Gender in/and/of Health Inequalities

Dorothy Broom

Abstract
The relationship between gender and health inequalities is potentially
complicated, raising questions for health and social research, practice and
policy. In this paper, I use two population health case studies - obesity and
smoking - to explore the interplay between gender and socioeconomic
position. The cases show that, on its own, neither dimension of inequality
affords a comprehensive picture of these significant risks to public health.
Furthermore, historical change in the socioeconomic and gendered
distribution of these health risks suggests that gender is best considered as a
dynamic and layered form of differentiation, rather than as a simple or stable
dichotomy A more nuanced approach to the analysis of gender and health
has the potential to generate both more fruitful research and more effective
health and social policy
Keywords: gender, smoking, obesity

11
Gender in/and/of Health Inequalities

Inequality has always been a fundamental interest of social science. Yet the relationship
between various forms of social inequality - including gender - remains a vexing
problem, not only for researchers but for people who supply health and social services,
particularly to populations characterised by multiple forms of disadvantage. How should
inequality be understood? How do its various forms relate to one another? Is gender as
important as poverty, race and ethnicity? What happens when one form of difference
is emphasised over the others, either within social scientific research, or in service
provision? What happens when gender is ignored? And if gender is not to be ignored,
how should we thirik about what have been called its 'intersections' with race and class?
This paper cannot address all those questions, but it makes a beginning by summarising
briefly the challenges of dealing with multiple forms of social difference simultaneously,
pointing to the weaknesses inherent in omitting gender, and then exploring in some
detail the history of two significant population health risk factors - obesity and cigarette
smoking - as case studies of the interplay between class and gender in health. I conclude
by proposing a different metaphor for thinking about gender: one may that enable us to
incorporate its complexity as we deal with the practical challenges of social research and
social action.

Complicating class: the problem of gender

Proponents of competing models of class and stratification have long debated whether
the primary focus belongs on production or consumption, income or wealth, Marx or
Weber. Students of race and ethnicity have asked how far the disadvantages of racial
minorities can be explained by lower income and education, and how much these
are compounded by prejudice, discrimination and segregation. Discussion of such
questions has revolved mainly around empirical and political choices, rather than the
epistemological assumptions underlying how to measure difference.
With the emergence of feminist social science during the 1970s, it became apparent
that traditional methods of measuring social position were deficient in at least two
respects: first, they typically assign everyone's status on the basis of the occupation of the
male head of household; and second, they ignored unpaid labour (most of which was
done by women) (Acker 1973; Crompton &C Mann 1986). Attributing status to a whole
family on the basis of a single individual leaves out contributions of other members of
the household (which may be substantial), and it ignores the differential distribution of
resources within households, assuming instead that everyone shares equally (Edwards
1984). It also implicitly assumes that only single women are in paid employment (or at
least paid jobs only matter for women who are not married). And it completely omits to
address the sexual division of labour or the power relations of the reproductive sphere
(Delphy 1981).
While academic feminists were taking their male colleagues to task for these
discriminatory failures, women of colour were criticising the failure of white feminists to
appreciate the significance of race and ethnicity, and the ways white women benefit from
the oppression of other' women and men (Hartman 1976; Hooks 1984). They insisted
on the diversity of women against the tendency to essentialise and homogenise gender
categories, as if all women share identical interests and are fundamentally the same as all
other women and different from all men.

12 Australian Journal of Social Issues Vol.43 No.1 AUTUMN 2008

Thus, the advent of gender studies challenged everyone to devise new ways of
conceptualising fundamental social theories and the traditional methodologies of social
research. More than three decades later, a proliferation of publications on the topic has
succeeded mainly in complicating matters further, rather than resolving the problems
thrown up by the insistence that gender must be included as a fundamental element of
social structure (Andersen 2005).
The appearance of queer theory makes the consideration of gender itself more complex
than is typically acknowledged (de Lauretis 1991). The default option where gender
is concerned is to invoke some version of the sex/gender distinction, which operates
discursively a bit like deploying the ^X'HO definition of health: it supplies a legitimate
placeholder to cover lack of consensus about meaning (in either case). These deeply
ambiguous and contested concepts nevertheless exert enormous influence, and the
instability surrounding their meanings permits them to be deployed in a wide variety of
ways. Indeed, their popularity with the media, politicians and academics may stem in
part from their instability. Their shifting (even contradictory) social and political effects
may also rely on the impossibility of consensus. For example, policy claims based on
gender are presently off Australia's national agenda, while at the same time, the WHO
Global Commission on the Social Determinants of Health gives pride of place to the
concept (Irwin et al. 2006). Elsewhere (Broom 2005), I have explored the feminist
genealogy and academic uses of gender, so here it will sufñce to restate my conviction
that the sex/gender distinction is an occasionally useful heuristic device that should not
be confused v^dth a conceptually reliable boundary (Gatens 1983; Scott 1988). A recent
news article referring to the gender of a foetus reinforced that conviction.
Consequently, I use the terms sex and gender (and masculine and feminine) for
convenience, without implying a clear division. The sex/gender dichotomy is widely
deployed in discussions about health (among many topics). Sometimes relying on the
dichotomy may not make much difference, but when it comes to health and other fields
where the body is highly salient, a naive commitment to the distinction may be hazardous
because it suggests that some aspects of health are 'purely biological' (sex) and hence
exempt from the signifying and socially organising processes of gender; while others
are entirely' psychological or social and thus somehow immune from the materiality of
the flesh (Broom 2001). As Connell (2003: 371) has put it, 'the cultural constmaion of
gender does not express but constantly overrides bodily difference and similarity', so sex
and gender are always implicated in one another.
In these challenging circumstances, how might we usefully think about gender? The
term is occasionally used as a synonym for women,' but clearly, gender means more
than women. Indeed, it means more than women and men as demographic categories.
It must also incorporate dimensions of sociality and social structure. Ridgeway's (1999:
192) definition is a good start:
Gender is a system of social practices within society that constitutes
people as different in socially significant ways and organizes relations of
inequality on the basis of that difference.
To add to that, I understand gender as a fundamentally relational concept whose
elements constantly refer to and interact with one another (mostly implicitly), and in

