Investigating inconsistencies regarding health equity in select World Health Organization texts: a critical discourse analysis of health promotion, social determinants of health, and urban health texts, 2008–2016

Abstract

Background

Scholarly critiques have demonstrated that the World Health Organization (WHO) approaches the concept of health equity inconsistently. For example, inconsistencies center around measuring health inequity across individuals versus groups; in approaches and goals sought in striving for health equity; and whether considerations around health equity prioritize socioeconomic status or also consider other social determinants of health. However, the significance of these contrasting approaches has yet to be assessed empirically.

Methods

This study employs critical discourse analysis to assess the WHO’s approaches to health equity in select health promotion, social determinants of health, and urban health texts from 2008 to 2016.

Results

We find that the WHO: (i) usually measures health equity by comparing groups; (ii) explicitly specifies three approaches to health equity (although we identified additional implicit approaches in our analysis of WHO discourses); and (iii) considers health equity inconsistently both in terms of socioeconomic status and other social determinants of health, but socioeconomic status was given substantially more attention than other individual social determinants of health.

Conclusions

There is misalignment with the WHO’s stated approaches to tackle health inequity and its discourses around health equity. This incongruence increases the likelihood of pursuing short-term solutions and not sustainably addressing the root causes of health inequity. Critical discourse analysis’ focus on power allows for understanding why ‘radical’ approaches are not explicitly expressed to ensure that governments will be agreeable to addressing health inequity.

Background

Health inequity, “differences which are unnecessary and avoidable, but in addition, are considered unfair and unjust” [1], is a major focus—the central focus according to some—in public and global health [2]. Despite this emphasis on health equity, a recent scoping review found that the World Health Organization (WHO) uses ambiguous and contradictory approaches to health equity [3]. Those authors identified three inconsistencies: (i) individual versus group measurement of health inequalities; (ii) whether the goal is a minimum baseline for everyone or limiting the difference between the best and worst off; and (iii) whether the focus is on reducing inequities between those with different socioeconomic status (SES) or broader social determinants of health (SDH) [3]. The stances of the WHO are as observed by scholars, which are positioned as inconsistencies in the scoping review. However, none of these observations are empirically tested. The goal of this study is to empirically investigate how the WHO’s work aligns or not with these three observed inconsistencies.

To empirically investigate, our study assesses a purposeful selection of WHO texts in the domains of health promotion, SDH, and urban health to determine what WHO texts reveal about the organization’s underlying concepts and approaches across these inconsistencies. Critical discourse analysis (CDA) is an interdisciplinary qualitative method for analyzing power relations within language use, which are understood to manifest in social practice. CDA is an appropriate method for uncovering discourses in relation to their social context, and its objective is to bring about positive change. Discourses can be understood as “ways of writing and talking about a phenomenon” [4]. We sought to make explicit understandings and assumptions that may be only implicit. By making understandings and assumptions explicit, it can subject them to scrutiny. This is important particularly when the concepts themselves are highly significant but subject to multiple interpretations, as is the case for the concept of health equity. Undertaking this work is important to determine if in fact the WHO does approach health equity in an inconsistent manner, as relayed through WHO texts in our domains of interest, and as relevant to the three inconsistencies outlined in the aforementioned review. Analyzing whether these inconsistencies are manifested through discourses will allow for better understanding potential shortfalls in addressing health inequity and subsequently, identify opportunities to clarify ideas around health equity and associated recommended actions. Shortfalls can vary in consequence(s), potentially including neglecting important populations or actually undermining health equity. Our inquiry is intended to improve clarity around important questions for operationalizing health equity—a central concept in public and global health.

Methods

Critical discourse analysis

Taylor argues there is increasing importance afforded to language and that social change is driven by discourses [5]. By analyzing discourses, we seek to better understand who is using the language, why, how, and when [6]. CDA is an “approach to conceptualize and study discourse as social practice” [4]—in other words, by asking what ideas or assumptions are relayed and how does this translate into action?—by bridging the “micro” (discursive practice) with the “macro” (state, government, and policy) [7]. CDA is useful in documenting marginalized, hybrid, multiple, and competing discourses and shifts in discourses, and those that translate into policy [5].

CDA is also appropriate when applied to analyze organizations’ discourses, such as the WHO, as organizations hold social power and shape public discourses [8]. As such, we understand power to encompass more than traditional laws, regulations, and subjugation, but also knowledge produced through institutions. In other words, the power and knowledge that shapes everyday lives [9], in line with governmentality [10]. In other words, power can manifest through knowledge and discourses that shape our understandings [10, 11]. Through providing this critical “perspective” to analysis in line with the critical tradition that embraces value-free science [8], CDA can elucidate ideological representations in background knowledge [7] and, accordingly, facilitate policy analysis [5]. One of the strengths of CDA is the understanding that social life and power are shaped by language [4]. Other methods, such as content and thematic analyses, face shortcomings of not aligning with the critical tradition to investigate discourses and accordingly informed our choice to undertake CDA to assess select texts for the three inconsistencies outlined above.

