In this article we explore systematically the different conceptions of health equity in key national health policy documents in the United States, the United Kingdom, and France. We find substantial differences across the three countries in the characterization of group differences (by SES, race/ethnicity, or territory), and the theorized causes of health inequalities (socioeconomic structures versus health care system features). In all three countries, reports throughout the period alluded at least minimally to inequalities in social determinants as the underlying cause of health inequalities. However, even in the reports with the strongest attachment to this causal model, the authors stop well short of advocating the redistribution of power and resources that would likely be necessary to redress these inequalities.
The health enjoyed by members of groups defined by economic, racial and ethnic, geographic, or gender differences, among others, can vary dramatically. Readers of this journal are well aware of some of the gaps in life mortality, morbidity, and access to health care between, for example, African Americans and white Americans, those living in rural and urban areas, or people with a college degree and those with less than high school education. Given this multiplicity of inequalities in health status and health care, how do governments choose to frame the issue of health equity, identify the underlying causes of inequalities in health, and craft appropriate policy responses? In this article we analyze major government reports from the United States, the United Kingdom, and France over the last thirty years to identify common trends and differences in these governments' approaches to health equity. Comparative analysis reminds us that the approach taken in any one setting is not an inevitable outgrowth of the fact of group differences in health. Instead, how governments frame health equity as a problem for politics, and the policies they advance to try to address that problem, are contingent. While the weight of history, institutions, and political cultures may tip the balance toward one framing or another, political communities can and do make choices about what to do in the face of inequities in health. The present analysis strives to clarify the choices that we have already made, and to offer, through comparative analysis of countries that are similar to ours, a broader vision of what might be possible in the future.
Policy elites at the international level and in European countries often use the term health inequalities more or less interchangeably with health inequities. The European Union, for example, defines health inequalities as “preventable and unjust differences in health status or in the distribution of health determinants between different population groups.” Similarly, in French policy discourse, the term “social” in the apparently neutral phrase inégalités sociales de santé (ISS) references health differences linked to socioeconomic position, which are generally regarded as unjust. In other words, in the discourse of European policy makers and researchers, the term health inequality has come to denote a policy problem (Bardach 1996) that has been identified as such at least partly because it is seen as inequitable. In the United States, by contrast, discussions of group differences in health more frequently use the term disparity, which in many instances does not include a fairness judgment.
Our findings suggest that the difference in language and framing does not simply reflect the fact that research communities in the United States and Europe have developed somewhat independently of each other. American and European policy communities conceptualize the problem of health inequalities differently, and in ways that are related to the causal models and policy solutions that are advanced in different countries to address health equity. However, we find that in all three countries, the policy recommendations that are espoused in government documents fall short of the remedies implied by understandings of the processes generating inequity in health. This suggests that reducing health inequalities may be even more politically difficult than researchers generally think; but it also shows that there is room for a reframing of health disparities in the United States to accommodate a broader scope of policy responses.
In most of the international and cross-national research literature in epidemiology and public health, scholars measure or attempt to explain health inequalities that are related to differences in socioeconomic status (SES, usually operationalized as income, type of occupation, or educational attainment). In the international policy community, researchers and activists characterize health inequalities as preventable, unfair, and hence deserving of policy attention to the extent that they result from underlying social inequalities that are produced by our political, economic, and social systems (CSDH 2008; Whitehead 1991). Yet, differences in health status across social groups defined by characteristics other than SES—for example, by ethnicity or race; gender; disability status; or geography—could also be considered preventable and unjust, depending on the circumstances. So, while important policy documents produced by the World Health Organization (most recently, the 2008 report of WHO's Commission on the Social Determinants of Health, chaired by Englishman Sir Michael Marmot) frame health equity primarily in terms of SES, there is nothing “natural” or inevitable about this framing.
Consider how a recent report from the Centers for Disease Control (CDC Health 2015) presents infant mortality rates. Differences are listed by race and ethnicity (11.3 African American infant deaths per 1,000 births, compared to 5.1 white infant deaths; 5.1 for Hispanics, 8.1 for American Indian or Alaska Natives; and 4.2 for Asian or Pacific Islanders), and by state and territory (ranging from 11.1 in American Samoa to 4.2 in Massachusetts) (CDC Health 2015, table 12). Data on infant mortality by income level or occupational status are noticeably absent.
Of course, different forms of inequalities may overlap, and when we decide to disentangle them it entails choices that are ultimately conceptual and political, rather than merely technical. Take as an example the relationship between racial (black-white) and socioeconomic health inequalities in the United States. Because race and SES are correlated but not completely, it is possible to estimate the “effects” of race and SES on health separately using observational data. Scholars have debated whether this is a desirable practice, however. Williams (1999) argues that many traditional measures of SES have distinctly different effects on whites and African Americans, so that controlling for SES can “account for” much of the observed racial disparity in health status while substantially underplaying the role of institutional and individual forms of racism on health status. Reed and Chowkwanyun (2012) go one step further, demonstrating that the practice of identifying and measuring racial disparities in health is itself rooted in the development of American social science and policy research, where some of the more complex relationships between race and class have been buried under a dominant narrative about durable racial inequalities.
By framing inequalities as either “about” SES or “about” race or some other group, policy makers can simplify, or even perhaps obscure, the complexity of the overlapping boundaries between social groups that are defined in different ways. Policy frames involve a definition of the social problem, a causal story about where that problem comes from, and a policy prescription—each of which may invoke a moral evaluation that determines who is responsible for causing and/or treating the problem (Entman 1993; Stone 1989; Verloo 2005). How policy frames are constructed and employed therefore has implications for politics—who is to blame, who is responsible—and can be expected to shape the policies intended to address the problem.