13
Gender ¡n/and/of Health Inequalities

SO doing, mutually constitute one another. In addition, gender can be analysed as a

central element of social structure, as an organising social process, (Petersen 2003),
as a core dimension of culture and as a fundamental aspect of personal identity. It
is simultaneously productive and constraining, an accomplishment and an ongoing
performance, symbolic and material (Auerbach 1999), embodied in individuals and
encoded in institutions, material objects and behaviors.
Gender is both produced by and in turn structures major social institutions, including the
state, technology, paid work, and the home (Connell 2002; Mellstrom 2003; Williams
2001). Embodied individuals are shaped by gendered institutions and social practices,
and at the same time, they question, revise and resist those forces. That is, gender entails
both structure and agency. It manifests at all levels of analysis, from the individual to the
global. Consequently, we should not be surprised by the fact that gender is so fraught
with contradictions, confusions, incoherence and misunderstanding (Connell 2003).

Complicating health: class and gender

In addition to the conceptual complexities, we face some daunting methodological
challenges in the effort to consider gender and health inequalities together. For the
last two decades, the term 'health inequalities' has been used mainly to refer to the
systematic patterns in health status associated with social class or socioeconomic
position. The ubiquitous finding that higher class is associated with better health (Adler
ÔC Ostrove 1999) is occasionally disputed by the selection of another health outcome,
or a different measure of socioeconomic position. Thus, occupational status, income,
wealth, and education are not always consistently related to any given health outcome or
risk factor, let alone a different indicator of health. Although the approach of assigning
a woman to the socioeconomic status of her husband or father has been extensively
criticised by feminist sociologists, it is still often employed, much like the oversimplified
dichotomy between sex and gender. Such simplifying assumptions may be necessary
for quantitative measurement, but they ignore what may be pertinent considerations
including the ways in which the social system and the economy are invisibly and
inherently gendered, and the ways gender is variously classed. These are the focus of this
paper.
Much health inequalities research (like other research on class and status) has
concentrated mainly or exclusively on men, or else the results have not been broken
down by gender (Macintyre & Hunt 1997). Typically the measured associations between
SES and health outcomes have been stronger among men than women (Matthews et
al. 1999). It is unclear whether this difference arises from the limited accuracy of the
usual indicators of status (occupational prestige and income) when applied to women
(Annandale & Hunt 2000; Arber 1989) or from gender differences in the effects of
socioeconomic position on health.
A further methodological difficulty is created by the limited tinderstanding of the health
effects of unpaid work (Csete 2005). Domestic labour and caring work (still highly
gendered) are omitted from official national accounts (Waring 1988). They are usually
absent from studies of health and work and from research on the lifeways of people in
different socioeconomic locations. Finally, the epidemiological and policy emphasis on
conditions that lead to hospital admission or death tends to understate the burden of

14 Australian Journal of Social Issues Vol,43 No,1 AUTUMN 2008

disease from more diffuse and less dramatic illness conditions (including reproductive
health problems) which may be more common among women. I cannot offer solutions
to these problems, but I identify them here to acknowledge that they generate a lot of
background noise against which we strain to hear faint whispers that might permit us to
develop a more informative approach.
With these qualifications and complications in mind, I now explore two particular
instances of population health patterns in an effort to unpack what might be at stake
in bringing gender and health inequalities together. My hope is that these examples
can help us understand what gender can contribute to a more theoretically informed
approach to population health as a source of accurate and useful descriptions, and
effective health and social policy. •
The examples I consider here are obesity and smoking, by many accounts, two of the
most consequential health risk factors for contemporary populations in rich and poor
nations alike. Each illustrates somewhat different aspects of the nexus between health
inequalities in gender. I use them to develop a more detailed approach to theory and
research on gender and health inequality than that generated in most public health
writing.

Inequalities in healthy weight^

Like most other tisk factors for ill health, excess weight - the current 'lifestyle disease'
- tends to be more prevalent among people further down the social and economic
scale, so that sectors of the population who are most disadvantaged in other respects
are also more likely to be too heavy (Ball & Crawford 2005; Sobal &c Stunkard 1989).
Internationally, this description has applied mainly to rich societies, and it is very
different from how one would have described the situation a century ago when thinness
tended to be associated vwth poverty, malnutrition and disease (particularly tuberculosis),
so that heaviness was more prevalent among the wealthy, especially wealthy men. The
reversal from the historic direa relationship to the present inverse relationship happened
in less than a hundred years (Caballero 2007)^. Members of ethnic and racial minority
groups are also now typically at higher risk of obesity. In the USA, obesity is more
prevalent among Blacks and Latinos than among other Americans (Wang & Beydoun
2007; Yancey et al. 2006). In Australia, the National Health Surveys show that the
Aboriginal and Torres Strait Islander population is nearly twice as likely to be obese as
other Australians. Obesity is less prevalent among urban dwellers than in regional and
rural areas.

Here I use published statistics and data from several Australian Health surveys^ to
explore the patterns in this country. Because it is a health risk factor, we are not surprised
that obesity is inversely related to socio-economic position. The advantage of having

1 The focus here is on obesity because it is less ambiguous as a driver of deleterious health consequences
than overweight.
2 The prevalence internationally is changing rapidly with globalisation and new forms of development.
Paradoxically, in some middle-income nations, there is evidence of both excess nutrition and
malnutrition, sometimes in the same households (Caballero 2007; Doak et al. 2005).
3 This seaion draws heavily on a recent book chapter co-authored with Sharon Friel (Friel Si Broom,
2007).

15
Gender in/and/of Heaith Inequalities

more education is clear from research reported by the AIHW (Australian Institute of
Health and Welfare 2003). People with post-school qualifications are much less likely
to be obese ( 15%) than those who lack such qualifications ( 19%). Similarly, people with
higher incomes are less liable to be obese than those on lower incomes.