Focus on the WHO

Because the knowledge produced by the WHO shapes broader understandings, we focused our attention on the work of the WHO. The WHO was created in 1948 as the United Nations specialized agency for health. In its constitution, signed by representatives of 61 states in 1946, the objective of “the attainment by all peoples of the highest possible level of health,” principles, and functions of the WHO are outlined [12]. The functions outlined in the constitution, precisely 22, describe the role and duties of the WHO and include acting as the “directing and co-ordinating authority on international health work;” maintaining collaborations; assisting governments upon request; providing administrative and technical assistance; promoting conventions, agreements, and regulations; promoting and conducting research; developing an informed public opinion; and establishing standards, such as around food products [12]. However, it has been argued that the WHO plays two main functions. The first being core work, which includes normative tasks, and supplementary, which includes technical cooperation [13].

The WHO operates as a key political institution aligned with global systems of power and capital. Politically, the WHO provides an authoritative voice in global health. The WHO guides policy and practice, such as through producing influential reports and providing technical support to member states. The WHO also shapes discussions around health across the world, which is particularly apparent during the COVID-19 pandemic where the WHO bridged the scientific community with lay people. Evidently, the WHO and its work guides broader global and public health practices and is worthy of inquiry.

Data sources

We selected the three domains of inquiry—health promotion, the SDH, and urban health—because they all provide different perspectives on health equity. Health promotion takes a high-level approach to health equity [14]. The SDH reflect an upstream perspective. Urban health was included because for many health conditions, cities are where inequities are the greatest [15]. We identified WHO activities to help operationalize our search for texts in these three areas. We focused on texts from two WHO global conferences on health promotion in 2011 and 2016. We focused on the Commission on Social Determinants of Health’s (CSDH) work, as the WHO’s premier investigation in this area [16]. For urban health, we chose to focus on global reports on urban health, such as the WHO’s Urban Health Equity Assessment and Response Tool (Urban HEART) [17], which was launched following the CSDH’s final report. We selected prominent and influential texts within these three domains using a multi-step process. We began with search results obtained during a previous scoping review [3, 18]. That scoping review identified 2558 search results investigating the WHO’s concept of equity in health [3]. For this paper, we wanted to identify the most influential texts as reflected in these search results. We reviewed these results keeping a list of WHO texts that were cited, discussed, or mentioned. This yielded about 100 different WHO texts. In tracking these results, we focused our attention on documents appearing multiple times and focused on health equity. We based our assessment on several criteria. We included texts published between 2008 and 2021, inclusive. We began with 2008 because the final report of the CSDH was published that year, drawing attention to the SDH and laying out an agenda for achieving health equity [16]. We considered this to be a WHO publication because it was commissioned by the WHO and issued under its authority, even though the report was overseen by external commissioners. We assessed the influence of WHO texts based on how many times texts were mentioned in the search results. For example, the Rio Political Declaration on Social Determinants of Health [19] was mentioned in nine hits retrieved in the scoping review search [20,21,22,23,24,25,26,27,28]. We also recognize the limitations to a quantitative assessment of citations. In policy and practice, texts may be prominent but not mentioned explicitly. To counteract this, we applied our knowledge of the field to identify prominent and influential texts in these domains (e.g., major global report, political declaration). We chose multiple types of texts because what the WHO says and does is expressed in a variety of ways. We considered different communications channels, including technical reports, press releases, speeches, and commissioned reports. Given that we wanted to analyze at the institutional level, as opposed to another unit of analysis, we looked for these various outlets. We also sought interrelated texts, meaning texts that are linked to each other. This decision aligns both with our intentional choice to uncover discourses across texts and methodologically with CDA to assess interrelated systems of knowledge. Please see Fig. 1 for an illustration of these interrelations, where a solid line depicts a text that cites another, and a dashed line depicts a text being mentioned but not explicitly cited [29]. We wanted to undertake a global analysis, so we did not apply any geographic restrictions. Applying these aforementioned criteria resulted in the identification of nine texts for inclusion in this study and another [29], detailed in Table 1. These texts were discussed by the authorship team and agreed upon prior to analysis. All selected texts are publicly available online, so institutional review board research ethics approval was not required.

Fig. 1

Depiction of cross-referencing between selected texts

The source of this figure is [29]

Table 1 List of WHO texts we analyzed

Approach to analysis

We sought to uncover discourses through analysis at different levels: (i) language use in text (speech, writing, images such as graphs, or a mixture of these), such as through analyzing the selection of words; (ii) discursive practice or the communication of beliefs, such as understanding what interpretation or attitudes are conveyed through the selection of such words; and (iii) social practice and structures, such as through understanding how power is reinforced through presenting such beliefs. Select WHO texts (Table 1) were each read carefully and coded a priori by the first author into the three inconsistencies identified through the scoping review: individual versus group measurement; approaches and goals in striving for health equity; and, when striving toward health equity, if the focus was largely on SES/SEP or considered other SDH [3]. In addition to analyzing stated text appropriately (e.g., focus on SES/SEP or other SDH) based on the aforementioned scoping review, attention was paid to the “unsaid,” as “ideologies are primarily located in the ‘unsaid’ (implicit propositions)” [7], in line with CDA. In other words, not restricting the analysis to the a priori codes established in the scoping review. This is most evident when considering approaches taken in tackling health inequities or goals sought, where analysis also sought to illuminate key lines of debate playing out in the texts, as demonstrated through presenting unstated approaches. Thus, the assumption was that discourses may extend beyond just simply aligning or not with what was presented in the scoping review. Texts were coded using NVivo 12 [36] software and overarching findings are presented in the “results” section within the respective inconsistency grouping.

