Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=uajb20
Download by: [Masarykova Univerzita v Brne] Date: 19 February 2016, At: 02:39
The American Journal of Bioethics
ISSN: 1526-5161 (Print) 1536-0075 (Online) Journal homepage: http://www.tandfonline.com/loi/uajb20
The Social Determinants of Health: Why Should We
Care?
Adina Preda & Kristin Voigt
To cite this article: Adina Preda & Kristin Voigt (2015) The Social Determinants of
Health: Why Should We Care?, The American Journal of Bioethics, 15:3, 25-36, DOI:
10.1080/15265161.2014.998374
To link to this article: http://dx.doi.org/10.1080/15265161.2014.998374
Published online: 18 Mar 2015.
Submit your article to this journal
Article views: 2265
View related articles
View Crossmark data
Citing articles: 13 View citing articles
Target Article
The Social Determinants of Health: Why
Should We Care?
Adina Preda, University of Limerick
Kristin Voigt, McGill University
A growing body of empirical research examines the effects of the so-called “social determinants of health” (SDH) on health and
health inequalities. Several high-profile publications have issued policy recommendations to reduce health inequalities based
on a specific interpretation of this empirical research as well as a set of normative assumptions. This article questions the
framework defined by these assumptions by focusing on two issues: first, the normative judgments about the (un)fairness of
particular health inequalities; and second, the policy recommendations issued on this basis. We argue that the normative
underpinnings of the approach are insufficiently supported and that the policy recommendations do not necessarily follow
from the arguments provided. Furthermore, while many of the policies recommended—such as improving people’s living
conditions and reducing inequalities in wealth and power—are justified in their own right, the way these recommendations are
tied to health is problematic.
Keywords: fairness, health equity, health inequality, health policy, social determinants of health
There is a growing body of empirical research into
what has become known as the “social determinants of
health” (SDH). The central claim arising from this body
of research is that various social factors have a strong
influence on population health and on inequalities in
health outcomes across social groups. Much of the
attention the SDH framework has received from
researchers outside the field and the general public has
focused on a number of high-profile reports, as well as
articles in major scientific journals summarizing these
reports. This work includes most recently and prominently
Michael Marmot’s report on health inequalities
in the United Kingdom (Marmot 2010), and the work of
the Commission on the Social Determinants of Health
(Commission on the Social Determinants of Health
2008) and the World Health Organization (WHO) report
on the health divide in Europe (WHO 2014), both
chaired by Marmot. These reports put forward a number
of policy recommendations based on a specific
interpretation of the empirical findings in the epidemiological
literature. The recommendations are also based
on a number of normative assumptions pertaining primarily
to the injustice of social inequalities in health as
well as empirical assumptions about the best way to
address them.
In this article, we critically examine the normative
underpinnings that lead to these recommendations. Our
concern is with a certain framework that we call the
“health equity through social change model” (HESC),
which encapsulates a certain way of thinking about the
relationship between social factors and health. This is
primarily reflected in the prominent work of Michael
Marmot but is also adopted, albeit sometimes only in
part and/or only implicitly, by many other social epidemiologists.
While this model has been highly influential
both within epidemiology and outside the field, its
assumptions, especially its normative assumptions, are
rarely examined.
We should make clear from the start that we are not
unsympathetic to many of the conclusions and recommendations
presented in this body of work. In particular,
we agree that many of the policy recommendations—
such as improvements to people’s living conditions
and reductions of inequalities in wealth and power—
are required as a matter of social justice. However, the
way these recommendations are tied to health and
health equality in the HESC model is problematic. In
this article, we focus on two issues: first, the (sometimes
only implicit) normative judgments and assumptions
about the (un)fairness of particular health
inequalities; and second, the policy recommendations
issued on this basis. We argue that the normative
underpinnings of the HESC model are not sufficiently
supported and that the policy recommendations do not
necessarily follow from the arguments provided and
may be inconsistent.
Address correspondence to Kristin Voigt, McGill University, 1130 Av Des Pins O, Montreal, Quebec, H3A 1A3, Canada. E-mail: kristin.
voigt@mcgill.ca
ajob 25
The American Journal of Bioethics, 15(3): 25–36, 2015
Copyright © Taylor & Francis Group, LLC
ISSN: 1526-5161 print / 1536-0075 online
DOI: 10.1080/15265161.2014.998374
Downloadedby[MasarykovaUniverzitavBrne]at02:3919February2016
We begin by summarizing the main claims of the
HESC model in the second section. The third section
criticizes the account of health equity underlying this
model. The fourth section questions the move from these
normative judgments to the policy recommendations
issued. The fifth section summarizes our conclusions and
suggests possible ways in which future research could
help support the normative conclusions the HESC model
seeks to reach.
THE “HESC MODEL”
This section first outlines the main empirical findings in
the epidemiological literature that identify links between
social factors and health outcomes. We then identify and
clarify the central normative claims advocates of the HESC
model make, which complement the empirical findings.
Finally, we sketch the main policy recommendations that
are a crucial aspect of the HESC model and summarize the
assumptions we are concerned with.
Concepts and Empirical Findings
What, exactly, are the SDH and how can they be distinguished
from what we might call “nonsocial” determinants
of health? A wide range of factors appear to fall
within the definition of “social determinants of health.”
According to the WHO, the social determinants “are the
circumstances in which people are born, grow up, live,
work and age, and the systems put in place to deal with illness.
These circumstances are in turn shaped by a wider
set of forces: economics, social policies, and politics.”1
Definitions
of the SDH often explicitly exclude the health care
system. For example, Gopal Sreenivasan (2008) defines a
social determinant as “a socially controllable factor outside
the traditional health care system that is an independent partial
cause of an individual’s health status” (emphasis
added). Some of the SDH literature, by contrast, includes
health care as a social determinant. For example, the Commission
on the Social Determinants of Health (CSDH)
report notes that the “health-care system is itself a social
determinant of health, influenced by and influencing the
effect of other social determinants” (CSDH 2008, 8).
Diverse empirical studies conclude that there is a correlation
between socioeconomic status (SES) and/or its
various components on the one hand and health outcomes
(either in terms of an aggregate measure or with respect to
specific conditions) on the other (for a review and assessment
of the available evidence see Braveman, Egerter, and
Williams 2011c). Perhaps the most dramatic figures cited
in the SDH literature relate to differences in life expectancy
across different countries. For example, Marmot and
colleagues (2008, 1661) contrast life expectancies of more
than 80 years in countries such as Japan or Sweden with
that of less than 50 years in many African countries. Even
within individual, high-income countries, discrepancies in
life expectancies between different socioeconomic groups
can reach similar magnitudes. In Glasgow, Scotland (UK),
for example, male life expectancy is as low as 54 years in
the most deprived areas but 82 years in the most affluent
(Marmot 2007, 1153). Importantly, it is not only the case
that deprivation or poverty negatively affects health;
rather, health outcomes are correlated with socioeconomic
status across the entire socioeconomic spectrum, with stepwise
improvements in health outcomes as socioeconomic
status increases, even above levels where poverty or deprivation
could plausibly play a role. This is referred to as the
“social gradient” in health.