Obesity, SES and gender

Obesity is also systematically distributed by gender. International statistics from the
International Obesity Task Force show that in cotmtries wdth developed economies the
prevalence of obesity is usually 15-20% among men and 20-25% among women. Rates in
the USA are now over 30% for both women and men, in contrast to Japan where obesity
remains under 5%. In countries with relatively low gross national product such as those
in Central and Eastern Europe, Asia, Latin-America and Africa, overall rates of obesity
are typically lower than in rich societies (Samoa is an exception), but the direction of the
sex difference is consistent (Loueiro 6c Nayga 2005; Seidell 2000), and in most Pacific
Island nations and countries of the Middle East, the female excess is quite marked.
While faaors such as genes and hormones may contribute, they do so in the context of
the gendered meanings of bodies, gender norms for appearance, differential access to
resources, and socially organised relationships to the proximal causes of body weight:
food and physical activity.
The picture becomes more complex when gender and class are considered together.
Studies from several developed nations find that the inverse relationship between SES
and obesity is relatively consistent among women, but it is less clear or absent among
men (Ball et al. 2002; Cutler et al. 2003; Eckersley 2001; Langenberg et al. 2003;
McLaren 2007; Molarius et al. 2000; Sobal & Stunkard 1989; Wardle et al. 2002; Yu
et al. 2000). Our analysis of four recent Australian surveys examined the prevalence of
obesity by levels of income for men and women. We noted that the expected negative
association was fairly consistent among women, but was less distinct or even absent
for men. Indeed, there was a suggestion of a positive relationship between obesity and
income which is contrary to both the typically inverse association between SES and
health risk, and also to the pattern observed for women. We also noted some anomalies
in the obesity gradients among men when analysing by occupational rank. Education
presented a more consistent picture for both sexes of declining obesity prevalence with
increasing amounts of education (Friel &c Broom 2007).
That is, although obesity in Australia is most prevalent in population groups with the
least education and lowest income, the nature of the relationship between SES and
obesity appears to differ for men and women for certain indicators of socio-economic
position. Women with low education levels and low income are clearly more liable
than other women to be obese, and men with less education are also more prone to be
obese then better educated men. The possible sex difference appears in the suggestion
of a direct relationship between male obesity prevalence and income, although this not
consistent.
Education appears to operate similarly for both sexes, perhaps because it enables the
person to access and tmderstand information on healthy diet and physical activity.
Income is more ambiguously related to obesity, as it permits high level constunption,
including consumption of food and leisure activity that may encourage or discourage

16 Australian Journal of Social Issues Vol.43 No.1 AUTUMN 2008

healthy weight. In the Australian surveys we examined, high income was associated with
healthy weight among women, but not among men. Perhaps the sexes are prompted to
buy different goods and services with their incomes.
Thus, despite the fact that measures of SES are correlated with one another, they may
generate somewhat different patterns in the distribution of obesity because they entail
different resources affecting weight-related behaviour in different ways, and because
some indicators of SES may not be equally appropriate for women and men both.
Eurthermore, the association between SES and weight may be inherently complex.
Eor example, while high SES is usually health protective, very long work hours (often
required to achieve high status) may partly undermine the status advantage by limiting
time for exercise or to prepare home-cooked meals. However, while high status jobs
typically require long hours, long hours on the job give no guarantee of achieving high
status. Thus, people in low status jobs that require long hours (Ostry et al. 2006) have the
worst of both worlds, and may be particularly vulnerable to obesity (Burton & Phipps
2004); and depending on their human capital and their physical environment, they may
be particularly disadvantaged in trying to prevent or manage weight problems.
Offer has asserted that the influence of class on weight occurs through gender. He
claims that women's weight is more strongly determined by socioeconomic status than
men's and women also care more about it' (Offer 1998). The Australian data presented
above partly align with the claim. If it is correct that women's weight is more strongly
determined by their status', women would be wise to care more about it'. Whether Offer
is right or not, his observation alerts tis to the awkward reality that neither SES nor body
weight can be fully understood in reductionist terms, nor - 1 add - can gender.
There is indirect evidence for the logic of the class-through-gender argument. A study
of US adolescents reported that obese girls were significantly less likely to go on to post
school education than their non-obese peers, but that obesity made no difference to the
post-school education of boys (Crosnoe 2007). Internationally, research has suggested
that obese women incur more significant economic penalties for their weight than
obese men do, whether the penalty is measured in terms of income (Mitra 2001) or
wealth (Zagorsky 2005). In US research, obese white women are most disadvantaged
financially by their weight, whereas white men and black women who are obese suffer
comparatively limited economic disadvantage, while obese black men pay no penalty
at all (Zagorsky 2005), perhaps because they are already so disadvantaged. The chances
of being married are more compromised for women than men who are obese. These
findings show how race, class and gender interact to complicate the interplay between
weight and class, and they indicate that violating the norm of slimness is particularly
salient for those women who might compete for relatively high-paying jobs, or who
might marry men with such jobs.

These observations accord with women's intense interest in diets and weight control,
and with men's comparative indifference or resistance to organised weight loss activities.
They may also help to account for the puzzling possibility of a flat or even direct
association between SES and obesity among men but an inverse association among
women. If high incomes enable the consumption of excess of food, and demanding
but sedentary jobs discourage physical activity, higher class may become a risk factor
for unhealthy weight, particularly among upper-class men who may not pay such a high

17
Gender in/and/of IHealth Inequaiities

social or economic price as upper-class women for being too heavy. Perhaps men's blue
collar jobs still entail enough physical activity to reduce their risk of obesity, compared to
white collar jobs and working class women's jobs, although there are conflicting findings
on this point (Mummery et al. 2005; Salmon et al. 2000). Such important questions
require extensive investigation, but as yet there are no conclusive studies that resolve
them.
I now turn to my second example.