We observe in this article that health equity frames vary across countries. A systematic, comparative analysis of how national policy elites frame differences in health status serves to denaturalize health inequities, and to make more easily visible some of the institutions, ideas, and interests that shape contemporary definitions of inequality. To our knowledge, only one other study adopts a comparative approach to the discursive construction of health equity that includes the United States (Docteur and Berenson 2014; for comparative work not involving the United States, see Lynch 2016; Vallgårda 2007,1). Docteur and Berenson (2014) compare American health equity frames to those employed in European Union policy documents, but the EU in fact has little competence over the health and social policies of its member states. We instead compare the frames employed by authors of health policy documents in the United States and those that appear in documents produced by actors with the authority to affect public policies related to health in the UK and France. We conducted a systematic sampling of government documents—that is, documents produced by policy makers with the intention of shaping public policy—to examine these differences across the three cases.
Research Design and Methods
Frame Analysis of Government Reports
Discourses or conversations about policy problems, like discussions about anything else, are inherently incomplete. Frames do their work by activating schema in the minds of recipients of messages, who then use these schema unconsciously to “fill in the blanks” between elements of a problem in order to construct responses that seem reasonable (van Gorp 2007). Within the arena of public health, policy frames can affect whether an issue gets on the agenda (Stone 1989), shape policy responses to a problem (e.g., Kenterelidou 2012; Saguy and Riley 2005), and affect public beliefs and attitudes about policy choices, including policies related to health inequalities (e.g., Gollust and Lynch 2011; Lynch and Gollust 2010; Rigby et al. 2009).
Tracing how frames emerge can also reveal which actors in a policy field possess sufficient material and symbolic resources to impose their framing of an issue, and show how social institutions channel those resources. In other words, studying framing allows us to “tie [ . . . ] problem definitions to an analysis of power” (Vliegenthart and van Zoonen 2011: 108). A comparative analysis of the framing of health inequalities helps to denaturalize health inequalities, making it clear that they are not just facts “out there” to be measured and dealt with, but rather actively constructed as policy problems. We hope that this can encourage a deeper engagement with dimensions and causes of inequalities that might otherwise be missed, and may shed light on new ways to combat inequities.
We define health inequality policy frames as problem definitions, causal stories, policy prescriptions, and attributions of causal and treatment responsibility that can be detected in the form of key words, phrases, and analytic tropes in texts that contain the observations of the policy elites (researchers, policy advocates, bureaucrats, and elected officials) who constitute the health policy field.
We focus in this article on reports issued by national governments that deal either exclusively or in part with the problem of health inequalities, in order to identify the constitutive claims that national policy-making elites make about health equity: what qualifies as a health inequity, and why. Government reports are a standard source of information about how health policy elites understand health inequalities at a definitional level, and hence what policies are likely to be adopted in order to combat them (see, e.g., Docteur and Berenson, 2014; Graham 2004; Vallgårda 2007). Freeman (2006) articulates clearly the rationale for choosing to analyze these documents: “Government is a text-based medium, no less in public health than in other areas of public policy, and a feature of the politics of health equity across countries is that it turns on the production of a key text” (Freeman 2006: 52). The process of producing government reports on health inequalities helps to build constituencies for particular ideas and policies within the policy elite, as contributors negotiate over common language; the documents themselves become “a source of authority, a means by which influence is established and exerted, such that the production of the document may be thought of as a process of underwriting as much as writing” (Freeman 2006: 54; see also Raphael 2011).
Sampling: Country Cases
We selected three country cases for this comparative analysis. The United States, by virtue of the nature of this special issue, is the reference case. The UK in many respects constitutes a “most different system” comparison (Przeworski and Teune 1970). In 1998 the former British National Health Service was devolved into four autonomous units in England, Wales, Scotland, and Northern Ireland. Since then, there has been some divergence in the systems, particularly between Scotland on the one hand, and England and Wales on the other. The constituent countries of the UK nevertheless still all have single-payer, public national health systems that incorporate preventive and public health as well as medical care, and that provide medical care free or at low cost at the point of service to all legal residents of the UK and other European Union countries. As a result, policy makers in the UK may be more likely to consider the health system as an adequate tool for addressing health status inequalities, while at the same time inequalities in access to health care are unlikely to play as central a role in health inequalities discourse there as they do in the United States. Despite these differences, though, many policy areas relevant to health lie outside of the health system; and the UK is an important case for comparative analysis because of its outsized influence on the health inequality policy agenda throughout Europe.
We also consider the French case, which, in relation to the United States and in contrast to the UK, is closer to a “most similar system.” Like the United States, France has a health system in which public health and prevention plays a very limited role compared to insurance coverage for medical care. Multiple public and private payers and providers constitute the medical care system, which, as in the United States, has led policy makers to be preoccupied with issues like the uneven supply of medical services, uninsurance, and cost-related barriers to accessing health care. Another important similarity between the United States and France for the purposes of this study is the political salience of race and ethnicity, which is clearly higher in the United States and France than in the UK (Crowley 1993; Maxwell 2012). Analyzing the framing of health inequalities in the United States and France together thus allows us to see how the politics of race intersects with health inequality policy frames.
In order to assess the features of different health inequalities frames in policy discourses in the United States, the UK, and France, we conducted systematic qualitative content analysis, described below, of a sample of reports written or commissioned by the national government of a country that are mainly concerned with the issue of health inequalities during the period 1980 to 2012. This period begins with the release of the Black Report in the UK, which is widely recognized in the secondary literatures as the starting point for national-level policy attention to health inequalities in the industrialized democracies of the West. We end in 2012 in order to ensure that our search strategy, which depended in part on the secondary literature, would capture all of the relevant documents produced in that period.
