book laments, much of the discussion of health inequality in. France at present .... chapter on cardiovascular disease includes a re-hash of the major risk factors ...
© International Epidemiological Association 2001
International Journal of Epidemiology 2001;30:628–631
Printed in Great Britain
Essay Review Health Inequalities in France M BARTLEY Les Inégalités Sociales de Santé A Leclerc, D Fassin, H Grandjean, M Kaminski, T Lant. Paris: INSERM/La Decouverte, 2001, pp. 448, 225FF. A series of major changes in the organization of employment in France, as in other nations, over the period between 1980 and the present has given rise to a new concern for health inequality. One of the most striking policy changes has been the need to extend cost-free access to medical care to large numbers of people whose position in the labour force and in society has become ‘precarious’. Although the great majority of the French population are covered by some form of medical insurance, visits to the doctor and prescriptions must be paid for, at least in part, at the time of need and reclaimed afterwards. Between 70% and 80% is reimbursed by the state and the rest is covered by medical insurance taken out jointly by the person and the employer. This leads to a bias against the unemployed, SDF (‘sans domicile fixe’, people with no stable residence) and other similar groups. Once unstable employment increased, the system began to fail larger numbers of people. Since 1998, therefore, new health insurance laws and even some new institutions have been put in place to deal with these ‘excluded’ groups. It is therefore understandable that, as the introduction to this book laments, much of the discussion of health inequality in France at present concentrates more on the plight of the ‘socially excluded’. This is to the detriment of efforts to understand the ‘gradient’: the differences in health which extend from top to bottom and differentiate the life chances of the most privileged occupational groups not just from the most oppressed, but from the next most privileged group as well. In their introductory chapter, Fassin et al. make an elegant job of connecting the changes in the wider society, which have increased social marginalization, to wider issues of social structure. They highlight the political gains to be had by encouraging the public discourse on growing inequality to be held in terms of ‘social exclusion’—leaving everyone who can just scrape together the money for that second hand Mondeo to feel they are not the ones being left behind. This theme, of the ways in which concepts of exclusion and marginalization detach discussion of the most extreme manifestations of inequality from wider social and economic policy debate, persists throughout much of the book. This is a very long book with a large number of shortish chapters, an entirely appropriate format for its objective of giving wide access to as much information on health inequality as possible. But the reviewer risks droning on from one topic to another. An alternative is to group the chapters more or less according to their topic matter rather than their (often rather strange) order. The 27 chapters fall roughly into three types: those which review international epidemiology evidence; those which describe specifically French health inequality; those which reflect on historical and policy aspects. Within the
epidemiological chapters there are again two types: those which are mainly descriptive, or which follow a conventional risk factor epidemiology format, and those which combine evidence from studies with more innovative theoretical approaches. There is only one chapter dealing with methodology, and this focuses on qualitative methods.
Historical and policy aspects of health inequality The first chapter by Bourdelais whets the appetite by showing data from life tables for Geneva and Rouen in the 17th and early 18th centuries. However, this turns into a simple, if most interesting, set of relative risks, with a brief discussion on the ways in which health inequality arrived on the agenda of media and political debate in the 19th century that could have come from any British or American social policy text. This policy history is extended into the period following the second world war in chapter 2. Here Drulh points out the irony of a health policy that aims to give the great majority of the population, regardless of income or class, access to a freely chosen doctor, in that this tends to give support, both financial and ideological, to a purely individualistic view of health. The general public, in France as in the UK and USA, perceive access to medical care as a far more important issue than health inequality. In making large amounts of public money available to underwrite access to care, French governments, he argues, have reinforced this. In her chapter on the relationship between research and policy, Berthod-Wurmser asks why has Great Britain been the leader in health inequality research for so many years? It is an interesting question, of the kind that puts the more waffly types of policy case studies on their mettle. She answers it elegantly. In Great Britain, after the introduction of the National Health Service, it became clear that equity in access to health care was failing to reduce inequalities in outcome. In addition, the policy instruments used to allocate resources to health care providers forced planners to take account of differences in demand between social and regional groups. As a test of the plausibility of this interpretation, Berthoud-Wurmer takes a second case study, that of The Netherlands. Before 1986 there was no debate on health inequality in The Netherlands. After this time, individuals were given a wider choice of which health insurance organization they wished to belong to. In principal, anyone may choose their insurance provider on the basis of which services they wish to receive. However, insurance companies are now also allowed to refuse applications (in contrast to the older system, more similar to that in France, in which most people were covered by providers that were obliged to accept any member of the relevant population or occupational group). In order to prevent ‘cherry picking’ of the healthiest customers, state funding reimburses the insurance companies, but only for the costs of ‘essential’ medical care. The amount of funding received depends,