The genders of smoking^

Like obesity, smoking follows the socioeconomic gradient in developed societies. People
lower down the socioeconomic scale are more likely to smoke than wealthier, higher
status people. And like obesity, the gradient shifted during the 20th century, particularly
if we define 'smoking' as the regular use of commercially manufactured cigarettes^.
Indeed, the advent of mass-produced cigarettes both enabled and drove the dramatic
expansion of the market. Initially expensive and a signifier of discretionary income,
cigarettes were democratised, particularly by war. As its health hazards became more
widely known, and its capacity to display elite status diminished with lower prices, its
popularity among the well-to-do declined, gradually transforming it into a marker of
working<lass status or poverty (Barbeau et al. 2004; Turrell &c Mathers 2000).
In an equally striking shift, cigarette smoking had been a masculine preserve at the
beginning of the 20th century, but became a rare case of gender equality by the end,
at least in the Anglo-American democracies (there is more variation in nations of the
EU (Graham 1996)). That is, while globally many more men than women smoke, the
differential is shrinking in many nations; and in North America, the UK and Australia,
there is no longer any consistent sex difference (McDermott et al. 2002). Unlike
obesity, however, as far as I am aware, there are few anomalies in the class and gender
interactions surrounding smoking, although this is a topic worth investigating. The
dramatic transformation in the sex ratio is - for my purposes - the standout feature.
The changing relationship between gender and smoking is a remarkable story in which
cigarette manufacturers and tobacco marketers have consistently outpaced the public
heaith sector (Jacobsen 1986), and the tobacco industry's theoretical and empirical
superiority to public health where gender analysis is concerned was evident very early.
Recruiting women to smoke was not simply a matter of their increasing exposure to
previously male environments during the world wars and subsequently, although it is
likely that those experiences played a significant part. The Lucky Strike campaign 'Reach
for a Lucky instead of a sweet' (1925) clearly targeted women and drew an explicit
link between smoking and weight control, although another version of the campaign
(showdng sports stars) apparently targeted men (Blum 2005). In the 1920s and '30s,
product endorsements by movie stars began to glamorise smoking for women. But it is

My focus is on developed societies, especially English-speaking nations, and on dominant culture

populations within those societies. The history, cultural and economic processes must be different
among minority populations such as migrants and Indigenous people, but data are not available to
address these important matters; I can only acknowledge them.
Other ways of smoking (particularly pipes), and using tobacco in forms other than smoking (snuff and
chewing tobacco) have a somewhat different social history, not covered here.

18 Australian Journal of Social Issues Voi.43 No.1 AUTUMN 2008

astonishing to learn that Philip Morris organised a lecture tour to give American women
lessons in cigarette smoking (Amos 6c Haglund 2000). Such marketing was apparently
required to overcome the unwelcome symbolic link between smoking and a disreputable
feminine sexuality.
The invention of commercially manufactured cigarettes generated the opportunity to
're-gender' smoking, and simultaneously created a commercial imperative to do so. Once
manufactured cigarettes became available, the industry wanted to expand the market for
their new product, and women were the obvious target for such expansion.
Amos and Haglund put it succinctly: 'In comparison with traditional smoking methods,
cigarettes were clean, easy-to-use, modem, and increasingly cheap' (Amos & Haglund
2000: 3). That is, the faaory-made cigarette redefined smoking as potentially appropriate
for women in part by producing an appropriate objea for a woman. If women were to
smoke in public (something sought by both tobacco manufacturers and some women
themselves), a major change to its image was required. During the first half of the 20th
century, the image of the cigarette as dirty and masculine was supplemented with the
idea that smoking could be 'not only respectable but sociable, fashionable, stylish, and
feminine' (Greaves 1996: 4).
At the same time, smoking began to signify the modern world (as opposed to traditional
customs), female independence and emancipation (Brandt 1996). A persuasive case is
made that the cultural meanings of women's smoking gradually shifted from implying
loose morals to indicating modernity, feminine glamour and sexual desirability (Greaves
1996), and by the 1970s, these connotations had become marketing staples. Documents
from the international tobacco industry reveal a highly sophisticated appreciation of how
class and age inflect gender to create an elaborately segmented cigarette market (Gollin
2005). For example, RJ Reynolds identified and developed advertising pitched toward
what it labelled the 'virile female': a young working-class woman with limited education,
a blue- or pink-collar job, and a present-oriented, fun-loving adventurous outlook
(Barbeau et al. 2004). The Philip Morris version of this woman was the 'maverick female
smoker.' Apparently, transnational tobacco has largely abandoned the effort to recruit
more educated smokers (either men or women) in the US, and is focusing its efforts on
working-class people.

UK and Ganadian research has shown that for working-class mothers isolated with
young children, 'cigarettes may provide the only moment when the struggle for financial
survival can be suspended and they can join a world of personal consumption that
most adults take for granted' (Graham 1994; Greaves 1996). In a cultural economy of
commodification, such dynamics may be potent indeed, and they suggest the potentially
perverse effects of a widespread health promotion strategy that urges pregnant women
smokers to quit for the health of the foetus. It may encourage temporary cessation, but
its implicit message (that the reason for a woman to quit is the well-being of someone
else) is liable to foster the low self-regard that often contributes to women's smoking in
the first place, and hence to relapse after the baby is born.

The masculine meanings of smoking have also been dynamic. For example, although
smoking has long been considered masculine, in early 20* century USA, pipes and cigars
were considered more manly and cultivated forms of tobacco use than cigarettes which