Results

Overall, the results of this study clarify scholarly ideas around these inconsistencies and demonstrate that the select WHO texts in the three domains of interest largely approach health equity measurement by comparing groups; explicitly specify three goals/approaches to health equity, but additional unstated approaches emerge when assessing the WHO’s discourses, which are striving for a baseline level for all, improving average health, reallocating resources, and striving for a full health potential for all; and also considers health equity in terms of both SES and other SDH. These three sections are discussed at length below.

Measurement of inequity

In seeking to determine if WHO texts discuss measuring health inequity through only focusing on individuals or only focusing on groups, findings demonstrate that WHO texts undertake data analysis to draw comparisons largely between groups to uncover health inequities and encourage lower levels of data aggregation to ensure health inequities are uncovered. These discourses are discussed in the proceeding sections.

Individual versus group measurement surveillance

We found that the WHO texts in our sample measure health inequity across groups, for instance stating “community-based participatory surveillance of urban health determinants should be a component of health and social outcome surveillance initiatives, including the monitoring of intra-urban differentials, to produce comparative analyses” [30]. Or similarly, in “looking at the health status of subgroups of city dwellers according to their socioeconomic status, neighborhood or other population characteristics” [33]. And measurement across groups is further reflected in the framework used in Equity, social determinants and public health programs [31] where the third level, differential vulnerability, relates to population groups, and the bottom two levels’ focus on individuals also includes a discussion of population groups; figures, such as graphing social gradients by asset quintiles and regions (Fig. 14.1 in [31]); tools, such as the Urban HEART which investigates inter- and intra-city health inequities [17]; recommendations, such as those by the CSDH for national governments to establish a “health equity surveillance system” that “would present data stratified by social groups within countries, and would include measures of inequity in health and determinants between these groups” [16]; and observations of others’ work, such as reflections on a government of Cambodia and Asian Development Bank project focused on providing primary healthcare, where it is noted that “it is unclear if the units of analyses were the geographical areas—as they should have been—or individual children and women” [31]. This finding also aligns with the WHO’s Health Inequality Data Repository that seeks to “…identify differences in health between different population subgroups” [37]. Evidently, the WHO texts analyzed aligned with measuring health inequities across groups rather than individuals, and in addition, this choice was met with calls for lower levels of aggregation.

Calls for lower-level aggregation

The assessed texts contained brief mentions of the inability of averages to distinguish health inequities, which aligns with the WHO’s choice to measure health inequities across groups. The rationale for exercising caution when using averages also applies in presenting the need to use lower levels of aggregation; in other words, to uncover more complex patterns, as one analyzed WHO report notes “differentials and variances tend to get lost in aggregation” [31]. In fact, urban averages and data, when disaggregated at the neighborhood or district level, provide different portrayals of health inequity [33]. Instead of “relying on city averages [which] has further obscured inequalities within cities” [33], “average results for specific groups of people or specific risk factors (e.g., gender groups, income groups, ethnic groups or neighborhoods) [can be] revealed” [32]. Thus, choosing to not disaggregate data at a lower level means that health inequities across certain populations and individuals will not be revealed. This subsequently entails that such inequities are unlikely to be addressed through deliberate action. This focus on disaggregated data aligns with the WHO’s Health Inequality Data Repository that “uses health data disaggregated by relevant inequality dimensions (i.e., demographic, socioeconomic or geographical factors)” [37] and the PROGRESS stratifiers investigated (i.e., place of residence, race or ethnicity, occupation, gender, religion, education, socioeconomic status, and social capital or resources) [38]. With various methods for determining health inequities, it is important to ensure the measurement of indicators and associated collection of data for health information systems around health inequity are “designed as a system for action: not simply for the purpose of gathering data, but also for the purpose of enabling decision-making for the interventions that target the identified social determinants” [31], as noted by the WHO. For this reason, data analysis and how health inequities are determined is important. In other words, how health inequities are acted on is partially predicated on the measurement, but also the type of health equity approach(es) and/or goal(s) that are being sought. This latter component is the focus of the next section. However, it is worth noting that discourses around lower-level aggregation are reinforced and can lead to measuring health inequities in this way, which is further presented below.

Health equity approaches and goals

We sought to determine what approaches and goals the WHO texts analyzed align with, if any, when aiming to reduce health inequity. Select WHO texts analyzed explicitly mention three “main approaches” which are as follows: (i) "targeting disadvantaged population groups or social classes" (hereby referred to as “targeting populations”), (ii) "narrowing the health gap" (hereby referred to as “narrowing the gap”), and (iii) "reducing inequities throughout the whole population" [17, 33] (hereby referred to as “levelling up”). But in addition to these approaches, analysis of the texts demonstrated that the WHO’s language supports four unstated goals or approaches, which are as follows: (iv) striving for a baseline level for all, (v) improving average health, (vi) reallocating resources, and (vii) striving for a full health potential for all (collated in Table 2 for ease of reference).