To identify the universe of relevant documents, we first surveyed the secondary literature on public health and health policy in each country to construct a timeline including all mentioned government or government-sponsored publications; and we searched the websites of (a) the national health ministries, (b) any subsidiary organs that these ministries' websites linked to, and (c) the government publications offices to identify any additional policy documents related to health inequalities. Finally, we conducted Google searches for documents whose titles included the word “health” and variants on “inequality,” “disparity,” “difference,” “divide,” and “gap” in order to identify any documents that might have been omitted based on the literature and government website searches. From the lists of documents generated using these strategies, we then selected documents that met the following criteria: They are (1) “major” reports, that is, commissioned or released by the top level of the organization in question, rather than by a subsidiary department; AND (2) they are primarily concerned with health inequalities, that is, (a) the title contains the term health inequalities or health disparities, OR (b) the bulk of the report is dedicated to the problem of health inequalities. We also included major sections of general reports on the health system that were commissioned directly by the health minister, the national executive, or legislature and that met criterion 2(a) above. We eliminated any publications that met the above criteria but that were not “unified,” that is, they were constituted by individual chapters on diverse topics relating to health inequalities and attributed to separate authors. Table 1 lists the sample of reports for each country.
We identified the publication's date and producers (e.g., political appointees, national biomedical research centers) in order to contextualize the report and situate it within the broader stream of health policy and political events in each country. We then conducted a systematic close reading of each report according to a set of preestablished criteria ( Appendix). Our reading focused on whether health inequalities referred, in each report, to differences in health status and/or differences in health care (outcomes); over which groups (e.g., by socioeconomic status, racial and ethnic categories, gender, disability status, geographic designations) the inequalities occurred; and whether different frames were associated with different language, for example, “difference,” “inequity,” and “disparity”—terms that convey different degrees of moral charge. We next turned to examining the underlying causal explanations and proposed policy remedies for health inequalities. Causal explanations for health inequalities tap into policy makers' understanding of the true drivers of health equity. Policy recommendations indicate the practical response of governments, but could be contingent on particular factors such as program renewals, the fiscal climate, or the proximity of targets. Summaries of the results of the analysis of the text and figures in these reports are presented in tables 2 and 3.
We begin our analysis of the framing of health inequalities with the landmark British document released in 1980, the Black Report. Commissioned by a Labour government, the Black Report was delivered to the incoming Conservative Thatcher government. Under the Conservatives there were no significant government or commissioned reports on health inequality, and the issue was only taken up again in 1998 when the incoming Labour government commissioned the Acheson Report. Over the course of the next ten years, Labour's health ministry followed up on the Acheson Report with a series of documents outlining a policy response and tracking progress toward the goal of reducing health inequalities. In 1999 the government released the Saving Lives public health white paper and outlined a policy program aimed specifically at health inequalities in Reducing Health Inequalities: An Action Report, and in 2002 the government established an inter-ministerial “Cross-Cutting Review” to summarize progress to date and outline a long-term policy agenda. The 2003 document Tackling Health Inequalities: A Programme for Action, which included a foreword from Prime Minister Tony Blair, laid out a plan to achieve the national targets for 2010 of reducing inequalities in infant mortality and raising life expectancy faster for the most disadvantaged part of the population. Five years later, in Health Inequalities: Progress and Next Steps, the Department of Health refined its policy response in light of slower-than-expected progress toward meeting the 2010 targets.
In 2009 the government commissioned Sir Michael Marmot to conduct an evaluation of the previous ten years of policy efforts, summarized in Tackling Health Inequalities: 10 Years On. With the exhausting of the national targets in 2010, the government commissioned another independent review on health inequalities, again tapping Sir Michael Marmot to lead the commission that resulted in the Marmot Review. This review, like the Black Report, was commissioned by a Labour government but delivered to a Conservative one. The incoming Conservative government published its own public health white paper in 2010, but it did not contain a sustained emphasis on health inequalities, and hence was not included in our sample.
All of the British reports focus mainly on group differences in health outcomes, mentioning health care only as one cause among many of inequalities in health status. The latter differences are uniformly termed “health inequalities” (not disparities, differences, or “variations,” the term used by Thatcher's Conservative government after the release of the Black Report). While some of the reports reference health inequalities across geographic units (such as those between regions or health authorities within England), racial and ethnic groups, gender or disability status, by far the dominant group framing was class. For example, while the Marmot Review mentions “systematic differences” in health across gender and ethnic lines, the first sentence in the report, in the “Note from the Chair,” states the main point: “People with higher socioeconomic position in society have a greater array of life chances and more opportunities to lead a flourishing life. They also have better health” (p. 3). Later in the Introduction, ethnic and gender inequalities are characterized as “additional sources of disadvantage and exclusion” that go above and beyond SES (p. 39). Furthermore, British documents rarely discuss the non SES-related drivers of variations in health status across regions or smaller geographic units. Small area variations, in particular, are consistently presented as a proxy for the effects of socioeconomic deprivation. Of all of the documents reviewed here, the 2002 “Cross-Cutting Review” alone stands out for having a sustained focus on geographic, ethnic, and gender health inequalities as well as inequalities defined by SES. Overall, attention to inequalities across racial and ethnic groups (6 figures), across geographic areas (17 figures, 4 maps), and between men and women (23 figures) each garnered only a fraction of the attention devoted to class-based inequalities (60 figures).
The cohesive focus on SES inequalities in British documents is associated with a similarly cohesive set of causal stories to explain the origins of these inequalities. The Black and Acheson reports both use a political economy framework to explain health inequalities, positing a through-line linking economic and social structure, the experiences of occupants of different classes (Black) or socioeconomic positions (Acheson), and health outcomes. Worse health among working-class people was not a result of unhealthy working conditions or insufficient resources alone, however; the subjective experience of low socioeconomic position had its own independent effects on health behaviors and health outcomes—a position reiterated in the “Cross-Cutting Review.” Subsequent Labour government reports engaged in less pointed class analyses, but nevertheless assigned primary causal responsibility for health inequalities to underlying inequalities in “social conditions” or “social determinants.” By 2010, however, unabated health inequalities justified a return to more politicized language. The authors of the 2010 Marmot Report drew freely on the language of the 2008 WHO report of the Commission on Social Determinants of Health (CSDH), which was also led by Marmot, to declare that health inequalities were a result of “inequities in power, money, and resources” (CSDH 2008: 16, 37), and even cited the Phelan, Link et al. (2004) theory of fundamental causation.