19
Gender in/and/of Health Inequalities

were considered European and effete, or else rough and uncultured (Starr 1984). In
popular cinema and novels, cigarette smoking might identify a man as a villain, or lacking
in character. But these connotations w^ere soon destabilised and enlarged. For example,
beginning in the late 19''' century, cigarette packs often included cards with images of
sporting stars, and collecting and trading these cards became a popular hobby (Blum
2005). Like the film stars glamorising smoking for women, sporting heroes endorsed
particular cigarette brands, lending them a new masculine appeal.
The free issue of tobacco to soldiers in both world wars was part of the evolution in
masculine signification of cigarettes. It reflected and consolidated smoking as a portable
pleasure, accessible in the absence of all others. It reconfirmed smoking as an activity of
men (heroic men in this case) during an era when women were taking it up in increasing
numbers, and it linked smoking to patriotism for everyone. The traditional Australian
smoke-o' is both classed and gendered, the working man's entitlement to a break
from the toil of the job. The cigarette-smoking villain of popular culture was joined by
resourceful, independent and sexually compelling romantic leads like Humphrey Bogart.
Health research suggests that some of the factors driving smoking and obstructing
quitting differ subtly between the sexes. For example, smoking appears to be more
strongly linked to emotional distress and concern about weight among women (Berlin
et al. 2003; Camp et al. 1993; Honjo &C Siegel 2003), factors which may contribute to
women's generally lower rates of cessation (Abrams et al 1995; Jeffery et al. 2000; Patton
et al. 1998). Men, especially young men, may be particularly motivated to abstain from
smoking by participation in sporting activities and a desire for physical fitness (Rodriguez
Sc Audrain-McGovern 2004; Thiri Aung et al. 2001). By contrast, adolescent girls have
few alternatives to cigarettes to mark themselves as cool' (Plumridge et al. 2002).
To understand such processes requires attention to the social functions of smoking
(Klein 1993) because it is a social praaice as well as an individual habit. For example,
females appear to be more likely than males to acquire cigarettes through non-
commercial means (Castrucci et al. 2002), so giving and receiving cigarettes contain
strongly gendered and sometimes sexual elements. Sociologically oriented studies show
that smoking operates to support sociability and to build solidarity (Barbeau et al. 2004;
Bialous 2005; Greaves, 1993), elements that are also likely to have implications for
gender. We seem to have come a long way since Virginia SUms. The 'virile female' won't
smoke a brand of which her boyfriend disapproves. 'Real women' don't need to smoke
girlie cigarettes to display their femininity.
These histories are intersected by the anti-tobacco movement and health education
(throughout the 20* century) striving to introduce new meanings, but also drawing upon
discourses that elided health and morality. Paradoxically, health promotion discourses
help to configure smoking as a form of resistance to expert authority, to condescending
adults, to the good girl i m ^ e (Banwell & Young 1993). More recent research suggests
that some high status young women may be smoking in order to demonstrate that they
are not uptight (Lennon et al. 2005).
Oddly, little interest is shown in how the gendered symbolic changes and converging
sex ratio might be affecting the meanings of smoking to and among men. If it is
true that gender is relational, surely it is not possible for women's smoking to be so

20 Australian Journal of Social Issues Vol.43 No.1 AUTUMN 2008

transformed without also changing men's smoking. I have found no research to suggest
it, but I wonder whether some of the decline in smoking prevalence among men may
be attributed to the féminisation of smoking, both symbolically and behaviourally.
That is, the potential for smoking to signify forms of masculine identity will be altered
when equal numbers of women smoke. That is not to imply that smoldng has ceased
to be masculine, or that it simply has become feminine, but that its signifying potential
is constantly shifting. Perhaps one of the new meanings relates to the creation of new
subject positions in which working-class white men alienated by corporatism, feminism
and professional power adopt anti-health lifestyles as a way to reassert heterosexual
masculine authority.

By contrast to rich societies, in the majority of less-developed nations very few women
smoke, a differential of keen interest to the tobacco industry. Consequently, the current
focus of transnational tobacco is on expanding untapped markets, particularly among
women in less-developed nations and Asia, where women's rates of smoking have
remained extremely low (Csete 2005). We should expect the industry to seize on the
opportunities created by 'free trade' agreements and by weak regulation of tobacco
advertising, and yet again to show considerable sophistication in their appreciation of
how gender and culture can be used to open up this market. Where they have done so
in the former Soviet bloc, the impact on smoking prevalence among women has been
nothing short of spectacular (Amos & Haglund 2000).

Obesity and smoking: gendering health inequalities^'

Obesity and smoking have more in common than the dubious distinction of being
among the most significant preventable causes of ill-health and premature death.
Although surveys are equivocal, it seems that some people smoke as a method of weight
control. The proliferation of competing authorities' and conflicting information are
making it increasingly difficult to balance various health (and other) priorities in daily life
(Dixon &c Winter 2007). Additionally, obesity and smoking are both socially stigmatised
and objects of intense public health scrutiny and numerous individualised heaith
promotion campaigns. Furthermore, for women at least, both smoking and emotional
eating are linked to psychological distress in a truly vicious circle in which anxiety
provoked by the behaviour itself (eating or smoking) triggers more of the behaviour and
further anxiety (Camp et al. 1993).
The two case studies pose fundamental challenges for how theoretical and research
questions should be framed. Should the effort be to explain how gender affects the SES
gradient in obesity.' Or should the question be: how does SES affect sex differences
in excess weight? For smoking, the focus is perhaps clearer: how can public health
mobilise gender effectively to diminish the appeal of smoking to poor and working-class
populations, and to slow the relendess global spread of tobacco? And how can they do
so without reinforcing destructive gender stereotypes?
Whatever the specific question, it is apparent that public health will continue to have
limited success if it cannot incorporate gender into the study of health inequalities. In
so doing, the understanding of gender must be theoretically informed and empirically

6 • This section draws on an unpublished manuscript co-authored with Jane Dixon.

21
Gender in/and/of Health Inequalities

detailed; that is, it must recognise the diversity among men and among women, not
simply the differences between them as binary categories. Gender can only be a resource
for improving public health if it reflects a level of analysis at least as sophisticated as that
of transnational tobacco. And that will require a theoretically nuanced understanding of
how gender operates. Again, the tobacco case is illustrative.
Commentators often s u r e s t that the gendered meanings of smoking have changed,
and that the old connotations have been replaced. I propose that the case of smoking
illustrates a more subtle and intricate process than simple substitution. Instead, it seems
to me that the previous connotations persist as new ones are brought into play. Indeed, I
believe that efforts to offer new significations (whether to enlarge the market or improve
population health) are layered on the traces of the previous meanings. That is, gender is
more like a palimpsest (remember the child's 'magic slate' toy?) than it is like a stencil.
Earlier connotations are not erased completely, but continue to permeate the way people
interpret and embody current meanings. It is through such layering that smoking can
remain simultaneously a symbol of rugged masculinity and feminine attractiveness.

To conclude
In summary, then, trends in both smoking and obesity are patterned by class and gender
in dynamic and complicated ways: patterns that do not easily yield their secrets. During
the last century, the social distribution of smoking prevalence has shifted dramatically.
Overall, it has conformed to the assumption that health risks (once they are identified
as such) are likely to be avoided by elites while working class and disadvantaged people
continue to be exposed. However, a variety of social and commercial developments
over the same time have generated dramatic changes in prevalence by gender, with
women taking up smoking while rates among men have levelled off, resulting in
a rough convergence. However, w^ith obesity, a focus on gender has confused the
socioeconomic pattern, possibly signalling an anomalous case in which the class gradient
is comparatively consistent for women, but inconsistent or even reversed among men,
raising the question whether there might be other such 'anomalies'.
These examples suggest that there is no simple formula for how we should relate gender
and health inequalities. There are, however, some lessons for researchers. Clearly, a
comprehensive understanding of any health condition still needs epidemiological data
that permit comparative analysis by gender, race/ethnicity and several measures of SES.
Furthermore, it requires both quantitative and qualitative information on the historic,
cultural, economic, political, physical and social environments in which the subjeas of
epidemiological research are located. We must also appreciate that similar prevalence
rates may be generated by gendered processes (as in the case of smoking), so it is unwise
to assume that gender is irrelevant simply because there is no sex difference in the rate
of a risk or illness. A comprehensive understanding will require attention to commercial
interests and activities, as well as the actions of the state and civil society (Hinde &
Dixon 2005). Such a prescription can only be filled by multidisciplinary collaborations
involving a range of population health researchers.

22 Australian Journal of Social Issues Vol.43 No.1 AUTUMN 2008

Finally, the relationship between gender and health is one in w^hich gender shapes a
variety of exposures and experiences. And at the same time, the gendered enactments
and embodiments of health and risk may reflect resistance as well as conformity to
social norms (including gender norms), and may entail the production of new classed
and gendered subject positions. Because gender is effectively multiplied by other forms
of social difference, and because of the dynamics of history, a binary view of gender has
great potential for unintended consequences in the hands of the public health researcher
or practitioner.

Consequently, there is as yet no general, all-purpose formula for understanding how

gender and other inequalities organise the health of individuals or populations, or
how to respond to those patterns. For now, we may have to be content to generate
detailed pictures of their interplay in the case of particular health conditions in an
effort to develop more accurate empirical studies, theories rooted in the empirical, and
interventions that can enhance the health of the people with priority given to identifying
and addressing the needs of the most vulnerable.
References
Abrams, J. S., Phillips, P. H., & Friedman, M. A. (1995) 'Commentary: A reappraisal of
research results for the local treatment of early stage breast cancer Journal of the
National Cancer Institute, 87(24), 1837-1845.
Acker, J. (1973) "Women and social stratification: a case of intellectual sexism, American
Journal of Sociology, 78(4), 174-183.
Adler, N. E., & Ostrove, J. M. (1999) 'Socioeconomic status and health: what we know
and what we don't'. In N. E. Adler, M. Marmot, B. S. McEwen & J. Stewart (Eds.)
Socioeconomic Status and Health in Industrial Nations: Social, Psychological, and
Biological Pathways, New York, New York Academy of Sciences.
Amos, A., 6c Haglund, M. (2000) 'From social taboo to "torch of freedom": the
marketing of cigarettes to women'. Tobacco Control, 9(1), 3-8.
Andersen, M. L. (2005) Thinking about women: a quarter centtiry's view'. Gender &
Society, 19(4), 437455.
Annandale, E., & Htint, K. (2000) 'Gender inequalities in health: research at the
crossroads'. In E. Annandale & K. Hunt (Eds.) Gender Inequalities in Health,
Buckingham, Open University Press.
Arber, S. (1989) 'Gender and class inequalities in health: understanding the differentials'.
InJ. Fox (Ed.) Health Inequalities in European Countries, Aldershot, Gower
Publishing Company,.
Auerbach, J. ( 1999) 'Gender as proxy'. Gender & Society, 13(6), 581-583.
Australian Institute of Health and Welfare. (2003) 'Are all Australians gaining weight?
Differentials in overweight and obesity among adults, 1989-2001', A/HWßM//ei/«,
11 (December), 1-24.
Ball, K., & Crawford, D. (2005) 'Socioeconomic status and weight change in adtdts: a
review'. Social Science & Medicine, 60(9), 1987-2010.

23
Gender in/and/of Health Inequalities

Ball, K., Mishra, G., & Crawford, D. (2002) 'Which aspects of socioeconomic status
are related to obesity among men and women?', Intemational Journal of Obesity,
26(4), 559-565.
Banwell, C. L., & Young, D. (1993) 'Rites of passage: smoking and the construction of
social identity'. Drug and Alcohol Review, 12, 377-385.
Barbeau, E. M., Leavy-Sperounis, A., & Balbach, E. D. (2004) 'Smoking, social class, and
gender: what can public health learn from the tobacco industry about disparities in
smoking?'. Tobacco Control, 13(2), 115-120.
Berlin, I., Singleton, E. G., Pedarriosse, A.-M., Lancrenon, S., Rames, A., Aubin, H.-J., et
al. (2003) 'Modified reasons for smoking scale: factorial structure, gender effects
and relationship with nicotine dependence and smoking cessation in French
smokers', Acü/cíz'oM, 98 (November), 1575-1583.
Bialous, S. A. (2005) "Women's health under fire: does it need to go up in smoke?' In I.
Kickbusch, K. Hartwig & J. List (Eds.) Globalization, Women, and Health in the
21st Century, London, Palgrave McMillan.
Blum, A. (2005) 'Tobacco in sport: an endless addiction?'. Tobacco Control, 14(1), 1-2.
Brandt, A. M. (1996) 'Recruiting women smokers: the engineering of consent', Journal
of the American Medical Womens Association, 51(1-2), 63-66.
Broom, D. H. (2005) 'Rethinking the genders of wellbeing'. In L. Manderson (Ed.)
Rethinking wellbeing, Perth, API Network.
Broom, D. H. (2001) 'Public health, private body', Australian and New Zealand Journal
of Public HeaUh, 25(1), 5-S.
Burton, P., & Phipps, S. (2004) Families, Time and Money in Canada, Germany,
Sweden, the United Kingdom and the United States, Halifax, NS, Department of
Economics, Dalhousie University.
Caballero, B. (2007) The global epidemic of obesity: an overview', Epidemiological
Reviews, 29(1), 1-5.
Camp, D. E., Klesges, R. C , &c Relyea, G. (1993) T h e relationship between body weight
concerns and adolescent smoking'. Health Psychology, 12(1), 24-32.
Castrucci, B. C , Gerlach, K. K., Kaufman, N. J., & Orleans, C. T (2002) 'Adolescents'
acquisition of cigarettes through noncommercial sources'. Journal of Adolescent
Health, 31(4), 322-326.
Collin, J. (2005) 'Opening a global goldmine: globalization, gender, and transnational
tobacco companies'. In I. Kickbusch, K. Hartwig & J. List (Eds.) Globalization,
Women, and Health in the 21st Century, London, Palgrave McMillan.
Connell, R. W. (2002) Gender, Cambridge: Polity Press.
Connell, R. W. (2003) 'Developing a theory of gender as practice'. Gender & Society,
17(3), 370-372.
Crompton, R., 8cMann, M. [Eds]. (1986) Gender and Stratification, Oxford, Polity
Press.
Crosnoe, R. (2007) 'Gender, obesity, and education'. Sociology.of Education, 80(July),
241-260.

24 Australian Journal of Social Issues Vol.43 No.1 AUTUMN 2008

Csete, J. (2005) 'A comfortable home: globalization and changing gender roles
in the fight against HIV/AIDS'. In I. Kickbusch, K. Hamvig & J. List (Eds.)
Globalization, Women, and Health in the 21st Century, London, Palgrave
McMillan.
Cutler, D. M., Glaeser, E. L., & Shapiro, J. M. (2003) 'Why have Americans become
more obese?', Journal of Economic Perspectives, 17(3), 93-118.
de Lauretis, T. (1991) 'Queer theory: lesbian and gay sexualities', differences: Ajournai
of Feminist Cultural Studies, 3(2), iii-xviii.
Delphy, C. (1981) 'Women in stratification studies'. In H. Roberts (Ed.) Doing Feminist
Research, London, Routledge & Keegan Paul.
Dixon, J., & Winter, C. (2007) "The environment of competing authorities'. In J. Dixon
6c D. H. Broom (Eds.) Seven Deadly Sins of Obesity: How the Modem World Is
Making Us Fat, Sydney, University of NSW Press.
Doak, C. M., Adair, L. S., Bentley, M., Monteiro, C , & Popkin, B. M. (2005) 'The dual
burden household and the nutrition transition paradox'. International Journal of
Obesity, 29(1), 129-136.
Eckersley, R. (2001) 'Losing the battle of the bulge: causes and consequences of
increasing obesity'. Medical Journal Australia, 174(June 4), 590-592.
Edwards, M. (1984) 'The distribution of income within households'. In D. H. Broom
(Ed.) Unfinished Business: Social Justice for Women in Australia, North Sydney,
George Allen & Unwin.
Friel, S., & Broom, D. H. (2007) 'Unequal society, unhealthy weight: the social
distribution of obesity'. In J. Dixon & D. H. Broom (Eds.) The Seven Deadly Sins
of Obesity: How the Modem World Is Making Us Fat, Sydney, University of New
South Wales Press.
Gatens, M. (1983) 'A critique ofthe sex/gender distinction'. InJ. Allen & P. Patton (Eds.)
Beyond Marxism? Interventions After Marx, Leichhardt, Intervention Publication.
Graham, H. (1994) 'Surviving by smoking'. In S. Wilkinson & C. Kitzinger (Eds.) Women
and Health: Feminist Perspectives, London, Taylor 6«: Francis.
Graham, H. ( 1996) 'Smoking prevalence among women in the European Community
1950-1990', Social Science & Medicine, 43(2), 243-254.
Greaves, L. ( 1996) Smokescreen - Women's Smoking and Social Control, Halifax,
Fernwood Publishing.
Greaves, L.J. (1993) The Meaning of Smoking to Women: Women, Smoking and
Identity. PhD thesis, Monash University (unpublished).
Hartman, H. (1976) 'Capitalism, patriarchy, and job segregation by sex'. Signs, 1(3 part
2), 137-169.
Hinde, S., & Dixon, J. (2005) 'Changing the obesogenic environment: insights from a
cultural economy of car reliance'. Transportation Research Part D, 10( 1), 31-53.
Honjo, K., & Siegel, M. (2003) 'Perceived importance of being thin and smoking
initiation among young girls'. Tobacco Control, 12(3), 289-295.