Table 2 Health equity approaches and goals

These seven approaches are presented in depth in the proceeding section. Defining these approaches and goals is important for understanding where and how WHO texts discuss targeting their efforts and to shed light on gaps in approaches. These potential gaps can include populations that are not targeted, actions that seek to improve health but not necessarily health equity, and others.

WHO texts clarify that these stated approaches are interdependent. However, we found that it was not always clear what precise approach(es) was/were being recommended. Notably, selection across approaches depends on numerous factors, such as disease targeted and patterns of health inequity, the latter of which largely relates to the measurement of health inequities as presented above.

• (i) "Targeting disadvantaged population groups or social classes"

Under this first approach, abbreviated to “targeting populations”, only the target group is assessed for improvements, such as the most disadvantaged or certain population groups [17]. Therefore, the remaining population is not assessed and improvements are not relative to other groups [17]. This approach of “targeting disadvantaged population groups or social classes” has been noted as being “mandatory” from an equity perspective [31] and essential to move toward achieving universal health coverage [34].

In addition to being seen as “mandatory” in Equity, social determinants and public health programs [31], this approach was recognized as being only a “partial response” by the CSDH [16]. This is because this approach does not always result in a reduction of health inequities [17], particularly when considering absolute measures as opposed to relative measures [33]. Despite this, there is a strong emphasis on targeting certain population groups, such as Indigenous Peoples, that is expressed as needing to be distinct from universal approaches [16], or social classes like the urban poor, which was the focus of the Hidden cities [33] and Global report on urban health [34] reports. For example, in discussing the role of the healthcare sector, the CSDH text states “it can promote health equity through specific attention to the circumstances and needs of socially disadvantaged and marginalized groups” [16]. And, in general, such “pro-poor” approaches—left undefined in this use—have been identified as needed and as an effective strategy for achieving equity in urban settings [30]. In fact, UN-Habitat’s analysis of countries with respect to Millennium Development Goal (MDG) 7, Target 11, found that successful countries had a political commitment to slum upgrading and pro-poor land and housing reforms [30]. Thus, this approach’s focus on disadvantaged groups and social classes has merits and was deemed useful in the short term [16]. However, it is important to observe that this approach was deemed both “a partial response” and “an effective strategy for achieving equity,” demonstrating an inconsistent view of the merits of this approach. This inconsistency seems to lie in the word “achieving” which suggests the accomplishment of a goal rather than addressing the issue or advancing toward a goal. However, this approach can perhaps be considered both a partial response and contributing to achieving equity.

• (ii) "Narrowing the health gap"

Under this second approach, abbreviated to “narrowing the gap”, the most disadvantaged are similarly targeted, but this is in reference to the best off [17]. Thus, the goal is reducing health inequity between the worst and best off, typically between the worst and best performing quintiles [17]. This differs from (i) “targeting populations” where there is no reference to a comparator group. As such, approaches that target segments of society that may be neglected, such as this approach and the one prior, are thought to bring about the largest population health benefits [35]. Similarly, while this approach was identified as being accepted by most for achieving health equity [33]—with who constitutes “most” being left unspecified—this notably contradicted the same sentiment shared about approach (iii), “levelling up”, in the same report [33]. And both this approach (ii) “narrowing the gap” and the prior approach (i) “targeting populations” do not necessarily “leave no one behind,” which was the call on the global community from the Shanghai Declaration [35], as groups between these quintiles are left unaddressed.

This approach and the language used to describe it arguably align with the emphasis on “closing the gap in a generation,” the title of the CSDH’s report [16]. This is illustrated through the comparison of groups (“measures are needed to support women to equally progress in work, to be on a par with men”) and Target 1 of the CSDH (“Reduce by 10 years, between 2000 and 2040, the LEB [life expectancy at birth] gap between the one third of countries with the highest and the one third of countries with the lowest LEB levels, by levelling up countries with lower LEB”) [16].

It is noteworthy that there are slight deviations in this approach. For example, in discussing reaching slum dwellers with respect to communicable disease control, one challenge is ensuring that “they are provided opportunities equal to the rest of the population to access proven interventions” [30]. Specifically, the goal is not ensuring the health of slum dwellers is equal to the level of the best off—as aligned with this goal of narrowing the health gap—but rather, potentially to the level of the societal average for opportunities to access interventions.

• (iii) "Reducing inequities throughout the whole population"

Lastly, under the third approach, abbreviated to “levelling up,” the whole population is targeted in seeking to improve all groups to the level of the best off [17]. It appears this approach, unlike the other two, attends to other quintiles aside from just the bottom quintile. This approach is thought to be the best for achieving health equity by “most” [33]—again, where “most” and the rationale are left unspecified. This approach has also been emphasized as, “achieving the various specific global health and development targets without at the same time ensuring equitable distribution across populations is of limited value” [31]. Thus, it is what cities should strive for to improve health equity [34]. However, there is a downfall of potentially not reaching vulnerable groups [33] and, on the opposite end of the gradient, those who are most powerful for the purposes of redistribution of resources. The latter of which was not indicated as a component of the three main approaches outlined in WHO texts, arguably inferring that the most powerful are not targeted in this way.

Notably, the role of the social gradient is highlighted when discussing this approach [17]. Emphasis on the social gradient is apparent through discourses around “how fairly health is distributed across the social spectrum” [16] and “in urban areas of several exemplar countries, under-five mortality rates decline progressively as family income rises. These results indicate that efforts to reduce inequities need to address the entire population, rather than focusing only on the poorest groups” [33]. This latter statement demonstrates WHO texts’ expressed position of the need for interdependency across these three approaches (further explained below).

