The paper by Daniels, Kennedy, and Kawachi is an important piece from the perspectives of social epidemiology, ethics, and policy. It raises critical issues about the determinants of health inequalities and proposes policies that might contribute to their reduction.

Specialized entities, both at the international level (e.g., the WHO) and the national level (e.g., various ministries of health), have great potential to influence policies aimed at the reduction of health inequalities.

We at the World Health Organization (WHO) give great importance to the reduction of health inequality. In a new framework for the assessment of health system performance proposed by WHO, reducing health inequality is one of four main goals for health systems. The other three goals are improving health status, enhancing the responsiveness of the health system to the legitimate expectations of the population, and protecting people in a fair manner from the financial consequences of caring for health. By explicitly listing the reduction of health inequality as one of the intrinsic goals of health systems, WHO illustrates the prominence that this problem should receive in the health policy agenda.

Daniels, Kennedy, and Kawachi make four policy recommendations for the reduction of health inequalities—all of which constitute social policies that do not primarily involve the health sector. They suggest that “to address comprehensively the problem of health inequalities governments must begin to address the issue of economic inequalities.” We agree that economic redistribution policies are intrinsically important, independent of their effect on the reduction of health inequalities, and as such should constitute good social policies regardless of the degree of health inequalities present in a society. Similarly, we consider health to be intrinsically important, independent of its association with other components of well-being.

When it comes to health, specialized entities, both at the international level (e.g., the WHO) and the national level (e.g., various ministries of health), have great potential to influence policies aimed at the reduction of health inequalities. They also have the capacity to be involved in inter-sectoral approaches aimed at improving specific determinants of health. But these entities can only do so much. For example, it may be possible to convince ministries of finance to raise taxes on tobacco, yet ministries of health typically have little capacity to influence broad economic redistribution policies. Therefore, these actors typically concentrate on within-the-health-sector approaches and on intersectoral initiatives to improve specific determinants of health.

Before making explicit policy recommendations we need to be better informed about how health systems of various countries perform in the achievement of their goals. Once we have a better understanding of the factors influencing the performance of a health system we will be better equipped to articulate policies that will lead to the reduction of inequalities.

Health Inequality

We define health inequality to be the differences in health across individuals in a population. We are using the individual as the unit of analysis and are interested in studying the inequality in the distribution of health. We propose to use health expectancy as the measure of health. Health expectancy—the number of years that an individual born today is expected to live in the equivalent of full health—reflects the risk of mortality and the risk of non-fatal health outcomes that an individual faces at each age. It is important that the measure of health reflects not only a risk of death but also the risk of being in ill-health.

Before we try to measure inequality of health expectancy, though, we first ask what, if any, components of health expectancy are either not amenable to change or arise from fully informed choices of individuals to decrease their health expectancy through the pursuit of risky activities. If there are differences that could never be remedied by intervention or new technology, one might argue that we should be uninterested in them. But what component of the distribution of health expectancy is not amenable to intervention? That due to genes? That due to chance during birth? In both cases, the argument that we cannot intervene to change the effects on the distribution of health expectancy seems specious. There is little evidence of significant cross-population variation in the contribution of genes. And with current improvements in technology and future progress, it is likely that even genes will become amenable to change.

What about volition? How much of the distribution of health expectancy for a population is due to fully informed choices of individuals who have a taste for risky behavior? This seems like a very slippery slope. What choices affecting health are fully informed? Would we exclude the effects of tobacco on health expectancy because smoking is a choice? Even if we claim that the choice was informed, should it be excluded? We argue that it should not be excluded. First, in most cases health risks are not adopted because of a love of risky behavior but rather for other, less informed, reasons.1 Second, the true volitional component of the distribution of health expectancy is likely to be very small and can well be ignored. This argument is similar to ones used to explain certain measures of income inequality, where the variation in the distribution of income due to different trade-offs between leisure and income within the population is routinely ignored in the measurement of income inequality.

Finally we ask, how can health expectancy be measured? Risk, after all, is not observed; only outcomes are. But the distribution of health risks can be reasonably approximated through a variety of techniques. Together, they allow us to measure the distribution of four key dimensions: child mortality risk, adult mortality risk, life expectancy and health expectancy through small area analyses, and non-fatal health outcomes.

1. Child Mortality Risk. We can observe the variation in the proportion of a mother’s children who have died, which provides information at a very fine level of aggregation (namely households) on the distribution of child death risk. Using simulation, we can evaluate the difference in the distribution of outcomes from that which would be expected based on a distribution of equal risk. Data on children ever born and children surviving for women of different ages are widely available from the Living Standards Measurement Studies (LSMS), the Demographic and Health Surveys (DHS) and many censuses and surveys. We have implemented this strategy for measuring child mortality.

2. Adult Mortality Risk. We do not have good data to measure the distribution of adult mortality. Information on the survivorship of siblings could in principle be used but it would refer to average mortality experience over decades and the technical challenges have yet to be solved. Other strategies need to be developed.

3. Life Expectancy or Health Expectancy for Groups. We can divide the population into groups that are expected to have similar health expectancies and measure directly the health expectation for those groups. Inevitably, this will underestimate the distribution of health expectancy. The more refined the groupings are, the more we will approximate the true underlying distribution. Small-area analyses hold out the promise of being one of the most refined methods for revealing the underlying distribution of health expectancy in a population. For example, a detailed age-sex-race group analysis of counties in the United States has revealed a range in life expectancy across counties of 41.3 years—almost as large as the range across all countries of the world.

4. Non-Fatal Health Outcomes. Measurement of non-fatal health outcomes on continuous or polychotomous scales provides more information from which to estimate the distribution of risk across individuals. Numerous surveys provide information on self-reported health status using a variety of instruments. The main problem to date with this information is the comparability of the responses across different cultures, levels of educational attainment, and incomes. For example, the rich often report worse non-fatal health outcomes than the poor. Problems of comparability must be resolved before such data sets can be used to contribute to estimation of health expectancy in the population.

For the WHO, the way forward will be to simultaneously pursue the development of methods and data sets to measure these different dimensions of the distribution of health expectancy. We recognize that there is a great need for new methods to integrate these different measurements into one estimation of the distribution of health expectancy in populations. Based on the wide array of measures used to summarize the distribution of income several measures of the distribution of health expectancy can be developed.

Determinants of Health Inequality

The measurement of health inequalities across countries is a crucial step towards a better understanding of its determinants. Once the performance of the health system has been assessed with respect to each goal, including the reduction of health inequalities, we intend to proceed with analytical work on the determinants of performance.

The study of the determinants of health and its distribution will involve those socioeconomic variables that are likely to play an important role. The relationship between socioeconomic status and health is a complex one and we would like to differentiate among the following four interactions, each of which is very important and needs to be better understood: 1) how the average level of socioeconomic status affects the average level of health in a population; 2) how the average level of socioeconomic status influences the distribution of health; 3) how the distribution of socioeconomic status affects the average health; and 4) how the distribution of socioeconomic status influences the distribution of health in a population.

Before making explicit policy recommendations we need to be better informed about how health systems of various countries perform in the achievement of their goals.

The framework presented above will provide us with the opportunity to study these relationships, since individuals have been used as the unit of analysis in the measurement of health and health inequality. Individual-level data provide us with a greater capacity to analyze this complex relationship than aggregate-level data. This approach has been criticized by those looking at differences in health status across social groups for ignoring the important relationship between socioeconomic status and health. On the contrary, our proposed measurement strategy will use socioeconomic status and its distribution as potential determinants of health inequality. When individuals are grouped by a socioeconomic variable and differences in health status across these groups are reported, the a priori assumption is made that the variable used to group individuals is the most important determinant of health inequality. Our approach does not make any a priori assumptions in the measurement of health inequality but uses potential determinants, including socioeconomic status, as explanatory variables. Both approaches, however, recognize that socioeconomic status and its distribution may be powerful determinants of health inequality.

The reduction of health inequalities is a key goal of health systems. Once the complex web of determination of these inequalities is better understood, there will be a pressing need for policies aimed at reducing them. The health system will play a central role in the formulation and implementation of these policies, whereby efforts directed specifically at health care institutions will have to be accompanied by initiatives involving other sectors.


1 The cost of being fully informed about the health consequences of different choices often is prohibitively high. Most individuals are forced to make choices with incomplete or incorrect information. When the choice to take on risk and the outcome are separated in time, the rate at which individuals discount the future can profoundly influence choices about health.