But, while some British health inequalities documents contained an implicit critique of market capitalism at the level of causal explanation, at the level of policy recommendations they were more muted. Despite references to the social gradient in health, some of the furthest “upstream” policy recommendations were aimed at reducing poverty and deprivation, particularly among children, rather than dampening income inequality more generally. Labour government documents after Acheson mentioned reducing income inequality as a solution to the problem of health inequalities, but in practice this meant only action on the very bottom of the income distribution (implementing a minimum wage, upgrading minimum income benefits for families with young children and the elderly, and adjusting taxes and benefits to incentivize work). These interventions did reduce the incidence of poverty, and together with investments in early childhood education and housing surely made a real difference for many families at risk of having poor health outcomes. But the policy recommendations were nevertheless at odds with the more trenchant underlying critique contained within the causal theories espoused by the authors of these documents.
The first major report to be produced by the US government on health inequalities was Report of the Secretary's Task Force on Black and Minority Health, spearheaded by then-Secretary of Health and Human Services Margaret Heckler in 1985 (also known as the “Heckler Report”). The report is credited for first drawing attention to the issue of racial disparities in health (Docteur and Berenson 2014), which became a core theme of the agency's health strategy as outlined five years later in Healthy People 2000 (HHS 1990). The strategy received a significant update in 2000 with the publication of Healthy People 2010 (HHS 2000). Around this time, a key group of government consultants at the Institute of Medicine began devoting resources and attention to the study of health inequalities. Commissioned by congressional leaders in 2003, Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care was the IOM's first major statement on the issue of health disparities. The text self-consciously focused on differences in access to health care rather than health outcomes. It was followed in the same year by a report that placed greater emphasis on population health: The Future of the Public's Health in the 21st Century (2003). This report was a joint project of multiple health-oriented government agencies: the Centers for Disease Control and Prevention, the National Institutes of Health, the Health Resources and Services Administration, the Substance Abuse and Mental Health Services Administration, the Department of Health and Human Services (HHS) office for Planning and Evaluation, and the HHS Office of Disease Prevention and Health Promotion. The HHS Action Plan to Reduce Racial and Ethnic Disparities (2011) is the most recent update to the agency's policy strategy; and the CDC's January 14, 2011, supplement to its Morbidity and Mortality Weekly Report entitled Health Disparities & Inequalities Report justifies the Center's ongoing attention to the issue.
The American framing of health inequalities in policy documents differs sharply from the British one. American authors consistently write of “health disparities.” Unequal Treatment and the 2011 HHS Action Plan include differences in health care as an outcome of interest, but label these differences, like inequalities in health status, as “disparities.” Unequal Treatment defines the term as “racial or ethnic differences in the quality of health care that are not due to access-related factors or clinical needs, preferences, and appropriateness of intervention” (IOM 2003b: 20–21). Here, the link between disparities and racial and ethnic inequalities is patent. The link is less direct elsewhere, for example, in the CDC supplement, which defines health disparities as “differences in health outcomes and their determinants between segments of the population, as defined by social, demographic, environmental, and geographic attributes” (CDC 2011: 7). This more contemporary report also reviews the other language used to describe differences in health status, before ultimately settling on “disparity” as the operative language.
Despite the CDC's conceptual efforts to decouple the language of “disparities” from racial differences, the fact remains that health inequalities over racial and ethnic groups are the central frame in each of the US reports. Of the seven American policy documents we included, three reports are devoted wholly to inequalities across racial and ethnic groups (HHS 1985; HHS 2011; and IOM 2003b), and four reports make these inequalities a primary area of study (CDC 2011; HHS 1990; HHS 2000; and IOM 2003a). Compared to the British reports, the attention to race and ethnicity is staggering. However, it would be wrong to say that there is no attention to differences across other social groupings. Although the vast majority of tables, charts, and graphs in the American reports that illustrated a group-based inequality focused on race (82 figures total), attention is also paid to differences between genders (34 figures total), as well as across SES groups (23 figures total), and even across geographic areas (14 figures, 2 maps total). Nevertheless, major American government reports on health “disparities” always address race and ethnic inequalities, and directly or indirectly this term has acquired a racial charge.
Causal attributions for disparities in US documents are typically linked to social determinants. Whether the reports use this language and WHO's framework to explain these determinants varies. Healthy People 2010, the Future of the Public's Health in the 21st Century (2003), and the CDC supplement (CDC 2011) use the language of “determinants.” Yet, both the oldest (HHS 1985 [also referred to as the “Heckler Report”]) and most recent (HHS 2011) reports speak of “societal factors” and “social conditions” instead (pp. 16 and 4, respectively). This suggests that the introduction of WHO discourse in the intervening period has made limited inroads in American reports on health inequalities.
Moreover, the reports frequently follow these discussions of “upstream” determinants with discussions of the “downstream” determinants: namely, access to medical care (e.g., CDC 2011; HHS 2000), and factors specific to individual biology and behavior (e.g., HHS 1985; HHS 1990). In the case of the 1985 Heckler Report, the “social characteristics” responsible for health inequalities are largely tied to the health sector: demographics, health education, health professionals, health care services, and financing (p. 13). Biological and behavioral attributions, for their part, often underscore salient US social cleavages, even in the most globally oriented documents. For instance, the Future of the Public's Health devotes an entire appendix to the various models of health determinants, including the Dahlgren-Whitehead model. Yet, the preface of this report points to the fact that “factors interact in complex ways with each other and with innate individual traits such as race, sex, and genetics” (p. 16). One document, the Unequal Treatment report, includes discrimination as an important source of racial disparities in health care: “Consistent with the charge, the study committee focused part of its analysis on the clinical encounter itself, and found evidence that stereotyping, biases, and uncertainty on the part of health care providers can all contribute to unequal treatment” (p. 18).
Given the relative weakness of the social determinants causal story and the strength of health care system-related explanations in US policy documents, it should be no surprise that policy recommendations are almost always focused on health care policies (e.g., access to primary care, improved cancer screenings). Even Healthy People 2010, which explicitly adheres to a comprehensive “determinants of health” approach (p. 18), develops twenty-eight focus areas for policy intervention that are largely dependent on changes within the health care system. The report, like several other US reports (HHS 1985; HHS 1990; IOM 2003a), also looks to community-based and local actors to spearhead these efforts. Only the recent reports have alluded to more comprehensive policy solutions. The foreword by CDC Director Thomas R. Frieden to the Center's supplement makes a clear plea to improve “health and social programs, and more broadly, access to economic, educational, employment, and housing opportunities” (CDC 2011: 2). The second section of the third goal of the HHS Action Plan calls for the adoption of a “health in all policies” approach and the piloting of a “health disparity impact assessment” for selected programs (HHS 2011: 28), but these statements are buried deep in the text of the report. The dominant emphasis in the American reports is on behavioral and health care-focused interventions. As Secretary Louis W. Sullivan stated unequivocally in the foreword to Healthy People 2000: “health promotion and disease prevention comprise perhaps our best opportunity to reduce the ever-increasing portion of our resources that we spend to treat preventable illness and functional impairment” (HHS 1990: vi).
The first report on health inequalities in France in the modern era was Les inégalités devant la santé, authored by Sylvie LeRoux at the request of the newly appointed Communist health minister Jacques Ralite and released in 1985. The issue of health inequalities surfaced again in preparations for the eleventh national Plan (a working group prepared a brief on health inequalities included in the health system planning document Santé 2010 [Soubie 1993]), but the next major report did not emerge until 1998. The High Commission for Public Health (HCSP) issued Allocation régionale des ressources et réduction des inégalités dé santé in response to changes in the health care financing system that were prompted by Santé 2010. It is noteworthy as the first report from the country's main public health body that uses the term health inequality in the title. The next HCSP report on health inequalities, Les inégalités sociales de santé: Sortir de la fatalité, came more than a decade later (2009), and is explicitly addressed to “social” (i.e., SES) rather than territorial inequalities in health. The most recent major French report on health inequalities that we consider was prepared in 2011 by the Inspectorate General for Social Security (IGAS), and is once again nominally directed at social (SES) inequalities in health. The authors of the earlier documents frame inequalities in both health and in health care as problems. Geographic differences in health care resources and spending are presented as injustices in their own right, particularly (but not only) when resource allocation failed to keep up with the differing health care needs of residents. Later documents focus more on health outcomes, but cost-related barriers to accessing health care were still mentioned as prominent causes of health inequalities.
France has a long tradition of social epidemiology, which one might expect would make SES the dominant frame for discussing health inequalities among French policy elites, as it is in the United Kingdom. Furthermore, since 1965, data on mortality by both socio-professional category and place of residence, collected by the French national statistics agency (INSEE), have been available to researchers. However, discussion of health inequalities in France since the mid-1980s has had a strong territorial emphasis, with the SES-centered language so central in the UK appearing less frequently. While French reports on health inequalities discuss both differences in health status across SES groups and parts of the country as “inequalities,” the territorial frame is far more prominent than in either the UK or the United States.
Consider some examples from the earlier French reports. The LeRoux report details the health consequences of hard, lightly regulated labor in France's countryside, factories, fisheries, and office buildings, but the report is nevertheless deeply territorialized in its structure, such that each chapter devotes some or all of its time to describing geographic inequalities. The 1998 HCSP report begins by noting that there are large differences in both health and health care supply between the regions, with the north of France generally disfavored on both counts. The report concludes that the best way to reduce inequalities in health status between regions would not be through health care spending, but by “une politique régionale” that devotes supplemental resources to disfavored regions (LeRoux 1984: 23).
Later reports continue this trend, while nonetheless attempting to integrate them into the emerging international consensus (illustrated, for example, in WHO's CSDH report) that framed health inequalities as primarily occurring over socioeconomic groups. The 2009 HCSP report on health inequalities characterizes territorial inequalities as linked to social inequalities in health, arguing that “the geographic environment constitutes one of the determinants of health.” However, the territorial analysis recommended and carried out in part in the report is at a finer-grained level than much previous government analysis in France—at the level of the community or neighborhood rather than whole regions. And unlike earlier reports, this one casts territorial inequalities primarily as containers for SES inequalities (SES inequalities are “anchored” in “les territoires” [pp. 76, 92]). And IGAS's Les inégalités sociales de santé: Déterminants sociaux et modéles d'action similarly reflects current WHO language on health inequalities—but the report goes on to argue that “public policies need to take into account the relationships that link social inequalities in health with other forms of inequality, above all territorial inequalities” (p. 22).
As in the United States, the French reports for the most part limit their concrete policy recommendations to actions in the health care system. The exception is the early LeRoux report, whose political context—the report was requested by a Communist health minister—perhaps encouraged policy recommendations focused on workplace-related interventions, including worker participation in oversight of health conditions. The 1999 HCSP report and IGAS reports, both of which drew extensively on UK and WHO expertise, call for increasing the knowledge base around health inequalities, generating political will to tackle the problem, and cooperation across government departments and levels to resolve health inequalities. But these documents do not recommend specific policy interventions outside of the health care sector.
Given the resources governments devote to preparing major reports, they are the most likely of any form of policy discourse to be fully informed by the scientific literature; but, as Freeman (2006) points out, they are also selective: the process of producing government reports on health inequalities helps to build constituencies for particular ideas and policies within the policy elite, as contributors negotiate over common language. This selectivity is a source of bias, but we see this bias as highly informative. Major government reports are unlikely to use problem definitions or highlight results that are supported by only a minority of those scholars or activists who work in the area of health equity and are deemed trustworthy by the government agency producing the report. Such reports are also unlikely to promote policies or understandings that are at odds with the perceived direction of political winds. These documents are, in other words, political. No single type of text offers a complete view on how all actors in a policy field conceptualize the object of their work. Analysis of scholarly research publications, news media coverage, or the texts of legislative debates, for example, could provide information about how health inequalities are framed in public debates. We chose to analyze major national government reports, rather than the scholarly literature, public sentiment, or internal government deliberations from which they often draw inspiration, because national government reports make authoritative statements of the policy direction that a government wishes to pursue.
Another potential source of bias in our data springs from the time frame we consider. While the Black Report is generally considered an unproblematic starting point for analyses of contemporary government efforts to reduce inequalities in health, the endpoint of 2012 could be more problematic. Already by 2012 there was some convergence across countries in the language used in government reports, as we have seen. Further convergence might well be visible if we extended the time frame to more recent reports. For example, the Affordable Care Act could push health equity discourse in the United States toward a greater consideration of health outcomes, now that some of the most pressing health care access issues have been addressed. The 2014 French government's Stratégie National de Santé is more consonant with the standard WHO-Europe framing of health equity—for example, recommending an interministerial body to coordinate action on health inequalities—and less focused on geographic differences in health care access than in years past. Nevertheless, the gap between language and action persists, and discourse about health equity under the Conservative government in the UK has moved decisively back in the direction of emphasizing personal health behaviors and local-level initiatives.
Discussion and Conclusion
Our comparative results reveal important differences between national health inequality discourses, as well as the historical development of these differences. A first important difference concerns the use of language signaling group differences versus language that contains a more explicit moral judgment about the unacceptability of these differences. British and French reports use “inequality” as the default term to describe group differences in health and health care that are preventable and unjust. The most commonly used term in the American reports, “disparity,” is also linked to a sense of injustice, through its resonance with Title VI, which rendered disparate treatment of racial groups illegal. However, the language surrounding these core terms in reports from the three different countries differs markedly, with more frequent and more urgent appeals to a sense of (in)justice and equity in the British and French documents than in the American ones.
A second important difference between the reports from the three countries concerns the dominant group used to illustrate or frame the issue of health equity. Since the 1980s, policy documents in the United States have put much more emphasis on racial and ethnic differences in health status than on differences across any other group. Meanwhile, British reports in the past thirty years have included only a handful of charts, tables, or graphs reporting differences in health status across racial or ethnic groups, and French reports have shown none. One possible explanation for differences in attention to different group dimensions of health equity across countries is that government statistics are available only, or primarily, for certain types of groups. The fact that in the United States we have better data on mortality and morbidity by race than by SES, for example, likely reflects the fact that (socially constructed) race has always been a highly salient fact about American bodies, while class is often regarded as less important and/or impossible to measure (Krieger, Chen, and Ebel 1997). Meanwhile, while official British statistics have recorded occupation for at least a century, data on ethnicity is only available starting in 1991 (only country of origin for foreign-born Britons was available from the mid-twentieth century). It may come as no surprise, then, that while race and ethnicity are central axes of comparison in US health equity documents, British health inequalities documents privilege class differences in health status. In France, where there are no government statistics on race and ethnicity to draw upon, socioeconomic inequalities compete for attention with territorial inequalities in government reports.
Data availability certainly affects what types of analyses are likely to be conducted and hence reported—but governments can and do make changes to policies about what types of data to collect. And even when new types of data become available, government reports do not necessarily use these data. The National Institutes of Health (NIH) Revitalization Act of 1993 required that data be collected in order to allow analyses by race and gender (but not SES), which led to an explosion of research and evidence on racial disparities in the United States (Friedman and Lee 2013). However, while the British census introduced questions about ethnicity in 1991 (White 2012), some local authorities recognize that the “collection and use of this data remains inadequate” (Greater London Authority 2010, emphasis added). Indeed, government attention to ethnic health disparities in Britain appears to be limited to selected local authorities, particularly in areas with high immigrant populations, such as London (Greater London Authority 2010).
To explain the inattention of the national government to racial and ethnic differences in health in the UK, Nazroo (2003) points to the influence of Michael Marmot's conclusions in a 1984 study of immigrant morality rates: “Published shortly after the Black Report had firmly placed inequalities in health on the research agenda, this study used a combination of British census and death certificate data to explore the relationship between country of birth and mortality rates. A central finding was that there was no relationship between occupational class and mortality for immigrant groups, even though there was a clear relationship for those born in the United Kingdom. These findings led to the conclusion that differences in socioeconomic position could not explain the higher mortality rates found in some migrant groups in the United Kingdom.” The study (Marmot, Adelstein, and Bulusu 1984: 277), which predated the expansion of available data on ethnicity in the United Kingdom, effectively removed racial and ethnic disparities from the research and policy agenda in the longer term. Curiously, one could imagine the same study producing the opposite conclusion in the United States, where policy makers emphasize how the differences in racial and ethnic health are not explained by class.