Correlations do not, of course, demonstrate causal connections.
While scientific papers on the SDH acknowledge
the difficulties in establishing causal connections between
different variables and are careful to describe the limitations
of the empirical evidence, proponents of the HESC model
typically portray these causal connections as straightforward.
For example, Marmot and colleagues state that
The poor health of poor people, the social gradient in health
within countries, and the substantial health inequities
between countries are caused by the unequal distribution of
power, income, goods, and services, globally and nationally.
. . . Together, the structural determinants and conditions of
daily life constitute the social determinants of health and cause
much of the health inequity between and within countries.
(Marmot et al. 2008, 1661, emphasis added)
Our aim in this article is not to dispute the existence of
a causal connection.2
The empirical literature supplements
findings about the correlations of various social factors
and health outcomes with studies that examine possible
causal pathways to demonstrate plausible causal links running
from social factors to health outcomes. For example,
the role of (lack of) job control and stress as contributing
factors in creating health inequalities is highlighted (Sreenivasan
2008). Thus, we will not dispute that social factors
can be seen as the “causes of the causes” (Marmot 2013,
289) of health inequalities—even if this is, in Marmot’s
(2013, 289) words, a “dubious concept philosophically.” At
the same time, it is important to bear in mind the complexities
surrounding the causal connections between particular
social factors and specific health outcomes, and the different
pathways that may be at work. The policy reports we
examine tend to deemphasize these complexities, which—
as we argue later, in the fourth section —becomes particularly
problematic when we consider policy strategies to
address health inequalities.
Normative Claims
It is not uncommon among social epidemiologists
to make certain (implicit) assumptions about the
1. http://www.who.int/social_determinants/thecommission/
finalreport/key_concepts/en
2. For a critique of the causal claims made in this model, see
Chandra and Vogl (2010).
American Journal of Bioethics
26 ajob March, Volume 15, Number 3, 2015
Downloadedby[MasarykovaUniverzitavBrne]at02:3919February2016
unfairness of health inequalities (see Mackenbach 2012,
767). While these may not be accepted by all epidemiologists
and are not always explicit, the policy reports we
focus on in this article, and some of the journal articles
summarizing them, do make explicit normative claims.
For example, the report of the Commission on the SDH
clearly states that
Where systematic differences in health are judged to be avoidable
by reasonable action they are, quite simply, unfair. It is
this that we label health inequity. Putting right these inequities—the
huge and remediable differences in health between
and within countries—is a matter of social justice. Reducing
health inequities is, for the Commission on Social Determinants
of Health . . . an ethical imperative. Social injustice is
killing people on a grand scale. (Commission on the Social
Determinants of Health 2008)
Thus, the normative judgment that is central to the
HESC model is that the health inequalities we find both
within countries and at the global level must—to at least
some degree—be considered unfair. Hence the reduction
of these health inequalities is required, not simply because
we want to improve health overall or because we want to
avoid the burdens associated with poor health, but rather
as a matter of social justice. We examine these claims in the
third section.
Policy Recommendations
The reports also issue a variety of policy recommendations.
For example, the recommendations offered by the
Marmot Review on inequalities in the United Kingdom
are: (1) Give every child the best start in life; (2) enable
all children, young people, and adults to maximize
their capabilities and have control over their lives; (3)
create fair employment and good work for all; (4) create
and develop healthy and sustainable places and communities;
and (5) strengthen the role and impact of ill
health prevention (Marmot 2010). The common thread
in these reports is the idea that inequalities in health
should be addressed by broad policy strategies that can
tackle social inequalities: “The implication we drew
from the gradient is that action to reduce inequalities in
health has to be across the whole society, not simply to
reduce poverty—universalist solutions are needed, not
targeted ones” (Marmot 2013, 287). While the more
recent report on the health divide and the SDH in
Europe acknowledges that some targeted measures are
also needed, it again stresses that they are not sufficient
for addressing health inequities:
One response to addressing health inequities open to all is to
ensure universal coverage of health care. Another is to focus
on behavior—smoking, diet and alcohol—that cause much of
these health inequities but are also socially determined. The
review endorses both these responses. But the review recommendations
extend further—to the causes of the causes: the
conditions in which people are born, grow, live, work and age
and inequities in power, money and resources that give rise to
them. (WHO 2014, xiii)
So the assumption here is that, if we are to address
social inequalities in health, we must go beyond medical
and behavioral approaches and include broader strategies
that redistribute income and wealth and make societies
more egalitarian. This assumption is problematic, as we
argue in the fourth section.
To summarize this section, we identified three main
sets of assumptions as central to the model we are
examining:
1. A set of empirical assumptions that identify socioeconomic
circumstances as the ultimate causes—or “causes
of the causes”—of health outcomes and inequalities.
2. A set of normative assumptions to the effect that social
inequalities in health are unfair and therefore must be
rectified as a matter of social justice.
3. Finally, the recommendation that the best and possibly
most effective way of redressing health inequalities is
by wide-ranging societal changes, most notably policies
that address differences in socioeconomic status (i.e.,
inequalities in the distribution of the SDH).
The last two sets of assumptions and therefore the
HESC model can be summarized in the heading used in
the WHO report on health inequalities in the European
Union (EU): “Health inequalities that are avoidable are
unjust: action is required across society” (WHO 2014, xiv).
In what follows, we unpack and examine these
assumptions.
HEALTH INEQUALITIES: WHAT ARE THEY AND WHEN
ARE THEY UNFAIR?
The normative assumptions about health inequalities are
central to the model. The literature on health inequality
distinguishes between those health inequalities that are
problematic (unfair, unjust) and those that are not. Often,
this is described in terms of the distinction between health
inequality (which captures all health inequalities) and
health inequity (which captures those health inequalities
that are unfair): “Health inequality is the generic term
used to designate differences, variations, and disparities in
the health achievements of individuals and groups . . .
health inequality is a descriptive term that need not imply
moral judgment” (Kawachi et al. 2002, 647). In this section,
we discuss how these distinctions are drawn in the HESC
model, so as to make explicit some of the normative
assumptions informing the framework and to highlight
possible tensions in the model.
Health Inequality: Variation Across Individuals or
Social Groups?
