, ed. Victor R. Fuchs (Chicago: University of Chicago Press, 1982), pp. 93-120.

14. Albert Bandura, "Self-Efficacy Mechanism in Physiological Activation and Health-Promoting Behavior," in Neural Biology of Learning, Emotion and Affect, ed. John Madden IV (New York: Raven Press Ltd., 1991), pp. 229-69.

15. Bruce C. Vladeck, "Unhealthy Rations," The American Prospect (Princeton, NJ), Summer (1991), p. 102.

16. For other explanations by economic theorists for tied transfers, see Neil Bruce and Michael Waldman, "Transfers In Kind: Why They Can Be Efficient and Nonpaternalistic," American Economic Review 81 (1991): 1345-51.

17. See Stephen Coate, Stephen Johnson, and Richard Zeckhauser, "Robin-Hooding Rents: Exploiting the Pecuniary Effects of In-Kind Programs," mimeo, March 1992, Harvard University.

18. Victor R. Fuchs, "From Bismarck to Woodcock: The 'Irrational' Pursuit of National Health Insurance," Journal of Law and Economics 19 (1976): 347-59.

19. Victor R. Fuchs, "National Health Insurance Revisited," Health Affairs 10 (1991): 1-11.

20. Personal communication from Ruth Watson Lubic, December 9, 1991.

Henry J. Kaiser, Jr. Professor, Stanford University and Research Associate, National Bureau of Economic Research. This paper was presented at the 1992 Cornell University Medical College Health Policy Conference, "Medical Care and the Health of the Poor," in New York City on February 27-28, 1992. Comments from Alan Garber, M.D., John Hornberger, M.D., Douglas Owens, M.D., and Richard Zeckhauser are gratefully acknowledged. Forthcoming in David E. Rogers and Eli Ginzberg (eds.), Medical Care and the Health of the Poor (Boulder, CO: Westview Press).
COPYRIGHT 1992 Omicron Delta Epsilon
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1992 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Fuchs, Victor R.
Publication:American Economist
Date:Sep 22, 1992
Previous Article:The economic transformation of Eastern Europe: the case of Poland.
Next Article:A daily inflation index.

Related Articles
Eradicating poverty: moving forward as one.
Child poverty isn't confined to Dickens' days; Inside View.
Lost in translation and political will: research and policy as a means to advance human rights.
Global health and local poverty: rich countries' responses to vulnerable populations.

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters