Critical Reflection

Health inequalities and social group differences:

what should we measure?
C.J.L. Murray,1 E.E. Gakidou,2 & J. Frenk3

Both health inequalities and social group health differences are important aspects of measuring population health.
Despite widespread recognition of their magnitude in many high- and low-income countries, there is considerable
debate about the meaning and measurement of health inequalities, social group health differences and inequities. The
lack of standard definitions, measurement strategies and indicators has and will continue to limit comparisons Ð
between and within countries, and over time Ð of health inequalities, and perhaps more importantly comparative
analyses of their determinants. Such comparative work, however, will be essential to find effective policies for
governments to reduce health inequalities. This article addresses the question of whether we should be measuring
health inequalities or social group health differences. To help clarify the strengths and weaknesses of these two
approaches, we review some of the major arguments for and against each of them.

Voir page 541 le reÂsume en francËais. En la paÂgina 541 figura un resumen en espanÄol.

Introduction lating effective policies, with which governments will

be able to reduce these inequalities.
Inequalities in health, both between and within We hope this article will contribute to the
populations, are a major public concern that demands increasing attention on social group health differ-
attention. For example, life expectancy at birth of ences, inequalities, and inequities by addressing the
native American males in some counties of the USA question of whether we should be measuring health
is 56 years, while that of Asian American women in inequalities or social group health differences. We use
other counties is above 95 years (1). The long- the term ``health inequalities'' to refer to composite
standing interest in health-related inequalities (2) has measures of the variation in health status across
increased since the early 1980s and includes health individuals in a population. Particular measures of
differences between social groups (3±6). Interest in health inequalities can reflect the range of variation
this subject has been expressed in the political arena from best to worst or the distribution of individuals
in the USA and Europe, as well as by WHO, and in within that range. This definition of the term ``health
several publications (7, 8). inequality'' has been used by other disciplines, such as
Despite wide recognition of the extent of the extensive study of income inequality across
health inequalities and social group health differences individuals (12, 13). For example, income inequality
in many high-income and low-income countries, is frequently measured using the Gini coefficient,
there is considerable debate about the meaning and which is a function of the distribution of individual
measurement of health inequalities, social group income. Measures of inequality of income or health
health differences, and inequities (9±11). The lack of are important because the same average level of
standard definitions, measurement strategies, and income or health could correspond to vastly different
indicators have limited and will continue to limit distributions of these variables across individuals in a
comparisons Ð between and within countries, and population. A concern for inequality is a concern for
over time Ð of health inequalities and, perhaps more the distribution of attributes such as income or health
importantly, comparative analyses of their determi- across individuals. In other words, average levels do
nants. Comparative studies are essential for formu- not convey sufficient information.
Social group health differences are considered
to be the differences across subgroups of the
1
Director, Evidence for Health Policy, World Health Organization, population, which may be based on biological, social,
1211 Geneva 27, Switzerland. Requests for reprints should be sent economic or geographical characteristics. Social
to this author.
2
group health indicators are summary measures of
Health Policy Analyst, Economics Advisory Service, World Health subgroups of the population, and as such they mask
Organization, Geneva, Switzerland.
3 part of the range of inequality present in the
Executive Director, Evidence and Information for Policy, World Health
Organization, Geneva, Switzerland. population. In much of the published literature,
health inequalities are taken to be synonymous with
Reprint No. BU0028

Bulletin of the World Health Organization, 1999, 77 (7) # World Health Organization 1999 537
Critical Reflection