Note, first, that when discussing health inequalities, epidemiologists
working within the SDH framework
Caring About Social Determinants of Health
March, Volume 15, Number 3, 2015 ajob 27
Downloadedby[MasarykovaUniverzitavBrne]at02:3919February2016
generally envisage health inequalities as differences in
average health outcomes between different social groups,
such as socioeconomic or ethnic groups. There may be
good reasons for adopting this approach, but it is important
to recognize that this is not the only way to understand
and measure health inequality. An alternative to
comparing average outcomes across social groups—also
mentioned in Kawachi and colleagues’ definition just
quoted—is to measure variation in health outcomes across
individuals in a particular population. This understanding
of health inequality was employed in the 2000 World
Health Report, following an argument made by Murray,
Gakidou, and Frenk (1999). The inclusion of individual
measures was motivated in part by perceived weaknesses
of group-based measures: Such measures average out outcomes
for individuals within groups, thereby risking the
loss of relevant information and obscuring outcomes for
individuals, who are ultimately the locus of moral concern
(Murray et al. 1999). At the same time, this approach was
criticized sharply by epidemiologists who considered the
shift from social group to individual differences to
“effectively remove . . . equity and human rights from the
public health monitoring agenda” (Braveman, Starfield,
and Geiger 2001, 679). Similarly, Marmot argues that the
focus should be on group differences because it is these
differences that matter from a normative perspective:
The causes of individual differences and the causes of
group differences, then, may not be the same. In the
thought experiment of equalizing all relevant environmental
conditions, there would still be individual differences
in health. . . . It is social inequalities in health that exercise
me and, if avoidable, that I label as unfair. (Marmot 2013,
287–288, emphasis added)
While we do not take a position on this debate, it is
important to highlight that this choice is not normatively
neutral. The normative commitments underlying this
methodological approach are rarely discussed, or even
acknowledged, in the literature.3
Avoidability, Amenability, and the Distinction between
Social and Natural Health Inequalities
How, then, should we decide when health inequalities
between different social groups should be considered
unfair or, to use Kawachi and colleagues’ distinction introduced
earlier, how do we determine when a health
inequality should also be considered a health inequity?
One substantive account of health inequity has been proposed
by Margaret Whitehead. Her account features prominently
not only in the HESC model but also in other,
related accounts (e.g., Braveman et al. 2011b; Braveman
and Gruskin 2003). Whitehead’s proposed definition is:
“Equity in health implies that ideally everyone should
have a fair opportunity to attain their full health potential
and, more pragmatically, that no one should be disadvantaged
from achieving this potential, if it can be avoided”
(Whitehead 1990, 7). She further claims that “in order to
describe a certain situation as inequitable, the cause has to
be examined and judged to be unfair in the context of
what is going on in the rest of society” (Whitehead 1990, 7).
She identifies seven main determinants of health
(inequalities):
1. Natural, biological variation.
2. Health-damaging behavior if freely chosen, such as participation
in certain sports and pastimes.
3. The transient health advantage of one group over
another when that group is first to adopt a health-promoting
behavior (as long as other groups have the
means to catch up fairly soon).
4. Health-damaging behavior where the degree of choice
of lifestyle is severely restricted.
5. Exposure to unhealthy, stressful living and working
conditions.
6. Inadequate access to essential health and other public
services.
7. Natural selection or health-related social mobility
involving the tendency for sick people to move down
the social scale.
According to Whitehead, “the consensus view from the
literature” is that factors in categories 1, 2, and 3 would not
normally lead to health inequities; health inequalities
resulting from determinants 4, 5, 6, and 7, however, are
unfair and can therefore be described as “health
inequities” (Whitehead 1991, 219).
What then is the criterion for distinguishing between
these causes? Whitehead clarifies that it is “avoidability”
that is relevant in distinguishing the first three causes
from the last four. Similarly, Marmot and colleagues
explain that “if systematic differences in health for different
groups of people are avoidable by reasonable
action, their existence is, quite simply, unfair. We call
this imbalance health inequity” (Marmot et al. 2008,
1661). This view is not restricted to Whitehead and Marmot;
avoidability is regarded as crucial in other accounts
of health equity, such as the one proposed by Braveman
and colleagues (2011b).
Thus, in the HESC model, “avoidability” appears as
both a necessary and a sufficient condition for health
inequity. In other words, proponents of the HESC model
maintain that only and all those health inequalities that are
avoidable are inequities. We argue that “avoidability,” as
understood in this framework, is neither necessary nor
sufficient.
In order to clarify how “avoidability” is understood in
the model, we should note that it is meant to draw a line
between “natural” and “social” health inequalities. Whitehead
explains that the natural variation between individuals
with respect to health outcomes is unavoidable and
hence not inequitable. Thus, we must focus on redressing
3. For further discussion, see Hausman (2007, 2013), Kawachi
et al. (2002) and Lippert-Rasmussen (2013).
28 ajob March, Volume 15, Number 3, 2015
American Journal of Bioethics
Downloadedby[MasarykovaUniverzitavBrne]at02:3919February2016
that portion of health inequalities that can be attributed to
social causes.4
The need to address health inequalities that have
social causes also motivates the focus on inequalities
between social groups rather than across individuals
that we discussed earlier: The model assumes that what
happens across social groups must have social causes,
whereas differences within those groups reflect natural,
biological variation (Whitehead 1990, 6). Similarly, as
Marmot (2013, 287–288) explains in the passage quoted
earlier:
The causes of individual differences and the causes of group
differences . . . may not be the same. In the thought experiment
of equalizing all relevant environmental conditions,
there would still be individual differences in health. These
may claim attention, both from geneticists and from those
who sought to avert genetic destiny by improving the lives of
people despite their inheritance, but for clarity I would not
use the term “inequality” to describe these genetic differences
among individuals. I would not think of the genetic lottery as
unfair or unjust.
It is worth noting, however, that inequalities in average
health outcomes between different social groups are
not necessarily caused by social factors. Differences
between social groups may also have nonsocial causes.
An important example of differences between social
groups that seem to result, at least in part, from nonsocial
factors is that of gender. In almost all countries of
the world, women have longer life expectancies than
men (although the size of this difference varies). But
according to the HESC model, to the extent that this difference
has biological rather than social causes, it must
be regarded as fair:
Women, in general, live longer than men. This difference is
likely due to biological sex differences, and is not, therefore,
inequitable. However, in cases where women have the same
or lower life expectancy as men—that is, where social conditions
act to reduce their apparently natural longevity advantage—inequality
is a mark of inequity. The injustice that the
Commission seeks to address comes from failure to achieve
levels of health that, but for lack of action, should be attainable.
(Marmot 2007, 1155)5
This account of health equity thus presupposes (1) that
only avoidable health inequalities can be unfair and (2)
that only socially caused inequalities are avoidable. It thus
follows that only social inequalities in health are unfair.
Both these assumptions are questionable, as we now
explain.
First, consider the requirement of avoidability. We
should note that an inequality can be avoided in two ways:
It can either be prevented from occurring, or it can be
redressed. Thus, avoidability can be understood either as
preventability or as amenability to intervention. In assuming
that natural inequalities are not avoidable, this account
endorses the first sense of “avoidability.” But if we examine
the reasons why avoidability might be considered a
necessary condition for unfairness, this conceptualization
seems inappropriate.