Just as data on ethnic differences are available, but largely unacknowledged, in the UK, other countries have enough empirical evidence to shift their dominant frames but choose not to do so. For example, in the absence of official government statistics on race and ethnicity in France, other forms of analysis could provide significant insight into racial and ethnic disparities. Some researchers have used census data on country of origin for non-native born French as a proxy for ethnicity (Berchet and Jusot 2012). Analysis of small-area variations in health, which is prominent in more recent French reports, could also serve a similar purpose, given the spatial concentration of ethnic minorities in France. The territorial frame has deep roots in French politics where, historically, concerns over the state's capacity to rule and service all parts of the territory evenly has been closely linked to concerns about equal citizenship (Lynch 2016), so the explosion of small-area analyses in recent reports is not surprising. What is, perhaps, more surprising is that French government reports do not explicitly link small-area variations in health to the racial/ethnic makeup of these areas, instead leaving it to the audience to draw their own conclusions about whether racial and ethnic inequalities are problematic or, indeed, even exist.
Why, then, do the frames in US, British, and French policy documents about health inequalities differ? And what do these differences tell us about the nature of health politics in these three countries? In this article we can do little more than speculate. But different conceptualizations of what “counts” as a health inequality worth reporting on in the three countries are, we suspect, related to the particular historical salience of different cleavages in society. In all three societies, health inequalities are framed in terms of the broader inequalities that are already most familiar in politics. The survival of the landed nobility and aristocracy in Britain has made the British “a people uniquely obsessed by ‘class’” (Lawrence 2000: 307) and, by extension, class differences. The legitimacy of class conflict as a mode of politics is inscribed in the British party system, where the main cleavage runs between the Labour and Conservative parties. In the United States, race is, of course, the most salient difference, codified into law and statute in a way that class has never been. Finally, the key register in which inequality is expressed in France is not class, but territory. Territorial unity was a central issue in the formation of the French state and French citizens (Braudel 1992; Weber 1976), the lack of access in France's “deserts” to the amenities and privileges of the Parisian way of life has been a theme animating French policy since WWII, and the determination to remove territorial disparities became a cornerstone of postwar French policy, codified in successive national economic plans (Baudouï 1999; Lynch 2016).
Because class in the UK, race in the United States, and territory in France are the most familiar and most institutionalized cleavages, they are, we suspect, the most readily accessible to researchers and policy makers, and the most likely to be incorporated into inequality documents. It does not, however, follow that these familiar cleavages are the most important in terms of the size or kind of inequalities they generate. For example, France and the United States both have significant SES-related health inequalities that receive less attention than do racial (in the United States) or territorial (in France) inequalities in health. And in both Britain and France, there is evidence of racial and ethnic inequalities in health (e.g., Berchet and Jusot 2012; Marmot, Adelstein, and Bulusu 1984) that receive little attention from official national reports.
The inequalities that policy documents avoid discussing too thoroughly may point us to areas of particular sensitivity. As painful as it may be to “discover” racial discrimination in health care in the United States, regionally inequitable distribution of health care resources in France, or class inequalities in health status in the UK, these inequalities are in some sense less threatening than would be class inequalities in the United States, territorial inequalities in the UK, or racial inequalities in France, precisely because they are more familiar and better institutionalized. Government reports may or may not offer meaningful solutions to the familiar inequalities, but recognizing that these familiar cleavages exist is not inherently threatening, and may in fact serve as valuable signals to particular support bases.
Whatever the source of differences in the group definition of health inequalities across countries, they coincide in these documents with a third important difference: the causal stories used to explain why health inequalities exist. In recent years, WHO has downplayed the role of health behaviors and health care in producing health inequalities, in favor of a causal story highlighting the role of social determinants and underlying inequities in power and resources. Policy documents in all three countries have adopted this “political economy” frame at least in part, but to different degrees. In the UK documents, the class structure of society—and even in some cases the market economy itself—plays an important causal role even upstream of the standard social determinants of health model. In the United States and France, on the other hand, the health care system remains a central actor in the reports' causal stories, reflecting incomplete health insurance coverage and significant geographic variations in access to care in both countries for much of the period under study. American documents additionally point to the role of personal responsibility (e.g., HHS 1990), and even discrimination (e.g., IOM 2003b), in producing some inequalities, while explanations in the British and French documents more frequently refer to larger-scale structures (e.g., income inequality) and processes (e.g., l'aménagement du territoire). American government reports have not, by and large, adopted the framing of health inequalities as socioeconomic inequalities in health status caused by an inequitable distribution of resources and power in society.
Not surprisingly, in light of these different understandings of the causes of health inequalities, documents in the three countries also differ in a fourth important way: in their policy recommendations. While reports in all three countries refer to whole-of-government, health in all policies, or cross-sectoral policy-making approaches, the extent to which the authors of the reports in different countries make recommendations outside of the health care system is striking. British policy documents contain multiple, concrete recommendations for interventions to reduce poverty, particularly among children, and in areas such as the tax and benefits system, transportation, housing, and the like. French and American documents instead make concrete policy recommendations mainly in the arena of health care, for example, increasing access to preventive care, expanding take-up of cancer screening programs and health education, and reducing cost-related barriers to accessing health care. Some elements—for example, health in all policies, health impact assessment—of the standard European suite of policy recommendations for reducing health inequalities have appeared in recent American work (e.g., HHS 2011). However, WHO plays a marginal role in American health policy, and there is no supranational equivalent to the WHO Regional Office for Europe or the European Union pressuring the United States to bring its policies into alignment with the reigning paradigm in health equity.
Despite these important differences between the health equity problem's framing in the UK, the United States, and France, there is the single remarkable similarity that unites these policy documents. That is, not one of these reports—not even those explicitly recognizing the role of income inequality in shaping health inequalities—made a policy recommendation that would entail significant redistribution of economic resources or power. While poverty reduction appeared as a major theme of British documents from the Black Report onward, reshaping the income distribution in a way that would significantly flatten the entire socioeconomic gradient decisively did not.