social group differences in health. For example, Are health inequalities interesting?
Valkonen reports that the standardized mortality
ratio (SMR) for upper white-collar workers in Finland Even if we can measure health inequalities across
was 57 in 1986±90 as compared to 121 for blue-collar individuals, some authors argue that this is not
workers (14). Mackenbach & Kunst reviewed various intrinsically interesting (9±11, 20); ``...the main
summary indicators of health inequality, all of which problem is that such a measure [of individual
are indicators of the magnitude of social group health inequality] answers a different Ð possibly rather
differences (11). When social group health differ- uninteresting Ð question about generalized varia-
ences are equated with health inequality, the critical bility within a society distinct from systematic
choice is that of the variable used to distribute the variability based on social stratification within
population into social groups. Analytical traditions society'' (20). Variation across individuals in health
vary: in the United Kingdom, social groups have been can be attributed at the simplest level to four factors:
defined using five categories of occupation-based chance, genes, the environment (broadly defined to
social class; in some countries in continental Europe, include all physical and social factors), and the
educational attainment or occupational categories interaction between genes and the environment. The
have been used; and in the USA, most research argument that individual variation is uninteresting
focuses on social categories defined in terms of racial must rest on the claim that the components of
groups. Social groups defined by location deserve individual variation due to chance and perhaps genes
special note. When the geographical areas used to are not important or are without normative sig-
categorize the population are small and represent nificance. In the era of the human genome project,
relatively homogeneous groups, they provide a more the claim that health inequality due to differences
refined categorization of the population which can between the genetic endowment of individuals is
come close to revealing the extent of individual uninteresting seems hard to defend.
variation in the population. Should we be so uninterested in health inequal-
The term ``health inequities'' refers to the part ities due to chance that we do not include them in our
of the existing inequalities that are unjust according to measurement of health inequality? First, it is difficult
some theory of justice (15). It can legitimately be to justify the claim that we should not measure health
argued that the focus of health policy should be on inequality across populations because some compo-
reducing health inequities. In this critical reflection, nent of inequality is due to random rather than
however, we focus on the measurement of health systematic variation. Variation in the magnitude of
inequality and social group differences. health inequality across countries is unlikely to be due
There is intrinsic interest in both health to chance alone, so that the fact that some of the
inequalities and social group health differences. To variation is attributable to chance events is irrelevant to
help clarify the strengths and weaknesses of these the total assessment. Second, most of us are concerned
two approaches, we review some of the major with inequality of risk and inequality of outcome.
arguments for and against measuring health inequal- Consider the analogy to income. If the government
ities and social group health differences. Focusing on taxes every individual one dollar and then selects one
inequalities measured as the distribution across individual through a lottery to receive all the proceeds,
individuals has many advantages. First, measuring real income inequality would increase. But this is only
the distribution of health across individuals is the due to chance; the expectation of income is equally
natural complement to measuring the average level of distributed prior to the outcome of the lottery. In the
health in a population. Second, the individuals with case of income and health, risks and outcomes both
the worst levels of health can be identified without matter. If all differences across individuals are simply
choosing ex ante the variables used to define social due to chance, the degree of inequality across
groups. Third, comparisons of the degree of inequal- countries would be the same. The empirical results
ity between populations or for the same population on health inequality across populations will resolve this
over a period of time are straightforward. There are concern.
no concerns about comparability of groups or
changes in group composition. There will continue
to be a rich debate on the best summary indicator of Can health inequality across
the distribution of health across individuals, as there individuals be measured?
is for income (13). But once an indicator is chosen,
comparability of results is assured. Fourth, by For some commonly used health measures such as
separating the definition and measurement of annual mortality rates, inequality at the individual
inequality from ex ante causal hypotheses (see below) level is not very meaningful. Survival over one year is
or normative positions, inequality itself becomes an a dichotomous variable: individuals are either alive or
object of scientific inquiry. Despite these advantages, dead at the end of the year. The proportion of the
there are few studies of health inequalities, as defined population that is dead (the period death rate)
here (16±19). The paucity of studies may in part be contains all the information on the level and
due to the following arguments against measuring distribution across individuals. Because there is a
health inequalities or those in favour of measuring one-to-one correspondence between the proportion
social group differences. dead and the distribution of the population in the