A possible reason why avoidability is relevant is the
idea that for an outcome to be unfair, it must be possible to
alter that outcome.6
Judging something as unfair implies
that it ought to be changed, which in turn implies that it
can be changed. But if this is the thought, it does not follow
that it is preventability that matters. Instead, we might
think, it is amenability to change that matters, that is,
whether or not an inequality can be redressed.7
We cannot
prevent the rain from falling but we can address (at least
some of) its negative effects. Thus, we must distinguish
more clearly between avoidability and amenability, where
the first refers strictly to preventability and the latter to
whether or not an inequality is amenable to intervention.
4. A similar approach is taken by Kawachi and colleagues (2002),
who note that health inequalities due to “pure chance (for example,
a random genetic mutation—unlucky but not unjust)” should
not be considered inequitable. The idea that health inequities
must have social causes is explicitly rejected by some contributors
to the SDH debate. For Braveman and colleagues (2011b), for
example, what matters is that “health disparities are systematically
linked with social disadvantage, and may reflect social disadvantage,
although a causal link does not need to be demonstrated.
Whether or not a causal link exists, health disparities adversely
affect groups who are already disadvantaged socially, putting
them at further disadvantage with respect to their health. . . . This
reinforcement or compounding of social disadvantage is what
makes health disparities relevant to social justice even when
knowledge of their causation is lacking.”
5. In other accounts, longer life expectancy is not regarded as a
matter of inequity because men, as a group, are more advantaged
in nonhealth areas. Braveman and colleagues (2011b), for example,
note that “shorter life expectancy among men in general, if likely
avoidable, would clearly be an issue of public health importance
based on the magnitude of potential population impact. However,
men as a group have more wealth, influence, and prestige, so this
difference would not be a social injustice and, therefore, not a health
disparity or equity issue” (S1, emphasis added)..
6. Note that this is different from the thought, famously attributed
to John Rawls, that natural inequalities are neither fair nor unfair
(Rawls 1999, 87). This thought does not make avoidability a criterion
for fairness; rather, it points to a more general framework
according to which justice or fairness are concepts that can only be
applied to social institutions. It is unclear which of these ideas is
envisaged in the HESC model.
7. While, as we noted earlier, avoidability features in a number of
definitions of health inequity, other definitions of health inequity
seem to stipulate criteria closer to amenability. For example, the
International Society for Equity in Health defines equity as “the
absence of potentially remediable, systematic differences in one or
more aspects of health across socially, economically, demographically,
or geographically defined population groups or subgroups”
(cited in Macinko and Starfield, 2002). Remediability is also considered
a criterion for health inequity by Starfield (2001).
March, Volume 15, Number 3, 2015 ajob 29
Caring About Social Determinants of Health
Downloadedby[MasarykovaUniverzitavBrne]at02:3919February2016
It is crucial to be clear about which of these two notions
is relevant because many unpreventable inequalities are in
fact amenable to intervention. In particular, medical interventions
can often redress natural inequalities. More generally,
even when medical treatments are not available, to
the extent that we can compensate for inequalities (including
those that were unavoidable and cannot be mitigated
after the fact), they are, in that sense, amenable to interven-
tion.8
Thus, by making avoidability rather than amenability
(either through direct mitigation of poor health or the
provision of compensation) a necessary criterion for
unfairness, a number of health inequalities that might otherwise
have been considered unfair, such as natural
inequalities, are identified as unproblematic. If it is the
ability to change inequalities that is relevant in drawing
the line between fair and unfair health inequalities, there is
no reason why only social, but not natural, inequalities
should be seen as unfair.
In fact, the account endorsed by Whitehead would suggest
that the reason why social inequalities in health are
unfair is that they are not voluntary: “where people have
little or no choice in living and working conditions, the
resulting health differences are more likely to be considered
unjust than those resulting from health risks which
were voluntarily chosen” (Whitehead 1991, 220). But if
what makes inequalities unfair is the lack of control or
choice over their source, then there is every reason to
regard natural inequalities as unfair as well. In other
words, the same reason that makes social inequalities in
health unfair also makes natural inequalities unfair. But
then “preventability” cannot be a necessary condition for
unfairness.
Nevertheless, while it is plausible to regard
“amenability” as a necessary condition for unfairness,
even that is debatable. Some philosophers would argue
that it is possible to consider an inequality unfair even if
there is nothing that can be done about it. Furthermore, the
account of health (in)equity underlying the HESC model
suggests that avoidability is a sufficient condition; in other
words, all inequalities in health that can be avoided, that
is, prevented, are inequities. However, it is not clear why
this should be so. Even if we accept that avoidability is a
necessary condition for unfairness for the reasons mentioned
earlier, we need a further reason to accept that all
avoidable inequalities are unfair: The fact that something
can be done about them is not enough to indicate that it
should be done.
A possible explanation for the view that avoidability is
a sufficient condition for unfairness is an implicit assumption
that it is unfair that we, as a society, prevent people
from attaining the level of health that they could otherwise
attain. But it is not clear that this thought can be supported
with arguments. Why is it unfair or at least problematic if
people’s position in society is reflected in their health outcomes?
We examine some possible answers to this question
in the next section.
Socioeconomic Inequalities, Health Inequalities, and
Residual Inequalities
As we have highlighted, an important assumption of the
model is that all health inequalities that result from social
inequalities are unfair; it is less clear, however, on what
grounds we should come to this conclusion. There are
two possibilities, both of which are at times suggested by
the HESC model: First, we may judge the distribution of
the SDH to be unfair for independent reasons, and it is
this unfairness that makes any resulting inequalities in
health unfair as well or, second, the distribution of the
SDH may be considered unfair because it leads to health
inequalities. This second view presupposes that health
inequalities are unfair in themselves. Both possibilities
raise problems.
The first line of reasoning seems implicit in parts of the
CSDH report, which notes that the “unequal distribution
of health-damaging experiences is not in any sense a
‘natural’ phenomenon but is the result of a toxic combination
of poor social policies and programmes, unfair economic
arrangements, and bad politics” (1, emphasis added). However,
if we agree that inequalities in the distribution of
social factors such as income or status are unjust for independent
reasons—as we do—it is not clear why we would
focus on health inequality rather than social inequality
more broadly. Social inequalities ought to be redressed
because (social) justice requires it, rather than because of
their effects on health. We do not deny that showing the
effects of social inequalities on health may strengthen the
argument for redistribution; however, this cannot be put
forward as the main reason for such redistribution unless
it can also be argued that inequalities in health are problematic
or unjust in themselves.
This is the claim that we now examine. There is some
evidence that the model puts forward this claim, albeit
without much supporting argument. For example, Marmot
states: “My position in the public debate is that, as a doctor,
I regard as unfair health inequalities that could be
avoidable by reasonable means. Therefore I regard as
unfair policies that exacerbate avoidable health inequalities”
(Marmot 2013, 284).
But is there any reason to claim that health inequalities
are unfair when they result from a fair albeit
unequal distribution of social goods? The literature
refers to such health inequalities as “residual inequalities.”