Although the CSDH’s “call for the health gap to be closed in a generation” [16] can also align with this approach, it largely depends on which gap you are addressing. Whether this targeted gap is between the worst and best off as in approach (ii) “narrowing the gap” or across the population as a whole, as this approach, (iii) “levelling up”, seeks to do. However, can these gaps be closed or is this simply rhetoric (i.e., lacking sincerity)? Without specificity around the details of the gap to be closed—which will vary based on ideas around health equity, demographic, geographic, and other factors—this contributes to the vagueness around health equity. For instance, how will any potential closure of the gap be signaled (e.g., across which groups? In relation to which health indicators?).

• (iv) Unspecified approach: striving for a baseline level for all

Despite not being explicitly presented as a “main approach” in select WHO texts [17, 33], some discourses seem to be about striving for a baseline level for all (i.e., a minimum level or standard). Arguably, this differs from the third “main approach” outlined within select WHO texts (“reducing inequities throughout the whole population”) as it does not have a comparator group of the best off. Rather, it seeks to ensure a certain standard level for all (e.g., “the Commission views certain goods and services as basic human and societal needs—access to clean water, for example, and health care” [16]). This approach of striving for a baseline is illustrated through texts discussing quality health services and ensuring everyone has access independent of the ability to pay [16, 34]; the need for primary care, as “inequities are exacerbated by health care systems that do not provide essential noncommunicable disease services through a primary health care approach” [31]; the need for an essential public package of interventions [31]; and the sample interventions of population-wide automatic water fluoridation for oral health [31], the distribution of free booster seats [31], and provision of basic services such as safe water, sanitation, and hygiene to improve undernutrition [34].

Actions that target the whole population, or universal approaches such as this, were identified as promoting the dignity and self-respect of those most in need and as more politically acceptable [16]. Support for this approach was demonstrated in discussing a study of subsidized versus free distribution of insecticide-treated mosquito nets in Kenya, where the latter led to “near-perfect equity” instead of a “reduction” [31].

Notably, seeking to ensure a certain baseline may be dependent on improvements across the whole population or all quintiles. For example, in discussing inequities in access to skilled birth attendance, it was noted that “very high overall levels are impossible without high levels of access in all quintiles” [31]. Therefore, while expressed as a potentially distinct goal, there may be interdependence with other goals.

• (v) Unspecified approach: improving average health

Like the above approach, there were also mentions of improving the health of the whole population through improving average health. Thus, this approach differs from (iii) “levelling up,” as it does not draw on a better-off comparator to establish benchmarks that are strived for. Nor does it align with (iv) “striving for a baseline level for all,” as it does not seek to establish a baseline for all. Instead, improving average health (with the specific outcome determined by the indicator in question) is sought across the whole population. This is illustrated through the CSDH’s Target 2, “halve, between 2000 and 2040, adult mortality rates in all countries and in all social groups within countries” and Target 3, “reduce by 90%, between 2000 and 2040, the under-five mortality rate in all countries and all social groups within countries, and reduce by 95%, between 2000 and 2040, the maternal mortality rate in all countries and all social groups within countries” [16]. These two targets were noted to be “based on the principle that decreases in mortality should be at least proportional across countries and across social groups within countries. More specifically, countries and social groups with the highest mortality levels should achieve at least the same proportional mortality decline as countries and social groups with lower mortality levels” [16]. Thus, the CSDH also noted that with the achievement of these targets, absolute health inequity will decline, whereas relative health inequity may or may not decline. One caution raised for this approach is that it is possible to reach certain targets (e.g., MDGs) without making improvements in the poorest quintile, which is plausible when considering which groups are more likely to adopt health interventions [31].

• (vi) Unspecified approach: reallocating resources

This unspecified approach is the “most direct way to address inequities” and entails “affirmative action in the sense that resources should be directed to specific areas, communities and population segments, either as a reallocation of existing funds or as a mobilization of additional funds” [31]. Discourses around reallocating resources from the more powerful to the less powerful come through when discussing the power relations and power asymmetries between groups in the population that are inherent to the pursuit of health equity. For example,

“inequity is intrinsically related to power relations and control of resources. Attempting to reduce inequities in public health inevitably means confronting the more powerful to benefit the less powerful, whether at the greater societal or the individual health clinic level. Comprehensive intervention strategies therefore need to include approaches to dealing with resistance and opposition” [31].

Thus, this approach would seek to reallocate health resources. Althoughle the first three approaches can theoretically entail reallocating resources, they describe actions in terms of adding resources to groups instead of shifting resources from the more powerful to the less powerful (i.e., these discourses are not made explicit). Ultimately, this approach is anticipated to invoke backlash or resistance from the more powerful, whereas that is not the case for the other approaches.