538 Bulletin of the World Health Organization, 1999, 77 (7)

 Health inequalities and social group differences

categories dead and alive, the proportion dead is fully on health differences across social groups. A number
informative of level and distribution. Nevertheless, of arguments can be advanced to support studying
even if we know the distribution of the population in social group health differences instead of health
the categories dead and alive, there will probably be inequalities. These are reviewed below. Ultimately,
differences in mortality rates across subgroups of the the main challenge of the approach of measuring
population. Measuring social group differences in social group health differences is the choice of social
period mortality is thus an important adjunct to groups, and the profound problems of comparability
measuring the population death rate. and interpretation that this introduces.
As soon as mortality data are used to calculate a
continuous variable, survival time, such as in a cohort
life table, individual inequality contains important Are social group health differences
information that is not included in the average survival
time of the population. The same average survival or
synonymous with health inequalities
cohort life expectancy can occur with widely different because of their moral significance?
distributions of ages at death. Similarly, inequality of To many analysts working on social group health
individual healthy life expectancy for a cohort could be differences, it is almost axiomatic that certain types of
measured with the appropriate data. For all polycho- social categorizations are the best way to examine
tomous or continuous measurements of health (e.g. health inequalities because of their moral importance
SF-36, Euroquol, Activities of Daily Living or the (for discussion, see 4, 9, 10, 21). The argument is that
Health Utilities Index), health inequalities across social groups and health gradients according to these
individuals can be relatively easily measured. One groupings are more important because groups at the
popular measure, period life expectancy, is the survival bottom of the social gradient have disadvantages in
of a hypothetical birth cohort exposed to currently other spheres of well-being such as income, wealth or
observed mortality rates. Measurement of inequality education (see 15, for discussion). The concern about
across these hypothetical individuals in terms of the age health differences between social groups stems not
at death is possible (17), but difficult to interpret from the health differences themselves, but from
because of the hypothetical nature of a period life table. their covariance with other socioeconomic variables.
To make this normative argument in favour of
constructing health inequality as the preferred
Are measures of health method of measurement rather than social group
inequality sensitive to changes health differences, imagine two populations. In both
populations, the average levels of health and the
in socioeconomic status? degree of health inequality across individuals are
Some authors argue that ``it clearly is a defect [of identical. In population A, those with lower educa-
measures of individual inequality] if one takes the view tional attainment have worse health status than the
Ð as many do Ð that what is interesting Ð and more educated individuals; in population B, indivi-
indeed worrying Ð about inequalities in health is not duals with low educational attainment have better
that they exist, but that they mirror inequalities in health than the more educated. It is reasonable to
socioeconomic status'' (see 9, page 546). This is not a argue that health inequality in population A is of
positive but strictly a normative argument, albeit a greater concern and deserving of public attention
defensible normative argument (15). Wagstaff et al. than that in population B. But a cogent moral
argue that only those health inequalities that correlate argument could also be made that the inequality in A
with other socioeconomic inequalities are interest- and B are equally of concern. Regardless of the moral
ing (9). If health is a critical component of human well- position taken, it would seem strange to argue that
being, with which most would agree, one wonders why there is less health inequality by definition in B than in
inequality of health should not be considered A simply because of the correlation between health
intrinsically important, independent of its correlation levels and education levels. The argument that social
with other components of well-being. The parallel group health differences are the best approach to
argument for income, that income inequality is measuring health inequalities confounds a positive
interesting only to the extent that it correlates with issue, the extent of inequality across individuals, and a
health or education inequality, would not be seriously normative question: which inequalities are unjust?
considered. Rather, it is clearly a conceptual and
analytical strength to separate the measurement of
health inequality from normative claims on the types Is social position the fundamental
of health inequalities that are considered inequitable
and deserving of public action.
latent variable determining health?
Some researchers, who agree that health inequality
should not be measured in a normative manner, hold
Social group health differences that social group health differences may be the best
positive approach (21). One of the long-standing
Notwithstanding the advantages of measuring health debates in sociology refers to the dimension along
inequality across individuals, there is extensive work