A “residual inequality” is “an avoidable inequality
in health the causes of which are otherwise fair”
(Sreenivasan 2009, 245). Is there any theoretical framework
that can support the claim that these inequalities
are unjust? There are very few theories of justice that
8. Of course, neither medical treatment nor compensation may be
able to fully redress the harm someone suffers because of a health
condition so in that sense it may not be fully avoidable. However,
there is often partial redress that can be provided after the poor
health outcome has occurred, even if it could not have been
avoided.
30 ajob March, Volume 15, Number 3, 2015
American Journal of Bioethics
Downloadedby[MasarykovaUniverzitavBrne]at02:3919February2016
have addressed this issue, and that may be precisely
because a theory of social justice may not have much to
say about residual inequalities. Theories of justice are
usually concerned with “overall” inequalities, that is,
inequalities in the distribution of an overall good, such
as well-being, or a package of goods, such as primary
goods or even capabilities. As long as the total bundle
is equally and/or fairly distributed, there seems to be
no reason to be concerned with inequalities in specific
goods, such as health.
The two major types of theory that have considered
this question—Rawlsian and luck egalitarian accounts—
imply (and their proponents even state explicitly) that
residual health inequalities are not unfair. According to
Daniels, whose work develops a Rawlsian approach to
health inequality, “a health distribution is unjust when it
derives from an unjust distribution of the socially controllable
factors affecting population health and its distribution”
(Daniels 2008, 27). This means that decisions about
the fairness or unfairness of health inequalities depend on
a prior normative judgment about the distribution of social
determinants from which they result. If the latter are fairly
distributed, the former are of no independent moral
concern.9
According to the luck egalitarian view defended by
Shlomi Segall (2009), health inequalities are problematic if
they do not appropriately reflect people’s choices or effort.
The question is then whether the health inequalities resulting
from social factors reflect people’s choices or effort. On
one interpretation, they do if they are the direct (causal)
consequences of the distribution of SES and if this distribution
reflects people’s free choices. Thus, luck egalitarianism
could be seen to imply that social inequalities in health
are not necessarily unfair (Segall personal communication).
Thus, the concern with social health inequalities that
motivates the HESC model does not find much support in
the existing philosophical literature.10
However, we think that there may actually be scope for
supporting such a claim within a luck egalitarian framework.
Although it has not been, to our knowledge, developed
in the literature, we could envisage an argument to
the effect that inequalities in health are of concern even
when they result from a fair distribution of the social determinants.
This argument challenges the assumption mentioned
earlier, namely, that people are responsible for all
the consequences of their choices. In other words, it challenges
the idea that people are responsible for their health
status in virtue of being responsible for their socioeconomic
position, which has shaped their health outcomes.
But this would be a complex and not uncontroversial argument,
which would require a closer analysis of the causal
pathways leading from social inequalities to health
inequalities, as well as a consistent account of individual
responsibility that may be in tension with the account
assumed in the HESC model. We discuss these assumptions
about responsibility in the next section. The main
aim of this section is to highlight that the proponents of the
model should clarify how, if at all, the unfairness of health
inequalities relates to the distribution of the SDH.
Responsibility and Individual Bealth Behavior
If health inequalities are not unfair in themselves and they
are not unfair in virtue of resulting from an unfair distribution
of the SDH, could there be any reason for calling for
measures that would address health inequalities via a
change in the distribution of the SDH? In other words,
what is the rationale for trying to reduce the “social
gradient” in health if the social distribution is just? As we
suggested in the previous section, one such reason could
be that the distribution of health outcomes is unequal
because it does not reflect people’s choices about health.
This broadly luck egalitarian type of argument finds some
support in the HESC model. We already noted earlier
Whitehead’s claim that “where people have little or no
choice in living and working conditions, the resulting
health differences are more likely to be considered unjust
than those resulting from health risks which were voluntarily
chosen” (Whitehead 1991, 220). In order to support
the view that social differences in health, to wit, the social
gradient, are unfair, proponents of the HESC model accept
that they must answer the possible objection that (a portion
of) these differences are attributable to individual behaviors.
If these behaviors are themselves voluntary, the objection
would claim, the result is not unfair. In response to
this, the proponents of the model challenge the voluntariness
assumption and point out that individual behaviors
are themselves “determined” by SES.
We know that many risky health behaviors, such as
smoking, tend to be more prevalent in lower than in higher
income groups and these differences in health behaviors
make a significant contribution to social inequalities in
health. It is thus crucial for proponents of the HESC model
to take a position on whether this portion of the social
inequalities in health is unfair.
Health shortfalls resulting from these patterned risky
behaviors are considered unfair by the HESC model.
Whitehead emphasizes that some health inequalities
resulting from individuals’ choices are unfair whereas
others are not. Recall that according to her account, health
9. We should note here that, despite the fact that residual inequalities
are not unfair on this account, a universal health care system
may still be a requirement of justice. On this, see also the debate
between Sreenivasan (2007) and Daniels (2007).
10. In places, Marmot claims to rely on the normative framework
provided by the capabilities approach (Marmot 2013, 294–5).
However, this approach does not provide support for the view
that health inequalities are unfair either. Insofar as the capabilities
approach calls for an equalisation of any distribuendum, it, like
other theories of justice, envisages an equal package of the relevant
goods, in this case a bundle of capabilities. Even if health can
be seen as a separate capability, as argued more recently by Sridhar
Venkatapuram, there is no argument to the effect that capabilities
to be healthy alone must be equalized (see Venkatapuram
2011).
March, Volume 15, Number 3, 2015 ajob 31
Caring About Social Determinants of Health
Downloadedby[MasarykovaUniverzitavBrne]at02:3919February2016
inequalities resulting from “health-damaging behaviour
where the degree of choice of lifestyle is severely
restricted” should be considered health inequities; health
inequalities resulting from “health-damaging behaviour if
freely chosen, such as participation in certain sports and
pastimes” (Whitehead 1990, 5), should not be considered
inequities.
Proponents of the HESC model seem to assume that the
patterning of a particular behavior is by itself sufficient to
establish that behavior is “largely determined” by social factors
and that, therefore, the inequalities resulting from such
patterned behaviors are unfair. Marmot (2007) notes that
Contemporary public-health interventions have often given
primary emphasis to the role of individuals and their behaviours.
The Commission recognises the important role of these
factors, but sets them in the wider social context to illustrate
that behaviour and its social patterning . . . is largely determined
by social factors. (Marmot 2007, 1158–9)
We do not deny the conclusion that social inequalities
resulting from differences in health behaviors are problematic;
however, the arguments presented to support this
conclusion are open to challenges. One objection is that the
social patterning of particular behaviors is not sufficient to
justify the claim, made by Marmot, that such behaviors are
“largely determined” by social factors—not least because
behaviors often vary widely within groups. Empirically,
the correlation of particular behaviors with social factors is
not sufficient to establish causation. Neither does the fact
that behavior is patterned establish the normative conclusion
that particular individuals are not responsible for
their choices.11
More importantly, we do not need to show, as the
HESC model seems to assume, that behavior is determined
by social factors in order to conclude that individuals may
not be fully responsible for the resulting health inequalities.