• (vii) Unspecified approach: striving for a full health potential for all

Lastly, discourses around seeking to ensure all achieve their full health potential arise as another unspecified approach to health equity. This is illustrated with language like “equity in health implies that ideally all individuals should attain their full health potential” [31], through the mention of Article 12 of the International Covenant on Economic, Social and Cultural Rights, “the right to the conditions necessary to achieve the highest attainable standard of health” [16], and as reiterated in the Rio Declaration on the SDH as “the enjoyment of the highest attainable standard of health is one of the fundamental rights of every human being without distinction of race, religion, political belief, economic or social condition” (originally stated in the WHO constitution, Declaration of Alma-Ata, the Ottawa Charter for Health Promotion, and others) [19]. This approach reiterates that health inequities are avoidable, and thus can be addressed, and acknowledges that individuals have inherent differences in capabilities that result in each achieving a differing level of health. As with Amartya Sen’s Capabilities Approach, there is recognition of the challenges around how to make such an approach operational, including by Sen himself [39]. For example, how can it be determined that one has reached their full health potential? Such a question will be context-sensitive and difficult to put bounds on (e.g., at what point should interventions stop?) Thus, this approach may be more of an ideal rather than a tangible sought-after approach.

Overlap of approaches

The three specified approaches were “interdependent”, as expressed in two WHO texts [17, 33]. This is exemplified through statements such as “we need to be concerned with both material deprivation […] and the social gradient in health that affects people in rich and poor countries alike” [16], “geographical or group-specific targeting and universal access are not contradictory policy approaches” [16], “develop policies that are inclusive and take account of the needs of the entire population with specific attention to vulnerable groups and high-risk areas” [19], and “achieving health equity depends substantially on changing the inequitable distribution of physical environments, with a focus on those with the highest health needs” [34], all of which demonstrate both the first and third approach.

When drawing on the “unspecified approaches,” the interdependence of the approaches also becomes clear. For example, goals (iv) “striving for a baseline level for all” and/or (v) “improving average health” may be linked with (iii) “levelling up” through the WHO’s statement that “healthy urban governance … that promote a higher level and fairer distribution of health in urban settings, is a key and critical pathway for reducing health inequity in cities” [30].

Focus on SES/SEP versus other SDH

Lastly, in evaluating whether WHO texts tended to focus on reducing inequities between those with different SES/SEP or also considered broader SDH, our analysis found that they appeared to do the latter. Although documents that sought to describe health inequities, such as Closing the gap in a generation and Equity, social determinants and public health programmes, tended to emphasize quintiles centered on SES/SEP and the social gradient when describing health inequity, the Urban HEART texts sought to discuss health inequity in terms of broader determinants, as exemplified through the breadth of Urban HEART indicators [17]. While both SES/SEP and other SDH featured heavily in discourses around health inequity, it is notable that SES/SEP was given substantially more attention than other individual SDH.

In addition to considering both SES/SEP and other SDH, WHO texts referred to broader freedoms as a prerequisite to health equity, such as political voice, support and solidarity, acceptance of values, and others as presented below. However, these aspects were not assessed by WHO texts, such as through investigating these as variables or graphing across these axes. While we recognize that these variables are not necessarily conducive to quantitative analysis, investigation of these variables could convey a desire to disrupt where power lies.

Focus on measuring and acting on individual socioeconomic status

Having the three “main approaches” built on quintiles centered on the social gradient certainly supports the view that WHO texts largely base their health equity discussion around SES/SEP. Interestingly, the Global report on urban health indicated the use of both Multiple Indicator Cluster Surveys and Demographic Health Surveys but excluded surveys that “did not have wealth information and, therefore, were not suited for inequality analysis” [34]. Given that proxy variables can be used to account for broader SES/SEP, this perhaps emphasizes the weight given to wealth in their analysis.

However, despite the large focus on wealth and SES/SEP, this is not to say other SDH were not considered, but rather, this focus may be interpreted as a sign that SES/SEP is central for improving health equity, as it is strongly related to other determinants of health. This is reflected through stating both that “the evidence points overwhelmingly to social stratification as a key determinant of illness, with inequitable health outcomes arising from differences in ethnicity, gender, age, income, education, locality and type of occupation” [32] and contextualizing this to a population group: “to be among the urban poor is an overriding vulnerability in itself that is often compounded by other factors, such as gender, age, migration status and place of habitation” [34]. This relationship between poverty and un/healthy behaviors was noted to be reciprocal, illustrated through an example of tobacco control [31]. Alternatively, the focus on those with lower SES/SEP can be attributed to an understanding that this group is a relatively easy group to target, given the availability of data when contrasted with data on ethnicity, religion, or others. Or that targeting of this group is less politically tense than seeking to target other groups (e.g., refugees) or change the systems that result in large differences in SES/SEP to begin with.

Considering other determinants of health

When discussing health equity or health inequity, there is a wide range of social determinants considered in WHO texts. For example, when providing a brief overview of inequity in health conditions, these texts included indigeneity, race, gender, education, income of country resided in, in addition to SES/SEP [16]. Similarly, this broader consideration of the social determinants is reflected in the division of determinants of health by the CSDH into “structural,” which includes income, education, gender, age, ethnicity, and sexuality, and “intermediate,” which includes living and working conditions, housing, and access to healthcare and education [30]; the SDH were divided into policy domain categories of physical environment and infrastructure, social and human development, economics, and governance in Urban HEART [17]; stating that “approaches to improve urban health equity must fundamentally address the structural roots of poverty and the broader social and environmental determinants of health” [34]; and, lastly, commitment of mayors to “address all—social, economic and environmental—determinants of health” [35].