Bulletin of the World Health Organization, 1999, 77 (7) 539

Critical Reflection

which individuals are differentiated into social distal socioeconomic factors, proximal individual
categories. In fact, such dimensions and the factors, and physiological factors was proposed by
corresponding descriptions (e.g. social class, stratum, Mosley & Chen (30). For high-income countries,
group, sector) are critical elements in defining the these causal webs will surely include important
major theoretical traditions in the social sciences. community-level characteristics, such as income
Whether or not they are conscious of the theoretical inequality or social networks that operate at the
implications, many studies in the health field utilize individual level (31±33). Analysis of social group
variables as indicators of underlying social constructs. differences may stimulate the search for causal
Such an approach has gained prominence in high- explanations through the complex webs of distal,
income countries; according to this view, absolute proximate and physiological determinants. This
deprivation is not a major determinant of health, even cogent reason for studying social group health
for the poor. What matters is an individual's relative differences highlights the continuing importance of
social position. For example, Mackenbach & Kunst measuring them for analytical reasons; it does not
refer to this as socioeconomic status (SES): ``SES qualify them as the best way to measure health
refers to an individual's relative position in the social inequalities.
hierarchy and can be operationalized as level of The biggest problem with this way of measur-
education, occupation and/or income'' (see 11, ing health inequalities is the choice of variable to
page 758). Relative social position in the language define social groups, and the subsequent inevitable
of measurement theory is a latent trait (22), which can problems of comparability across countries. A
only imperfectly be measured with a variety of growing literature proposes methods to enhance
proxies such as income, education, car ownership, comparability (e.g. 23) by standardizing summary
wealth or occupation. It is with reference to this social indicators of social group health differences and by
position hypothesis that comparisons of occupation- using the same variables to define social groups
based social class groupings in the United Kingdom across countries or, more often, by assuming that
and educational attainment-based groups in the different variables are in fact proxies for the same
Netherlands are justified (23). The notion that social underlying variable, i.e. social position. While these
position is a fundamental latent variable determining attempts at standardization and enhanced compar-
health is consistent with the Marxist tradition which ability are laudable, they will never be completely
defines social classes by their position in production satisfactory. Even if occupation-based social group
relationships (24, 25). health differences are larger in France than in the
Because social position is the key variable United Kingdom, there may always be some new
determining health, according to this view, defining variable that can be used to define other social
health inequality in terms of social group differences groupings in which differences are greater in the
is legitimate. There are two aspects to this approach United Kingdom than in France. One such variable is
that can be challenged. First, the hypothesis that a geographical location. In the USA, Murray et al.
latent variable Ð social position Ð exists and is the (1) revealed that the differences in life expectancy
key determinant of health is nearly impossible to between counties were much larger than differences
falsify. Any contradictory evidence can simply be across socioeconomic variables. Small area analysis
ascribed to the use of imperfect proxies for social may hold out the greatest promise for studying the
position in a given cultural or political context. extent to which social group health differences vary
Second, defining health inequality as the difference in across countries. Location can be defined in a
health status between social groups, with lower as culturally independent way for all countries, data sets
compared to higher social position, does not allow for are widely available on health by location, and
scientific inquiry into other key determinants of location provides a social categorization with many
health inequality across individuals. more categories. Even for small area analysis,
considerable problems remain in undertaking cross-
national comparisons of the extent of social group
Socioeconomic factors differences in health. Because the choice of variables
used to define social groups matters so much, it
and webs of causation would be highly desirable to define and measure
Much of the social epidemiology research on health health inequalities independent of any particular
inequalities in Europe and Latin America (26, 27) has social variable.
focused on the concept of social position, while in
North America epidemiologists and social demogra-
phers have looked at various socioeconomic factors Conclusions
as independent determinants of health operating
through a complex causal web (28, 29). A major Both health inequalities and social group health
analytical challenge is to define the causal pathways differences are important aspects of measuring
operating from distal socioeconomic factors to population health. In the face of the enormous
proximal individual behaviours, and ultimately phy- variation in health within populations, we cannot
siological factors. An example of such an approach simply focus on average levels of health. There are
for child mortality in developing countries linking convincing reasons to measure social group health

540 Bulletin of the World Health Organization, 1999, 77 (7)

 Health inequalities and social group differences

differences: they are normatively important; they inequality across individuals. By moving towards the
provide insights into causal pathways linking distal measurement of the distribution of health across
socioeconomic determinants and health; and they are individuals, the study of inequality will be put on a
relatively easy to measure. In fact, one particular sounder scientific footing. A shift towards measuring
approach to defining social groups, namely commu- health inequalities across individuals will undoubt-
nity location, has been much underutilized. Small area edly fuel a rich debate on the advantages and
analyses are possible using existing vital registration disadvantages of various summary indicators of the
data in many countries and are likely to reveal larger distribution of health. This debate can borrow from
health inequalities than previously recognized. How- the extensive literature on the measurement of
ever, health inequality should be defined in terms of income inequality. n