628
ESSAY REVIEW
however, on a formula giving the ‘risk class’ of each patient. It is a sort of individual level equivalent to the Deprived Area supplement given to British general practitioners, paid instead to the insurers. In calculating these levels of risk, the different amounts of ‘essential’ treatment likely to be needed by people in different social positions became evident. Beginning in 1990, The Netherlands has produced some of the most influential work on health inequality done outside the UK. France has not yet introduced this distinction between essential and nonessential care, and insured individuals (still the great majority) continue to be reimbursed by their health insurance without judgements of need. It is, in other words, the need for health care rationing which has given rise to literature on ‘health inequality’—a great irony which is little commented on in the UK itself. In reviewing the evidence on health inequality in older people, Grand et al. give us a useful summary of the way in which pensions are administered in France. The chapter by Lombrail, on access to health care takes aim at the idea that French citizens do enjoy universal health coverage. Although in principal no person is denied care for a health emergency, the author finds data indicating major inequality in access particularly to dental and preventive services. In a chapter on the general trends in inequality in French society, Bihr and Pfefferkorn show the efforts of neo-liberal austerity policies. After a steady rise between 1973 and 1979 net real incomes stagnated in the private sector and fell in the public sector until 1985. Between 1985 and 1993 while the average unskilled workers wage rose by under 0.7%, the average real income received from savings and investments rose by 59.4%, 85 times greater. Household equivalent disposable income between 1984 and 1994 rose from 173 000 Fr to 240 000 in professional households, from 152 000 to 172 000 in managerial households, but fell from 60 000 to 57 000 in the households of unskilled workers.
Health inequalities in France Descriptive chapters The chapter by Jougla et al. on ‘Mortality’ acts as a useful introduction for those not already familiar with it to the French Longitudinal Study. Social class and education were recorded in 1975, and two follow-up studies are reported here. These results have also been reported elsewhere, and show the expected differences between two broadly defined social classes and educational groups. The second study reported uses unlinked data: denominator populations taken from the censuses of 1982 and 1990. Here class differences in each major cause of death are given as directly standardized rates per 100 000 so that changes over the two time periods can also be seen. It shows, for example, that social inequality in the decline in mortality from cancers has been far greater than that in the decline in mortality from cardiovascular disease over this time period. Although there are very large inequalities in mortality in France, inequality in morbidity does not differ as much from other European nations, according to Girard et al. in their chapter on global health indicators. They end with a call for more homogeneous measures of morbidity in European health surveys. Lang and Ribet’s chapter on cardiovascular disease includes a re-hash of the major risk factors such as lipids, tobacco, Karasek’s work strain concept, unemployment, depression, social support and low birth weight.
629
They do point out that models of causality are not well developed in this literature, and that the presentation of large numbers of variables, all adjusted for one another, does not reveal chains of causality and may even obscure them. There are also businesslike reviews of the scanty evidence on inequalities in dental health and AIDS in France. As in other countries, AIDS incidence has fallen among gay communities (who still suffer the major burden of this disease) while it has risen among injecting drug users. This is having a predictable effect of introducing social inequality (or at least making this more evident) in the incidence of AIDS. It is pointed out that new therapies will also increasingly raise the question of unequal access to treatment. At present there are no data on this in France. Little information exists in France on health of children, despite regular monitoring of perinatal health and recent special studies on health of adolescents. In France, social protection measures for families with children are highly developed but it is hard to know if these have any effect on health. Tursz brings together bits of evidence from a wide range of data sources such as a continuous government survey of prescriptions; vaccination records; and school medical services. No class breakdown is available. The nearest one can get to social environment is studies of schools in educational priority areas, where many children are known to live in difficult circumstances. The number of children on at risk-registers in France has increased frighteningly, from 35 000 in 1992 to 82 000 in 1997. The authors admit that the casualization of work may have contributed to fragmentation of all social relationships, and that the psychological pressures applied to children by mid-ranking parents who ‘see their children as a way to rise up the social ladder’ are a hidden form of abuse.