25
Gender ¡n/and/of Health Inequalities

Hooks, B. ( 1984). Ain't I a Woman? Black Women and Feminism, Boston, South End
Press.
Irwin, A., Valentine, N., Brown, C , Loewenson, R., Solar, O., Brown, H., et al. (2006)
The Commission on Social Determinants of Health: tackling the social roots of
health inequities', PLoS Medicine, 3(6), 749-751
Jacobsen, B. (1986) Beating the Ladykillers: Women and Smoking, London, Pluto Press.
Jeffery, R. W., Hennrikus, D. J., Lando, H. A., Murray, D. M., & Liu, J. W. (2000)
'Reconciling conflicting findings regarding postcessation weight concerns and
success in smoking cessation'. Health Psychology, 19(3), 242-246.
Klein, R. (1993). Cigarettes Are Sublime, Durham, Duke University Press.
Langenberg, C , Hardy, R., Kuh, D., Brunner, E., & Wadsworth, M. (2003). 'Central and
total obesity in middle aged men and women in relation to lifetime socioeconomic
status: evidence from a national birth cohort, Journal of Epidemiology and
Community Health, 57(10), 816-822.
Lennon, A., Gallois, C , Owen, N., ßcMcDermott, L. (2005) 'Young women as smokers
and nonsmokers: a qualitative social identity approach'. Qualitative Health
Research, 15(10), 1345-1359.
Loueiro, M. L., Sc Nayga, R. M., Jr. (2005) Obesity Rates in OECD countries: An
international perspective, EAAE XI Congress, Copenhagen.
Macintyre, S., &c Hunt, K. ( 1997) 'Socio-economic position, gender and health: how do
they mteT2ict}\Joumal of Health Psychology, 2(3), 315-334.
Matthews, S., Manor, O., &C Power, C. ( 1999) 'Social inequalities in health: are there
gender differences?'. Social Science & Medicine, 48(1), 49-60.
McDermott, L., Russell, A., & Dobson, A. (2002) Cigarette Smoking among Women in
Australia, Canberra, Commonwealth Department of Health and Ageing.
McLaren, L. (2007) 'Socioeconomic status and obesity', Epidemiological Reviews, 29(1),
2948.
Mellstrom, U. (2003) 'Men, gender and the state'. In S. Ervo & T.Johansson (Eds.)
Among Men: Moulding Masculinities (Vol. 1), Aldershot, Ashgate.
Aiitra, A. (2001) 'Effects of physical attributes on the wages of males and females'.
Applied Economics Letters, 8(11), 731-735.
Molarius, A., Seidell, J., Sans, S., Tuomilehto, J., Sc Kuulasmaa, K. (2000) 'Educational
level, relative body weight, and changes in their association over 10 years: an
international perspective from the WHO MONICA Project', American Journal of
Public HeaUh, 90(8), 1260-1268.
Mummery, W. K., Scho&eld, G. M., Steele, R., Eakin, E. G., & Brown, W. J. (2005)
'Occupational sitting time and overweight and obesity in Australian workers',
American Journal of Preventive Medicine, 29(2), 91-97.
Offer, A. ( 1998). Epidemics of Abundance: Overeating and Slimming in the USA and
Britain Since the 1950s, Nufñeld College, University of Oxford.

26 Australian Journal of Social Issues Vol.43 No.1 AUTUMN 2008

Ostry, A. S., Radi, S., Lotiie, A. M., & LaMontagne, A. D. (2006) 'Psychosocial and other
working conditions in relation to body mass index in a representative sample of
Australian workers', BMC Public Health, 6(53).
Patton, G. C , Carlin, J. B., Coffey, C , Wolfe, R., Hibbert, M., &c Bowes, G. (1998) 'The
course early smoking: a poptilation-based cohort study over three years'. Addiction,
93(8), 1251-1260.
Petersen, A. (2003) 'Research on men and masculinities: some implications of recent
theory for future work'. Men and Masculinities, 6(1), 54-69.
Plumridge, E. W., Fitzgerald, L. J., & Abel, G. M. (2002) 'Performing coolness: smoking
refusal and adolescent identities'. Health Education Research, 17(2), 167-179.
Ridgeway, C , 6<: Smith-Lovin, L. (1999) 'The gender system and interaction'. Annual
Review of Sociology, 25,191-216.
Rodriguez, D., Sc Audrain-McGovem, J. (2004) 'Team sport participation and smoking:
analysis with general grovvah mixture modeling',/oMrwa/ ofPaediatric Psychology,
29(4), 299-308.
Salmon, J., Owen, N., Bauman, A., Schmitz, M. K. H., &c Booth, M. (2000) 'Leisure-time,
occupational, and household physical activity among professional, skilled, and less-
skilled workers and homeimkers', Preventive Medicine, 30(3), 191-199.
Scott, J. W. (1988) 'Deconstructing equality-versus-difference: or, the uses of
poststructuralist theory for feminism'. Feminist Studies, 14(1), 33-50.
Seidell, J. (2000) 'Obesity, insulin resistance and diabetes: a worldwide epidemic', British
Journal of Nutrition, 83(1), 5-8.
Sobal, J., & Stunkard, A. J. (1989) 'Socioeconomic status and obesity: a review of the
literature. Psychological Bulletin, 105(2), 260-275.
Starr, M. E. (1984) 'The Marlboro Man: cigarette smoking and masculinity in America',
Journal of Popular Culture, 17(4), 45-57.
Thiri Aung, A., Hickman, N. J., 8c Moolchan, E. T. (2001) 'Health and Performance
Related Reasons for Wanting to Quit: Gender Differences Among Teen Smokers',
Substance Use & Misuse, 38(8), 1095-1107.
Turrell, G., & Mathers, G. D. (2000) 'Socioeconomic status and health in Australia',
Medical Journal of Australia, 172(9), 434438.
Wang, Y, & Beydoun, M. A. (2007) 'The obesity epidemic in the United States-gender,
age, socioeconomic, racial/ethnic, and geographic charaCTeristics: a systematic
review and meta-regression analysis', Epidemiological Reviews, 29(1), 6-28.
Wardle, J., Waller, J., ôcjarvis, M. J. (2002) 'Sex differences in the association of
socioeconomic status with obesity', American Journal of Public Health, 92(8),
1299-1304.
Waring, M. ( 1988). Counting for Nothing: What Men Value and What Women Are
Worth, Wellington, NZ, Allen 6c Unwin.
Williams, J. (2001) 'From difference to dominance to domesticity: care as work, gender
as tradition', Chicago-Kent Law Review, 76, 1441-1493.

27
Gender ¡n/and/of Health Inequalities

Yancey, A. K., Leslie, J., & Abel, E. K. (2006) 'Obesity at the crossroads: feminist and
public health perspeaives', Signs: Journal of Women In Culture and Society, 31(2),
425-443.
Yu, Z., Nissinen, A., Vartiainen, E., Song, G., Guo, Z., Zheng, G., et al. (2000)
Associations between socioeconomic status and cardiovascular risk factors in an
urban population in China', Bulletin of the World Health Organisation, 78(11),
1296-1305.
Zagorsky, J. L. (2005) 'Health and wealth: The late-20th century obesity epidemic in the
U.S', Economics & Human Biology, 3(2), 296-313.

28 Australian Journal of Social Issues Vol.43 No.1 AUTUMN 2008