Responsibility is plausibly a matter of degree and most individual
behavior falls somewhere between the fully determined
and freely chosen ends of the spectrum. The
deterministic language used in the HESC model, coupled
with the quick move from social patterning of health behaviors
to the absence of individual responsibility, makes the
conclusion vulnerable by making it dependent on an
implausible but unnecessary claim about determinism.
Thus, although there is scope for arguing that social
inequalities in health are unfair even to the extent that they
result from individual behavior, the social patterning of
health behaviors is not sufficient to establish that claim. To
assess the fairness or otherwise of health inequalities
resulting from behaviors, we also need to know something
about the mechanisms that make individuals in lower
income groups more likely to adopt unhealthy behaviors
than individuals in higher income groups. The literature
has, of course, already identified many such mechanisms
for particular behaviors. Consider, for example, the mechanisms
that have been proposed as contributors to social
inequalities in smoking behavior, which include tobacco
advertising targeting poor neighborhoods, unequal access
to nicotine-replacement therapy, and differences in perceptions
of tobacco use (Voigt 2010).
Our main concern arising from the assessment of the
normative assumptions of the HESC model we have outlined
in this section is that much more needs to be done to
establish why social inequalities (and only social inequalities)
in health are unfair. A coherent view is hard to identify
in the literature. What is undeniable is that the current
distribution of social determinants of health is unjust and
needs to be rectified, but this is not for reasons of health
(even if the health benefits of a fairer distribution would of
course be welcome). In order to argue for a general restructuring
of society on the basis of a social gradient in health,
it must be argued that health inequalities are unfair or
unjust. Such an argument is not inconceivable, but it is
complex and open to challenges. However, even if the normative
claims are established, the policy recommendations
put forward by this model do not necessarily follow, as we
explain in the next section.
MOVING FROM JUDGMENTS ABOUT FAIRNESS
TO POLICY RECOMMENDATIONS
As we discussed in the second section, the HESC model
includes a variety of policy proposals, ranging from the
redistribution of power and wealth to improvements in
housing. The idea that social inequalities in health can
only be addressed through action on the social determinants
of health is a crucial aspect of the model. As Venkatapuram
and Marmot (2009, 86) explain, “It is always
implicit in the SDH literature that the logical social
response to the identification of social determinants of illhealth
is to transform them.” These recommendations are
framed in the language of justice: It is required as a matter
of social justice that we implement policies that change the
SDH in ways that will reduce health inequalities. However,
as we argue in this section, this argument moves too
quickly. We already argued that even if it is the case that
health inequalities are unfair, it does not follow that they
ought to be redressed by altering the distribution of the
SDH, if the SDH themselves are not unfairly distributed.
This section discusses a number of philosophical and
empirical reasons for resisting the policy recommendations
issued as part of the HESC model.
First, even if we accept the empirical assumption that
social factors are “the causes of the causes,” it does not follow
that the most effective way to alter health outcomes is
to alter the “ultimate” causes (Broadbent 2012). Furthermore,
the way the policy recommendations are presented
does not appropriately reflect the uncertainties surrounding
the empirical research on different population-level interventions.
Assessing the effectiveness of interventions that
11. For discussion of relevant questions about possible links
between responsibility and patterned behavior, see also Scanlon
(1995).
32 ajob March, Volume 15, Number 3, 2015
American Journal of Bioethics
Downloadedby[MasarykovaUniverzitavBrne]at02:3919February2016
address social determinants of health when it comes to
reducing social inequalities in health presents a number of
methodological problems. The standards of evidence that
have become prominent in medical contexts cannot be
straightforwardly applied to population-level interventions,
and because of differences in contextual factors, an intervention
that works well in one place can fail in another (Braveman
et al. 2011a; Broadbent 2013).
Second, empirical data shows that in European countries
social inequalities in health have persisted and in
some cases even widened, even where expansions of the
welfare state have reduced inequalities in income and
wealth (Mackenbach 2012). Evidence from the United
Kingdom, where there has been perhaps the most sustained
effort to reduce social inequalities in health through
large-scale social interventions, suggests that these efforts
have had disappointingly small effects on social inequalities
in health, with inequalities in some indicators not
only stagnating but in fact widening (Mackenbach 2010).
Such considerations should make us less confident that
large-scale social policies will indeed have the desired
effects on social inequalities in health.
Now, of course, it is not clear how to interpret the evidence
on associations between inequalities in income and
wealth and social inequalities in health (Mackenbach
2010). One possibility is that we simply have not yet seen
large-scale social changes of the sort envisaged by proponents
of the HESC model. However, it is arguably unrealistic
to expect such pervasive changes in the current political
climate. This underscores the need not only for a clearer
understanding of how different kinds of social policies are
going to affect health inequalities but also for clearly communicating
the complexities and uncertainties surrounding
this question to policy-makers. The move from
empirical observations about the effects of the SDH on
health outcomes to the conclusion that changing the distribution
of the SDH is clearly where we should intervene is,
therefore, problematic.
The claim that policies that reduce the unequal distribution
of social determinants of health are the most effective
way of intervening so as to improve health outcomes
and/or to reduce social inequalities in health seems particularly
problematic when it comes to outcomes that
involve health behaviors. Consider, for example, tobacco
use, which in many countries is a major contributor to differences
in health outcomes across social groups. Even
commentators who emphasize the importance of social
factors contributing to the patterning of smoking—such
as targeted advertising, and resource constraints in
accessing cessation aids—are skeptical that changes to
these social factors would be effective in bringing down
smoking rates in low-SES groups, particularly in the short
term. For example, Hilary Graham, whose seminal study
on tobacco use among working-class mothers highlights
the many ways in which living conditions in deprived
areas sustain smoking practices, advises caution with
respect to the efficacy of policies that address these living
conditions:
Given that smoking is addictive and that both disadvantage
and smoking have long-term and cumulative effects on
health, an improvement in socio-economic circumstances is
unlikely to result in either an immediate reduction in smoking
or an immediate improvement in health. (Graham 1998, 299)
Even when individual behavior does not intervene in
the link between social factors and health, there may be
other measures a government can take to improve people’s
health, as Jonathan Wolff points out based on the findings
in the SDH literature (Wolff 2011). For instance, government—or
even local authorities—can provide people with
the opportunity to rest or perhaps, on the contrary, to continue
working after retirement, depending on what would
be beneficial for their health (Wolff 2011). Such small-scale
interventions may be more effective in improving people’s
health outcomes as well as reducing health inequalities,
and they may also be more realistic policy options, especially
in the short term.
To repeat, we believe that many of the policies recommended
as part of the HESC model are required, as a matter
of social justice. It is the way these policies are linked to
inequalities in health that we find problematic and potentially
counterproductive. As Wolff puts it, it may present
the Minister for Health with a dilemma in that she will
have to argue for the diminished importance of her own
domain (Wolff 2011, 1), and this would be counterproductive.