Theoretical chapters The first of the chapters that both present evidence from French studies and reflect on that evidence in a less conventionally epidemiological manner is that by Kaminsky et al. They summarize what is known on inequality in perinatal health using data from the most recent special linkage study carried out in 1984 to 1989. They reveal that there is no social gradient according to the occupationally defined social class of the mother, and no difference in gradient according to marital status. The only social difference in mortality risk according to the socioeconomic circumstances of the mother is between economically active and inactive mothers, the babies of active independent mothers have around 30% lower risk than inactive, and those of active partnered mothers having around 20% lower risk. Kaminsky et al. also review new studies which have focused on ‘collective determinants’ of perinatal health such as the unemployment rate, the % of families living beneath the poverty level, mean income and education, and crowding. Choquet and Lagadic in their chapter take an even more innovative approach on ‘Adolescent health’. The present data from an unusual study from a sample of schools, with classes sampled within schools. Two kinds of school were included in the study: colleges or ‘lycees d’enseignment generale et technique’ (schools which allow progress to higher education), and ‘lycees professionels’, which do vocational courses. Analysis was carried out separately on children aged 12–15 and 16–19. Only a flavour of the fascinating findings can be given here, but these
630
INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
include: boys are far less likely to have nightmares than girls at ages 12–15 but the class gradient is stronger in boys. At 16–19 nightmares have diminished in boys but increased in girls, though gradients have decreased in both. Frighteningly, 7.5% of 12–15 year old sons of blue collar workers; (4.2% of sons of managers and professionals) and 6.8% (5.5%) of girls have made a suicide attempt at some time. This relatively high index of emotional ill health in middle class teenagers adds, the authors feel, to the evidence in agreement with West, that health differentials in adolescence are not strong. It also relates interestingly to the comment in Tursz’s chapter about emotional abuse aspiring middle class parents in an increasingly competitive social environment. Ravaud and Mormiche’s chapter on disability rightly stresses the overwhelming importance of social inequality in this sphere. Less privileged social groups suffer most from long-term disabling consequences of illness for two reasons: aetiological and economic. Not only is incidence of most disabling illnesses and accident higher in these groups, but their lack of economic and social power means that such illness as does occur will be less well treated and is more likely to lead to exclusion from waged work. The authors summarize the Wood model of disease, disability and handicap, but argue that it is more accurate to think of a continuum of disability which applies to all members of society. Those of us ‘able’ enough for some purposes will be disabled for other purposes if we count the ability to give constant care and support to others, or to sustain friendships and love relationships over long periods of time as well as climbing mountains. In their chapter on work, Volkoff and Thebaud-Mony question whether we really do belong to a new era in which heavy labour has disappeared to be replaced by a ‘knowledge society’. New technology has by no means assured a ‘natural evolution’ of work towards more benign conditions. Fragmentation of work careers, out-sourcing and large-scale sub contracting means that there are large numbers of peripheral workers who do the same work as their ‘core’ colleagues but under very different conditions in respect of job security, health and safety protection, and the right to negotiate wages and conditions. Most waged workers are under the constant threat of layoffs. Workers unable to keep up with the pace, who would in the past have been moved to less arduous if worse paid jobs, are now on the way to joining the socially excluded. Volkoff and Thebaud-Morny thus provide an excellent model for understanding the effects of work on health. It acknowledges that these are reciprocal and that health inequality is produced by these interactions over the life course. However they acknowledge that existing health statistics are not very well suited to this kind of analysis.
International reviews of health inequality Descriptive chapters It would have been a major omission to produce a volume such as this without situating French health inequality within the rest of Europe, especially as France compares rather badly to most other European nations, and in view of the quality of the very difficult multi-national comparative studies that Kunst and his fellow-workers in the European Working Group on Health Inequality in European Countries have been conducting. Readers of the health inequality literature will find nothing new
here although it provides a valuable summary for Francophone social scientists and epidemiologists. Hunt and Macintyre provide another useful review of international evidence, this time on gender differences in health. They point out that countries differ widely according to how typical it may be for women to be economically active and in which types of occupation. Brixi and Lang’s chapter on health-related behaviours falls between the ‘descriptive’ and ‘theoretical’. They remind us that ‘behaviours’ have to be understood as part of a far wider complex of advantages and disadvantages, pleasures and pains, costs and benefits. Rather disappointingly, they then go on to show a conventional set of tables with risk behaviours varying according to socioeconomic position of a type that could occur in any health education text. However they also point out the ways in which ‘risk factor epidemiology’ helps to keep the policy discourse on health inequality focused at the individual level. Kawachi and Berkman’s chapter is also a review of US and other literature on social support, a useful service for French-speaking readers. There is a brief review of ‘animal models’ by Berton, and Mormede, which is no different to many others in the Anglophone literature.