ReÂsumeÂ
IneÂgaliteÂs de sante ou diffeÂrences entre groupes sociaux : que devons nous mesurer ?
IneÂgaliteÂs de sante et diffeÂrences entre groupes sociaux individus entre ces limites extreÃmes. Les diffeÂrences
sont deux eÂleÂments importants de la mesure de l'eÂtat sanitaires entre groupes sociaux sont celles que l'on peut
sanitaire d'une population. Il existe dans les pays aÁ faible constater entre divers sous-groupes de la population,
ou aÁ haut revenu des ineÂgaliteÂs de sante et des sous-groupes dont les caracteÂristiques peuvent eÃtre de
diffeÂrences entre groupes sociaux dont l'importance est nature biologique, sociale, eÂconomique ou geÂogra-
largement reconnue, mais la signification et la mesure phique. Les indicateurs sanitaires relatifs aux groupes
des ineÂgaliteÂs ou des diffeÂrences, voire des iniquiteÂs, sur sociaux sont des mesures globales de l'eÂtat sanitaire des
le plan sanitaire entre groupes sociaux restent fortement divers sous-groupes de la population et en tant que tels,
sujettes aÁ controverse. L'absence de deÂfinitions norma- ils masquent en partie l'eÂventail des ineÂgaliteÂs qui existe
lise es, de strate gies de mesure et d'indicateurs aÁ l'inteÂrieur de la population.
approprie s limitent et continueront de limiter la Compte tenu des variations consideÂrables qui
comparaison Ð d'un pays aÁ l'autre, aÁ l'inteÂrieur d'un existent sur le plan sanitaire au sein d'une population, on
meÃme pays ou encore dans le temps Ð des ineÂgaliteÂs de ne peut pas se contenter de consideÂrer un niveau
sante et, ce qui est peut eÃtre plus grave, l'analyse sanitaire moyen. Il y a des arguments convaincants en
comparative des causes de ces ineÂgaliteÂs. Cette analyse faveur de la mesure des diffeÂrences sanitaires entre
comparative est pourtant essentielle pour les gouverne- groupes sociaux : elles sont importantes d'un point de
ments qui s'efforcent de trouver des politiques per- vue normatif; elles eÂclairent sur les chaõÃnes causales entre
mettant de reÂduire efficacement ces ineÂgaliteÂs. Nous deÂterminants socio-eÂconomiques lointains et sante et
souhaitons que cet article puisse contribuer aÁ attirer enfin elles sont faciles aÁ mesurer. De fait, il y a une
davantage l'attention sur les ineÂgaliteÂs, iniquiteÂs et meÂthode qui est treÁs insuffisamment utiliseÂe, aÁ savoir le
diffeÂrences de sante entre groupes sociaux en posant la lieu de reÂsidence des communauteÂs. Dans un grand
question de savoir s'il faut mesurer les ineÂgaliteÂs de sante nombre de pays, on peut proceÂder aÁ des analyses sur de
ou les diffeÂrences sanitaires entre groupes sociaux. petites zones en utilisant les donneÂes de l'eÂtat civil,
La mesure des ineÂgaliteÂs de sante et celle des analyses qui vont vraisemblablement re ve ler des
diffeÂrences sanitaires entre groupes sociaux sont toutes ineÂgaliteÂs sanitaires plus importantes qu'on ne le
deux intrinseÁquement inteÂressantes. Pour essayer de pensait. Il faut cependant deÂfinir les ineÂgaliteÂs de santeÂ
mieux voir les points forts et les points faibles de ces deux en fonction des ineÂgaliteÂs entre individus. En s'orientant
conceptions, nous examinons quelques-uns des princi- vers la mesure de la distribution de la sante entre
paux arguments qui sont avanceÂs pour ou contre la individus, l'eÂtude des ineÂgaliteÂs s'appuiera sur des bases
mesure des ineÂgaliteÂs de sante ou celle des diffeÂrences scientifiques plus solides. Une reÂorientation vers la
sanitaires entre groupes sociaux. Par mesure des mesure des ineÂgaliteÂs de sante entre individus ne peut
ineÂgaliteÂs de santeÂ, nous entendons la mesure composite manquer d'alimenter un deÂbat feÂcond sur les avantages
de la variation de l'eÂtat de sante d'un individu aÁ l'autre et les inconveÂnients que peuvent preÂsenter les divers
d'une population. Certaines mesures peuvent donner indicateurs globaux de la distribution de la santeÂ. Ce
une image de l'ampleur de la variation qui seÂpare le pire deÂbat peut se nourrir de la riche litteÂrature consacreÂe aÁ la
du meilleur ou rendre compte de la distribution des mesure des ineÂgaliteÂs de revenu.