We also want to note that beyond SDH, the WHO texts also discuss commercial and political determinants of health. First, the WHO texts mention the private sector and allude to the commercial determinants of health (with the only explicit mention of “commercial determinants of health” in Promoting health in the SDGs, Report on the 9th Global Conference for Health Promotion: All for health, health for all [35]), defined as “strategies and approaches used by the private sector to promote products and choices that are detrimental to health” [40]. This is most evident in Closing the gap in a generation, where the second recommendation is to “tackle the inequitable distribution of power, money, and resources” [16] and recommendations are put forward around market responsibility.

Considering individual freedom

The CSDH’s discussion of social inequity also appears to consider individual freedom—for example, through noting that “inclusion, agency, and control are each important for social development, health, and well-being. And restricted participation results in deprivation of human capabilities, setting the context for inequities in, for example, education, employment, and access to biomedical and technical advances” [16].

This is accentuated through using language like “leading a flourishing life” [16] and discussions of the “poverty of opportunity, of capability and of security,” which makes the point that “poverty should not be considered only in terms of ‘dollars per day’ of income, but also in terms of these social conditions, sometimes expressed as ‘relative marginality’” [30]. Evidently, this broader focus encompasses “aspirations for human security, including protection against poverty and exclusion, and aspirations for human freedom” [16]. It is about “political, economic, social, and cultural [dimensions]” [16], including “formal rights” and “the conditions in which rights are exercised,” “access to and distribution of material resources necessary to sustain life,” “relationships of support and solidarity,” and how “diversity of values, norms, and ways of living contribute to the health of all and are accepted and respected” [16].

Thus, these discourses move beyond typical considerations of SDH like age and gender, but include factors such as “political voice” [16, 30], “long-term security” [16], being “included in the society in which one lives” [16], “fulfilment of their aspirations and capabilities” [33], “right to the city” [34], and others.

Discussion

The findings of this study demonstrate that WHO texts largely approach health equity measurement through comparing groups; explicitly specify three goals/approaches to health equity ([i] “targeting disadvantaged population groups or social classes”; [ii] “narrowing the health gap”; and [iii] “reducing inequities throughout the whole population”), although discourses also support four additional unstated approaches ([iv] striving for a baseline level for all; [v] improving average health; [vi] reallocating resources; and [vii] striving for a full health potential for all); and consider health equity both in terms of socioeconomic status and other social determinants of health. Cross-comparing the findings across these three outlined inconsistencies allows for further reflections to emerge, as discussed below.

First, WHO texts analyzed present an unclear mix of approaches and goals to health equity. While an interdependence of approaches is indicated explicitly and implicitly in WHO texts, questions around which approaches should be sought, to what extent, in what order, etc. are left unanswered. These questions are important to discuss, as without interrogation, an unclear path for moving forward to tackle health inequities remains.

Despite WHO texts explicitly outlining three main approaches to targeting health inequity [17, 33], the emergence of unspecified approaches illustrates that more clarity is needed about whether the unspecified approaches should be considered and which approaches are most widely recommended in which contexts. For instance, with WHO texts’ large emphasis on improving the health of certain population groups, such as migrants, women, urban poor, and Indigenous Peoples, efforts in this area seem to align with approach (i) “targeting populations.” However, with the acknowledgement that this is both a mandatory and partial response—and also noting that while targeted programs can work toward universalism, it can be difficult to scale them up to address health inequity [16]—it raises questions on how best to proceed. Similarly, with approaches (ii) “narrowing the gap” and (iii) “levelling up” both being recognized independently as being the most accepted, this obscures the best way forward—and this is only when considering the main approaches specified in WHO texts. This unclear way forward is further compounded when considering the unspecified approaches. Considering the example from above when discussing overlapping approaches, where there was a potential linkage of approach (iv) “striving for a baseline level for all” and/or (v) “improving average health” with (iii) “levelling up”, it is unclear which approaches are being sought; therefore, the need to clarify aims becomes apparent.

Further, not all approaches can be materialized in practice, despite statements like “cities are capable of eliminating preventable, premature death for all people, and that is eminently within reach” [34]. As discussed above, achieving full health potential for all is difficult to make tangible. How are targets set? When does health equity work stop, if at all? Thus, while WHO texts state these aims, it seems there is an unacknowledged understanding that many of these aims are difficult to make tangible or, simply, they may be unattainable. This raises two questions: “what approaches can be materialized, and which ones are simply rhetoric?” and “are all approaches rhetoric if many can be considered unattainable?” As per CDA, it is important to acknowledge that these discourses are not just simply text regurgitated in documents, but they are social and political practices. Therefore, by promoting discourses of unattainable action, policy will follow—particularly policy that does not seek to disrupt current systems of power—thereby contributing to the persistence of health inequity. In other words, by promoting empty rhetoric, the status quo is undisrupted and meaningful change is not sought. CDA’s focus on power relations allows for understanding that certain discourses may be enabled over others. By presenting technical rather than politically disruptive options, the WHO is more likely to secure buy-in from Member States. This rhetoric promotes the relevance and legitimacy of the organization as one that can work with countries, and improve health equity, as opposed to one that is perceived as being radical and thus, disregarded. Preserving its legitimacy is also important to maintain its core work which includes normative tasks [13]. With the WHO’s normative role in setting standards is often focused on providing technical biomedical guidance, we can understand the inclination to move in this direction for work on health equity. It is our view that by only approaching health equity through discussing the first three approaches, the WHO limits its ability to achieve its stated objective to improve health equity. One may argue that the WHO has a hidden agenda to allow for some health inequities as a trade-off to maintain its power. This challenge is likely rooted in the WHO’s role as a political organization. It must speak the language of health equity, which would require major structural changes, while heeding the preferences of funders and powerful Member States for more minimal programmatic changes.