Furthermore, recommending particular policies as a
means to bringing down social inequalities in health
despite ongoing uncertainty about their effectiveness in
achieving this goal may lead to frustration, both by policymakers
and by the general public, when health inequalities
remain the same or even increase, in the face of large-scale
policy initiatives. If, on the other hand, we argue for these
policies as redressing social injustice simpliciter, noting
health effects as a possible though not certain “side effect,”
that problem is less likely to arise.
A further complication regarding the policy recommendations
issued as part of the HESC model concerns
the precise goals to be pursued through policy interventions
and the relationship between these goals. Two goals
in particular are emphasized by proponents of the model:
improving population health, and reducing health
inequalities (see also WHO 2014, xv). The possibility that
these two goals might diverge is sometimes acknowledged,
but when issuing policy recommendations, proponents
of the model suggest that improvements in overall
health and reductions in health inequalities tend to come
together:
We should have two societal goals: improving health for
everybody and reducing health inequalities. Others may see
them as being in conflict, but they are two separable goals.
Both are worthy and should be pursued. I have never argued
that an overall improvement in health should be sacrificed in
the pursuit of narrower health inequalities. Given my general
thesis that, to oversimplify, good health results from a good
set of social arrangements, I would look to sacrifice other
social goals (a self-serving movement towards making the tax
March, Volume 15, Number 3, 2015 ajob 33
Caring About Social Determinants of Health
Downloadedby[MasarykovaUniverzitavBrne]at02:3919February2016
system less progressive, for example) before accepting that
there had to be a tradeoff between these two health goals.
(Marmot 2013, 283)
However, it is not clear that that these two goals can be
simultaneously achieved in practice; it is certainly not
uncommon in public health contexts that interventions,
even if they lead to benefits for all relevant groups, benefit
these groups unequally so that inequalities increase as a
result (Mechanic 2002). This is in fact what seems to have
happened to life expectancy across different social classes
in the United Kingdom over the period of substantial
investment in large-scale policies to reduce health inequalities
(Mackenbach 2010, 2012; Department of Health 2009).
It may of course be possible to level down—that is, to
reduce inequalities in health and flatten the social gradient
by reducing the life expectancy of those at the top. But this
is clearly not what proponents of the HESC model suggest.
The thought is probably that aiming to bring everyone up
to the “highest possible level” of health—or “up-equalizing”—will
also reduce health inequalities, which is true.
But this is an unrealistic aim, both in theory and in practice:
If we successfully improve everyone’s health outcomes,
including those at the top, while also equalizing
them, there is no end to this aim. Furthermore, it is not
clear that we should pursue it as a matter of justice once
we take into account other considerations of justice.12
In
practice, of course, up-equalizing is likely far too expensive
a goal to be adopted by any government. Thus, these two
goals may conflict, and just assuming that they tend to run
in tandem glosses over the normative question of how to
weigh improvements in overall population health against
reductions in health inequalities when such conflicts occur.
CONCLUSION
In this article, we examined the framework for thinking
about fairness in health that informs many of the highlevel
reports and policy documents issued by a number of
international bodies. This framework starts from a number
of empirical findings that document the influence of social
factors on health outcomes and health inequalities and
draws certain policy conclusions via a number of—often
implicit—normative assumptions. We argued that these
normative assumptions need to be clarified and supported
with clearer arguments.
While we agree with many of the conclusions drawn
by proponents of the HESC model, accepting these conclusions
without a solid philosophical argument may lead to
unclear and possibly even contradictory prescriptions. The
role of social factors in creating health inequalities is
important and should not be overlooked. But it is also
important to understand exactly what is problematic about
the ways in which social factors shape health outcomes. If
the social factors themselves are unfairly distributed, calling
for their redistribution for reasons of health may
detract from important social justice concerns. Addressing
inequalities in income and wealth, for instance, is an
important justice concern, regardless of its effects on health
or health inequalities. But it may be damaging for health to
suggest that only far-reaching societal changes can lead to
improvements in this area.
Although we agree with Marmot (2013, 282) that thorough
philosophical discussion and ethical analysis is
beyond the scope of policy reports, building into these
reports a consistent account of what justice requires in the
area of health can only strengthen the recommendations
offered. Addressing these questions requires collaborative
attention from both philosophers and epidemiologists.
Our aim in this article has been to indicate some promising
avenues for such work. &
ACKNOWLEDGMENTS
We thank Nicholas King, Meena Krishnamurthy, Sridhar
Venkatapuram, and Gry Wester for their helpful comments.
We are also grateful to audiences at the Annual
Meeting of the Canadian Association for Practical Ethics,
University of Victoria; the Annual Conference of the Association
for Legal and Social Philosophy, University of Stirling;
the University of Limerick; and the Research Group
on Normative Aspects of Public Health, Centre for Interdisciplinary
Studies, University of Bielefeld.
FUNDING
This work was funded by Canadian Institutes of Health
Research grant ROH-115214, “Ethics, Social Determinants,
and Health Equity: Integrating Theory and Practice.”
REFERENCES
Braveman, P., and S. Gruskin. 2003. Defining equity in health. Journal
of Epidemiology and Community Health 57: 254–258. Available at:
http://dx.doi.org/10.1136/jech.57.4.254
Braveman, P., B. Starfield, and H. J. Geiger. 2001. World Health
Report 2000: How it removes equity from the agenda for public
health monitoring and policy. British Medical Journal 323: 678–681.
Available at: http://dx.doi.org/10.1136/bmj.323.7314.678
Braveman, P., S. Egerter, S. Woolf, and J. Marks. 2011a. When do
we know enough to recommend action on the social determinants
of health? American Journal of Preventive Medicine 40: S58–S66.
Available at: http://dx.doi.org/10.1016/j.amepre.2010.09.026
Braveman, P., S. Kumanyika, J. Fielding, T. LaVeist, L. Borrell, R.
Manderscheid, and A. Troutman. 2011b. Health disparities and
health equity: The issue is justice. American Journal of Public Health
12. For instance, up-equalizing health outcomes might require an
unequal distribution of educational resources or opportunities
and this may be judged to be unjust (Sreenivasan 2012). More generally,
the pursuit of equality in health may conflict with what justice
requires in other areas. But this is a larger issue that we cannot
address fully here.
34 ajob March, Volume 15, Number 3, 2015
American Journal of Bioethics
Downloadedby[MasarykovaUniverzitavBrne]at02:3919February2016
101: S149–S155. Available at: http://dx.doi.org/10.2105/
AJPH.2010.300062
Braveman, P., S. Egerter, and D. Williams. (2011c). The social
determinants of health: Coming of age. Annual Review of Public
Health 32: 381–398. Available at: http://dx.doi.org/10.1146/
annurev-publhealth-031210-101218
Broadbent, A. 2012. Epidemiology [audio podcast]. Available at:
http://www.minerva-podcast.com/post/22517061542/epi
Broadbent, A. 2013. Philosophy of epidemiology. Basingstoke, UK:
Palgrave Macmillan.