Resumen
Desigualdades sanitarias y diferencias entre grupos sociales: ¿que debemos medir?
Las desigualdades en materia de salud y las diferencias hay gran controversia acerca del significado y la medicioÂn
sanitarias entre grupos sociales son aspectos importan- de las desigualdades de salud, las diferencias sanitarias
tes para medir la salud de la poblacioÂn. Aunque se entre grupos sociales y las situaciones de inequidad. La
reconoce en general que hay importantes desigualdades falta de definiciones normalizadas, de estrategias de
de salud y diferencias sanitarias entre grupos sociales en medicioÂn y de indicadores ha limitado y seguira limitando
muchos paõÂses, tanto de altos como de bajos ingresos, las comparaciones Ð entre los paõÂses y en los paõÂses a lo

Bulletin of the World Health Organization, 1999, 77 (7) 541

Critical Reflection

largo del tiempo Ð de las desigualdades sanitarias y, situacioÂn de subgrupos de poblacioÂn, y que por
algo tal vez maÂs importante, los anaÂlisis comparativos de consiguiente ocultan parte de las desigualdades
los determinantes de las desigualdades sanitarias. Ese existentes en la poblacioÂn.
anaÂlisis comparativo de las desigualdades en materia de Dadas las grandes diferencias de salud dentro de
salud, sin embargo, sera esencial para formular polõÂticas las poblaciones, no podemos contentarnos con determi-
eficaces que permitan a los gobiernos reducir tales nar los niveles medios de salud. Hay razones convincen-
desigualdades. Esperamos que este artõÂculo contribuya a tes para medir las diferencias sanitarias entre grupos
la creciente atencioÂn prestada a las diferencias de salud sociales: son normativamente importantes; arrojan luz
entre grupos sociales, las desigualdades y la falta de sobre las võÂas causales que enlazan los determinantes
equidad, para lo cual se aborda la cuestioÂn de si socioeconoÂmicos distales y la salud; y son relativamente
deberõÂamos medir las desigualdades sanitarias o las faÂciles de medir. De hecho, un meÂtodo de definicioÂn de
diferencias de salud entre grupos sociales. los grupos sociales como es la ubicacioÂn de las
Tanto unas como otras revisten un intereÂs comunidades ha sido claramente infrautilizado. Es
intrõÂnseco. Para ayudar a elucidar los puntos fuertes y posible realizar anaÂlisis de zonas reducidas a partir de
deÂbiles de esos dos enfoques, analizamos algunos de los los datos del registro civil en muchos paõÂses, y esos
principales argumentos a favor y en contra de la anaÂlisis pueden muy bien revelar desigualdades sanita-
medicioÂn de cada una de esas variables. AquõÂ utilizamos rias maÂs importantes de lo que se creõÂa. Sin embargo, las
la expresioÂn desigualdades sanitarias para designar las desigualdades sanitarias deberõÂan definirse desde el
medidas compuestas de las diferencias en el estado de punto de vista de las desigualdades interindividuales.
salud entre un individuo y otro en una poblacioÂn. Esas DesplazaÂndose hacia la medicioÂn de la distribucioÂn de la
medidas pueden reflejar el margen de diferencia entre el salud entre los individuos, el estudio de las desigualda-
mejor y el peor de los casos, o bien la distribucioÂn de los des dispondra de una base cientõÂfica maÂs soÂlida, y ese
individuos dentro de ese margen. Se consideran desplazamiento impulsara sin duda un enriquecedor
diferencias sanitarias entre grupos sociales las que se debate sobre las ventajas y los inconvenientes de los
dan entre subgrupos de la poblacioÂn, subgrupos que se diversos indicadores resumidos de la distribucioÂn de la
definen en funcioÂn de caracterõÂsticas bioloÂgicas, sociales, salud. Dicho debate puede beneficiarse de la extensa
econoÂmicas o geograÂficas. Los indicadores del estado de bibliografõ a disponible sobre la medicio n de las
salud de grupos sociales son medidas que sintetizan la desigualdades en materia de ingresos.