At one level this is not surprising: Why, after all, would a government report seek to undermine the political economic system on which it rests? Even WHO's CSDH report, while it argued in general terms for redistribution, was nevertheless largely “silent on the topic of whose resources, and how and through what instruments” (Navarro 2009: 440; see also Birn 2009; Escudero 2009). Redistributing income or wealth downward, let alone altering the systems of production and accumulation that give rise to income inequalities, is a politically difficult demand in the societies of Western Europe or North America where neoliberal ideas and practices are dominant (Lynch 2017). On the other hand, it is worth bearing in mind that the British reports discussed in this article were all products of a Labour administration that explicitly recognized the role of economic inequality in producing health inequalities, and was strongly and publicly committed to reducing these inequalities. More generally, one might question why a government would go to the trouble of convening experts, gathering evidence, and releasing a landmark report on health equity if its leaders did not sincerely desire to address the problem. Viewed in this light, the mismatch between the theorized causes of health inequity and the proposed solutions in these reports is indeed surprising.
What further factors could account for the lack of congruence between the causal frame and the policy recommendations in all three countries? Bergeron, Castel, and Saguy (2013) point out that much of the framing literature has tended to assume that policy outcomes will be consistent with dominant frames. Discursive institutionalists, on the other hand, have aimed to “show empirically how, when, where, and why ideas and discourse matter for institutional change, and when they do not” (Schmidt 2010: 21). To Bergeron, Castel, and Saguy (2013), this implies that, even once adopted as dominant, a frame may not result in the choice of policy instruments that is coherent with the dominant frame. One possibility, then, is that the political economy causal frame meets none of the preconditions suggested by Bergeron, Castel, and Saguy (2013) for a close match between frames and policies. First, the main articulators of the political economy frame, epidemiologists and public health scholars, are well-enough integrated into health policy making to be invited to contribute to the report; but their expertise is seen as marginal to the process of politics. Hence, their policy recommendations may be ignored or downplayed during the writing up of reports in an effort to make the recommendations more politically palatable. Second, even health policy-making structures are organized around an ecology dominated by medical, rather than public health, actors (see, e.g., Smith 2013 on the English case). This serves to limit the reach of public health experts' recommendations even within the broader health field. Finally, the political economy frame itself implies a need for major institutional and political restructuring—in this case, significant redistribution of power and resources. These three factors, taken together, could well explain the mismatch between the causal understandings underlying the health equity frame and the policies that are ultimately recommended and enacted in its name. So, perhaps the silence of British, French, and American health equity documents on the topic of redistribution should not surprise us. Nevertheless, some redistribution may well be necessary in order to significantly improve health equity.
In this article, we have sought to shed light, through comparative analysis, on features of American public policy discourse surrounding health equity that might otherwise go unnoticed. Our government's most emphatic statements surrounding health equity have cast the issue mainly (though never exclusively) as a problem of racial disparities, and, while they have recognized the role of social determinants in shaping health, have nevertheless tended to focus on solutions based in the health care system. If we are to make strides toward greater equity in health, and not only in health care, we must go beyond the policy remedies currently envisioned in government reports. Not only must we act on the proximate social determinants of health like housing, transportation, or food availability. We must also demand that our political leaders put on the agenda the more politically risky remedies at which even European governments have balked, remedies that will fundamentally redistribute the underlying inequalities in power and resources across racial, ethnic, gender, geographic, and socioeconomic groups.
The authors thank Alyssa Kennedy and Alexandra Babinchak for their tireless research assistance, participants in the 2016 RWJF Investigator Awards in Health Policy meeting for their extraordinarily helpful feedback, and the coordinators of this special issue for their incisive editorial advice. We are particularly indebted to Jeb Barnes for his comments. We would also like to thank the numerous academic colleagues and health policy professionals in England and France who made time to speak with us and help us understand the health equity issues facing their societies.
Qualitative Coding Scheme for Government Health Equity Reports
A close reading of the government reports was carried out by the authors to identify the following themes. Particular attention was paid to the executive summaries, prefaces, and introductions to the reports when assigning weight to various frames. Additionally, each figure (table, chart, graph, or map) in the report that depicted an empirical relationship drawing on data was coded according to the outcome(s) and group frame(s) depicted.
Access to health care
Quality of health care
SES (including occupational status, education, income, wealth, employment status)
Geography (comparisons within the country of the report)
Other (e.g., cross-national comparisons)
Equity, inequity, justice, injustice, fairness, unfair, fair
Health care system
Access to curative care
Access to preventive care
Health behaviors, individual choices, risk factors with mention of structures shaping these choices
“Social Determinants of Health”
Other mention of social determinants
Deprivation (poverty, exclusion)
Discrimination, individual racism
Causes of causes
Underlying causes, “causes of causes”
Structural, institutional racism
Political economy (class structure, capitalism, inequality)
Health care system solutions
Upgrading quality, appropriateness
Greater emphasis on primary care and prevention
Changing provider behavior (e.g., end discrimination, more attention to determinants)
Expand use of screening
Changes to medical social services (e.g., home care for elderly, disabled)
Other health care
Behavioral change solutions
Health education and outreach (in schools, workplaces, communities)
Smoking cessation programs
Act on structures affecting individual health choices
Restrict marketing of food, alcohol, tobacco
Tax food, alcohol, tobacco
Restrict sales of food, alcohol, tobacco
Encourage availability of health foods
More opportunities, facilities for exercise
Act on social determinants
Education (not health education)
Income (raise minimum income benefits)
Limit unemployment spells through training, activation
Pollution (air, noise, toxic substances)
Dangerous working conditions
Act on economy
Limit unemployment spells by incentivizing hiring, reforming labor market
Redistribute income (not just by increasing minimum income benefits)
Limit free trade
Act on politics
Reduce structural or individual racism
Other (e.g., collect more data)
1. The Gusmano et al. 2010 comparative study of New York, Paris, and London is a useful guide to health policies that affect population health at the metropolitan level in the three countries we consider, but in that volume there is no attention to the alternative framings of health equity.