Chandra, A., and T. S. Vogl. 2010. Rising up with show leather? A
comment on fair society, healthy lives. Social Science and Medicine
71: 1227–1230. Available at: http://dx.doi.org/10.1016/j.
socscimed.2010.07.006
Commission on the Social Determinants of Health. 2008. Closing
the gap in a generation: Health equity through action on the social determinants
of health. Geneva, Switzerland: World Health
Organization.
Daniels, N. 2007. Rescuing universal health care. Hastings Center
Report 37: 3. Available at: http://dx.doi.org/10.1353/
hcr.2007.0019
Daniels, N. 2008. Just health: Meeting health needs fairly. Cambridge,
UK: Cambridge University Press.
Department of Health. 2009. Tackling health inequalities: 10 Years
on—A review of developments in tackling health inequalities in England
over the last 10 years. London, UK: Department of Health.
Graham, H. 1998. Promoting health against inequality: Using
research to identify targets for intervention—A case study of
women and smoking. Health Education Journal 57: 292–302. Available
at: http://dx.doi.org/10.1177/001789699805700402
Hausman, D. 2007. What’s wrong with health inequalities. Journal
of Political Philosophy 15: 46–66. Available at: http://dx.doi.org/
10.1111/j.1467-9760.2007.00270.x
Hausman, D. 2013. Egalitarian critiques of health inequalities. In
Inequalities in health: Concepts, measures, and ethics, ed. N. Eyal, S.
Hurst, O. F. Norheim, and D. Wikler, 95–112. Oxford, UK: Oxford
University Press.
Kawachi, I., S. V. Subramanian, and N. Almeida-Filho. 2002. A
glossary for health inequalities. Journal of Epidemiology and Community
Health 56: 647–652. Available at: http://dx.doi.org/10.1136/
jech.56.9.647
Lippert-Rasmussen, K. 2013. When group measures of health
should matter. In Inequalities in health: Concepts, measures, and ethics,
ed. N. Eyal, S. Hurst, O. F. Norheim, and D. Wikler, 52–65. Oxford,
UK: Oxford University Press.
Macinko, J., and B. Starfield. 2002. Annoted bibliography on equity
in health, 1980–2001. International Journal for Equity in Health 1.
Available at: http://dx.doi.org/10.1186/1475-9276-1-1
Mackenbach, J. 2010. Has the English strategy to reduce
health inequalities failed? Social Science and Medicine 71: 1249–
1253. Available at: http://dx.doi.org/10.1016/j.socscimed.
2010.07.014
Mackenbach, J. 2012. The persistance of health inequalities in
modern welfare states: The explanation of a paradox. Social Science
and Medicine 75: 761–769. Available at: http://dx.doi.org/10.1016/
j.socscimed.2012.02.031
Marmot, M. 2007. Achieving health equity: From root causes to
fair outcomes. Lancet 370: 1153–1163. Available at: http://dx.doi.
org/10.1016/S0140-6736(07)61385-3
Marmot, M. 2010. Fair society, healthy lives—The Marmot review.
Strategic Review of Health Inequalities in England Post-2010.
Available at: http://www.instituteofhealthequity.org/projects/
fair-society-healthy-lives-the-marmot-review
Marmot, M. 2013. Fair society healthy lives. In Inequalities in health:
Concepts, measures, and ethics, ed. N. Eyal, S. Hurst, O. F. Norheim,
and D. Wikler, 282–298. Oxford, UK: Oxford University Press.
Marmot, M., S. Friel, R. Bell, T. A. J. Houweling, and S. Taylor.
2008. Closing the gap in a generation: Health equity through
action on the social determinants of health. Lancet 372: 1661–
1669. Available at: http://dx.doi.org/10.1016/S0140-6736(08)
61690-6
Mechanic, D. 2002. Disadvantage, inequality, and social policy.
Health Affairs 21: 48–59. Available at: http://dx.doi.org/10.1377/
hlthaff.21.2.48
Murray, C., E. Gakidou, and J. Frenk. 1999. Health inequalities and
social group differences: What should we measure? Bulletin of the
World Health Organization 77: 537–544.
Rawls, J. 1999. A theory of justice, rev. ed. Oxford, UK: Oxford University
Press.
Scanlon, T. 1995. A good start. Boston Review April/May: 8–9.
Segall, S. 2009. Health, luck, and justice. Princeton, NJ: Princeton
University Press.
Sreenivasan, G. 2007. Health care and equality of opportunity.
Hastings Center Report 37: 21–31. Available at: http://dx.doi.org/
10.1353/hcr.2007.0033
Sreenivasan, G. 2008. Justice, inequality, and health. In Stanford
encyclopedia of philosophy, ed. E. Zalta. Available at: http://plato.
stanford.edu/entries/justice-inequality-health/
Sreenivasan, G. 2009. Ethics and epidemiology: Residual health
inequalities. Public Health Ethics 2: 244–249. Available at: http://
dx.doi.org/10.1093/phe/php030
Sreenivasan, G. 2012. Is it permissible to maximise individual
health? Paper presented at Justice in Health: Trade-offs and
conflicting values, McGill University, Montreal, Canada, April
27–28.
Starfield, B. 2001. Improving equity in health: A research agenda.
International Journal of Health Services 31: 545–566. Available at:
http://dx.doi.org/10.2190/DGJ8-4MQW-UP9J-LQC1
Venkatapuram, S. 2011. Health justice. Cambridge, UK: Polity
Press.
Venkatapuram, S., and M. Marmot. 2009. Epidemiology and social
justice in light of social determinants of health research. Bioethics
23: 79–89. Available at: http://dx.doi.org/10.1111/j.1467-
8519.2008.00714.x
March, Volume 15, Number 3, 2015 ajob 35
Caring About Social Determinants of Health
Downloadedby[MasarykovaUniverzitavBrne]at02:3919February2016
Voigt, K. 2010. Smoking and social justice. Public Health Ethics 3:
91–106. Available at: http://dx.doi.org/10.1093/phe/phq006
Whitehead, M. 1990. The concepts and principles of equity in
health. Copenhagen, Denmark: World Health Organization.
Whitehead, M. 1991. The concepts and principles of equity and
health. Health Promotion International 6: 217–228. Available at:
http://dx.doi.org/10.1093/heapro/6.3.217
World Health Organization (Regional Office for Europe). 2014.
Review of social determinants and the health divide in the WHO
European Region: Updated Reprint. Copenhagen, Denmark:
World Health Organization.
Wolff, J. 2011. How should governments respond to the social
determinants of health? Preventive Medicine 53: 253–255. Available
at: http://dx.doi.org/10.1016/j.ypmed.2011.07.008
36 ajob March, Volume 15, Number 3, 2015
American Journal of Bioethics
Downloadedby[MasarykovaUniverzitavBrne]at02:3919February2016