References
1. Murray CJL et al. U.S. patterns of mortality by county and race: 15. Marchand S et al. Class, health and justice. Milbank quarterly,
1965±1994. Cambridge, Harvard School of Public Health 1998, 76: 449±467.
and National Center for Disease Prevention and Health Promotion, 16. Deaton A, Paxtson C. Ageing and inequalities in income
1998. and health. American Economics Association Papers and
2. Antonovsky A. Social class, life expectancy and overall mortality. Proceedings, 1998, 88: 248±253.
Milbank Memorial Fund quarterly, 1967, 45 (2): 31±73. 17. Silber J. Health and inequality. Some applications of uncertainty
3. Marmot M et al. Social inequalities in health: next questions theory. Social science and medicine, 1982, 16 (19): 1663±1666.
and converging evidence. Social science and medicine, 1997, 18. Illsey R, LeGrand J. The measurement of inequality. In:
44: 901± 910. Williams A., ed. Health and economics. London, Macmillan, 1987.
4. Whitehead M The concepts and principles of equity and health. 19. LeGrand J. An international comparison on distributions of
International journal of health services, 1992, 22: 429± 445. ages-at-death. In: Fox J., ed. Health inequalities in European
5. Feinstein JS. The relationship between socioeconomic status countries. Adelshot, Hants, England, Gower Pub. Co., 1989.
and health: a review of the literature. Milbank quarterly, 1993, 20. Carr Hill R. Time trends in inequalities in health. Journal
71 (2): 279± 322. of biosocial science, 1988, 20: 265±273.
6. Beaglehole R, Bonita R. Public health at the crossroads: 21. Wilkinson RG. Socioeconomic determinants of health.
which way forward? Lancet, 1998, 351: 590± 592. Health inequalities: relative or absolute material standards?
7. Kaplan GA, Lynch JW. Whither studies on the socioeconomic British medical journal, 1997, 314: 591±595.
foundations of population health? American journal of public 22. Nunnally JC, Bernstein IH. Psychometric theory. New York,
health, 1997, 87: 1409±1411. McGraw-Hill, 1994.
8. Acheson D. Independent Inquiry into Inequalities in Health 23. Kunst AE et al. Mortality by occupational class among men
Report. London, HMSO, 1998. 30-64 years in 11 European countries. Social science and
9. Wagstaff A et al. On the measurement of inequalities in health. medicine, 1998, 46: 1459±1476.
Social science and medicine, 1991, 33: 545±557. 24. Navarro V. A historical review (1965±1997) of studies on class.
10. Manor O et al. Comparing measures of health inequality. health, and quality of life: a personal account. International journal
Social science and medicine, 1997, 45: 761±771. of health services, 1998, 28: 389±406.
11. Mackenbach JP, Kunst AE. Measuring the magnitude 25. Bronfman M, Tuiran R. La desigualdad social ante la muerte:
of socioeconomic inequalities in health: an overview of available clases sociales y mortalidad en la ninÄez. [Social inequality in death:
measures illustrated with two examples from Europe. Social social classes and mortality in childhood]. Cuadernos medico
science and medicine, 1997, 44: 757±771. sociales, 1984, 29±30: 53±75 (in Spanish).
12. Atkinson AB. On the measurement of inequality. 26. Laurell AC et al. Participatory research on workers' health.
Journal of economic theory, 1970, 2: 244±263. Social science and medicine, 1992, 34: 603±613.
13. Sen A. On economic inequality. Oxford, Clarendon Press, 1997. 27. Bronfman M. Infant mortality and crisis in Mexico.
14. Valkonen T. Problems in the measurement and international International journal of health services, 1992, 22: 157±167.
comparisons of socioeconomic differences in mortality. Social
science and medicine, 1993, 36: 409±418.

542 Bulletin of the World Health Organization, 1999, 77 (7)

 Health inequalities and social group differences

28. Hummer RA, Rogers RG, Eberstein IW. Sociodemographic 31. Kennedy BP et al. Income distribution, socioeconomic status,
differentials in adult mortality: a review of analytic approaches. and self-rated health in the United States: a multilevel analysis.
Population and development review, 1998, 24: 553±578. British medical journal, 1998, 317: 917±921.
29. Preston SH, Taubman P. Socioeconomic differences in 32. Geronimus AT et al. Excess mortality among blacks and whites
adult mortality and health status. In: Martin LG, Preston SH, in the United States. New England journal of medicine, 1996,
eds. Demography of aging. Washington, DC, National Academy 335: 1552±1558.
Press, 1994. 33. Anderson RT et al. Mortality effects of community
30. Mosley WH, Chen LC. An analytical framework for the study socioeconomic status. Epidemiology, 1997, 8: 42±47.
of child survival in developing countries. Population and
development review, 1984, 10 (Suppl.): 22±45.

Bulletin of the World Health Organization, 1999, 77 (7) 543