Theoretical reviews The first of the more theoretical reviews of international data is the chapter on cancer by Herbert and Launoy. In France, as now increasingly in other industrialized countries, cancer rather than heart disease is the largest single cause of mortality. One wonders what impact this epidemiological transition is going to have in the longer term on the whole discourse on social causation. Herbert and Launoy ask whether the observed social inequalities in cancer are more associated with incidence or survival, concentrating on female breast, lung and colorectal, the three most common sites. They show that whether or not there is a social gradient in incidence, gradients exist for survival, the more so in the more treatable forms of cancer. They do not attribute differences in either incidence or survival to single causes. Not all researchers are agreed on the relative importance of occupational exposures for lung cancer for example, or of non-parity and late childbearing for breast cancer, which has higher incidence in more socio-economically privileged women. Likewise in survival, the roles of the availability, cost and attitudes to use of screening services need to be considered as do differences in the quality of care offered to more and less privileged people. Another possibility is that those with lower levels of income and social power are less able to benefit from treatment due to poorer nutrition and higher levels of stress. For those without the benefit of paid sick leave, taking time off for lengthy treatment will result in the disappearance of regular income from employment. Their ability to keep warm and well fed will therefore depend on the level of social benefits available. Thus the situation of the building worker with a cancer diagnosis, even of a more treatable type, will be dramatically different to that of a senior manager or government official, who can undergo months of debilitating treatment with no impact on his/her standard of living whatsoever. However, these interesting observations are made on the basis of foreign research, as there is very little of this kind in France. In her highly sophisticated chapter on inequalities in mental health, Lovell takes a critical view of the measurement of both ‘class’ and ‘mental health’. She points out that most research
ESSAY REVIEW
does not distinguish between two types of measure of socioeconomic position: those based on ‘class’ and those based on ‘status’. Rather, occupationally based measures of class such as the French CSP (categorie socio-professionel), education and a variety of ‘socio-economic status’ (SES) classifications are used interchangeably and indiscriminately. She argues that this usage is underpinned by a functionalist notion of social stratification. According to this approach, the inherent capacities of individuals allow them to move between social strata by determining how much education, money, prestige, etc. they can accumulate. In contrast, the concept of social class is based on a conflict rather than a consensus model of society, in which those with different relationships to the structure of wealth ownership and employment have different degrees of power to accumulate social advantages. Class and ‘SES’ are therefore, Lovell argues, not the same thing, and it is possible to show that class measures relate more strongly to mental health outcomes than status. This Lovell offers as evidence that the true relation of social position to mental health is the link with aspects of class such as control and autonomy, rather than with relative position in a social hierarchy per se. Fassin’s chapter on qualitative approaches to health inequality is a rare and welcome exception in the literature of any nation. He points out that relevant work is to be found in urban sociology or social anthropology, in particular studies of small communities that have been carried out by the French school of urban ethnology can throw light on health inequality. While economics and epidemiology deal with the macrosocial processes that result in certain levels of income and certain risks of disease, it is, he argues, anthropology, sociology and perhaps psychoanalysis which are better suited to discerning ‘l’inscription de l’ordre social dans les corps ou … l’incorporation de l’inégalité’ [p. 135] One of the many strengths of the qualitative approach is that it enables us to explore what may be the different meanings given by different actors to the ‘same’ situation, as defined in macrosocial terms like class, marital status or employment status. One relevant example is the work of Becker on the difference between taking drugs now and then and ‘becoming a user’. One can apply this to smoking—policy makers might ponder on how best to prevent the young person who experiments or has occasional cigarettes with her friends proceeding to the
631
‘secondary deviation’ of defining herself as ‘a smoker’. There is no way to illuminate this process of self-redefinition other than by qualitative research. Finally, Fassin et al. reflect on ‘Future prospects for research and action’. The new law in 1998, which intended to improve access to care to the large and growing number who had no health insurance was focused on the perils of social exclusion. But the editors have concentrated on ‘the gradient’ of health throughout society. Too much concentration on social exclusion, they persuasively argue, ignores the social dynamics that conduct people into and out of this state, and risks treating it as a stable condition. It allows us to ignore too readily the processes by which work had been made more precarious and insecure. In gathering the material for the book they admit to having realised that an ‘inequality’ perspective does not take on board the dynamic processes over time any more than a ‘social exclusion’ perspective does. They set out four priorities for future research. 1) Better measures of both illness outcomes and socioeconomic input variables. 2) More attention to macro-social variables such as income distribution, social capital. 3) More attention to inequality in access to care. 4) More attention to inequality in impact of illness. They also call for more attention to early life, adolescence and old age. They call for a health inequality impact statement for all new policies, health or otherwise. Also for a wider debate on health inequality and what their existence means for the whole of society. In summary, the book contains many valuable insights—a shame that non-French speakers will not have access to these. Some might wonder why the book was written in French rather than English, the ‘lingua franca’ of science. Those who know France, and its citizens’ pride in their language, would be less surprised. Our colleagues from France, The Netherlands, Sweden, etc. usually have at least a reasonable command of two or more other languages in addition to English. In my personal opinion it is an embarrassment, particularly in the social sciences, that so few of us even have a reading knowledge of anything other than English. Rather than forcing everyone into a linguistic strait jacket, it would be preferable for international journals to commission translations of important papers, and for the rest of us to regard knowledge of languages as an integral part of the work of scholarship.