Second, WHO texts present competing discourses that would benefit from additional scrutiny. With WHO texts’ discourses around individual freedom, this requires more tangible examples of practice or analysis to support the WHO’s efforts. For example, in stating that “poverty should not be considered only in terms of ‘dollars per day’ of income, but also in terms of these social conditions, sometimes expressed as ‘relative marginality’” [30], which one of the health equity approaches or goals outlined above reflects this broader consideration of health equity? Because it is relative marginality, it appears there is a requirement for a comparator group, already eliminating certain approaches specified above. Overall, this broader approach considering individual freedom reflects the work of Amartya Sen, which is further substantiated with Sen’s work being cited in these discussions and Sen’s role as a Commissioner on the CSDH [16]. Focusing on individual freedom should, in theory, allow for evaluating diverse concerns and dimensions, such as targeting of structural contributors to health inequity. However, focusing on individual freedom has also been critiqued as being vague or abstractly presented—among other observations, such as being excessively individualistic and not considering the implications on others’ freedoms [41]—which we believe has the potential to complicate appropriately addressing power imbalances in society and dismantling oppressive structures. Ultimately, Sen’s scholarly work does not fully align with some of the approaches the WHO presents. Thus, it is evident that there are competing ideologies at play resulting in the manifestation of different discourses in WHO texts. These competing discourses are further compounded by political and economic interests in global public health—including colonial legacies that constitute the global health enterprise [42, 43] and—that result in more technical and less disruptive options to improve health equity.

And lastly, in considering the specified approaches alongside the unspecified approaches that emerge through analyzing discourses, we can further apply CDA to consider the operation of power. It can be observed that the three approaches explicitly presented in WHO texts add resources to select groups. These three approaches do not disrupt current systems of power by removing resources from another group. However, with the unspecified approach of reallocating resources, this shifts resources, thus disrupting power. In considering the WHO’s position seeking to support its Member States, it is understandable that “radical” approaches are not explicitly expressed to ensure that governments will be agreeable to addressing health inequity. Another study found that the term “unnecessary” in discussing health inequities was scrutinized by governments, which may be in an effort to shirk their responsibility to act on the root causes of health inequities [44]. With this knowledge, the WHO’s desire to approach addressing health inequities in a more technical manner may be strategic but may limit its ability to affect long-term systemic change. Further, this logic can also be applied to the observation that none of the WHO’s stated approaches discuss improving broader freedoms, such as political voice. If power shifts in this way, it can destabilize governments and inhibit the WHO’s legitimacy and ability to influence policymaking with Member States.

Limitations

This study sought to understand where discourses in WHO texts fell with respect to the three inconsistencies. A key strength is that it empirically tests findings from the scoping review that identified inconsistencies in the literature [3] using CDA methodology. One potential limitation arises from the ability of the selected texts to represent the WHO as a singular organization. For instance, texts produced by headquarters may not necessarily reflect the views or positions of all WHO regions or country offices or result in associated practices being taken up; or reflect that WHO rhetoric can change over time or with various influences, such as funders and other external shocks. In addition, by nature of the study design—a CDA seeks to uncover previously undiscussed discourses—there will be discourses that were not identified, discussed, or elaborated on that fall outside of these inconsistencies. And similarly, there may be additional discourses that align with these inconsistencies across domains outside of health promotion, the SDH, and urban health, or in texts not included in the study. This study is not intended to catalog all discourses embedded in WHO texts, nor can the selection of texts constitute the breadth of the WHO’s work across these domains. We believe it would be worthwhile for future analyses to investigate programmatic areas and other aspects of the WHO’s work.

Conclusions

Study findings demonstrate that there is misalignment with the WHO’s stated approaches to tackle health inequity and its discourses around health equity. This incongruence increases the likelihood of pursuing short-term solutions and not sustainably addressing the root causes of health inequity. Failing to discuss what health equity entails, what goals should be sought, and what the focus should be on, among other crucial questions, poses the risk that policy and practice may miss key considerations and that individual or group stakeholders will strive for different outputs and outcomes. Therefore, having discussions about what health equity means and establishing health equity outcome(s) sought should be clearly discussed at the outset of policy and practice and throughout. For example, is the focus to strive for health equity across an outcome or the determinants of said outcome, or both? Or when considering discourses around freedom, how will this be measured/assessed and subsequently strived for? The importance of clarity for aims and the need to define and refine aspirations early on in program and policy development is emphasized. Such discussions can also aid in moving toward more challenging conversations, such as around the redistribution of resources with Member States.

With the CSDH’s recommendation that “by 2010, the Economic and Social Council, supported by WHO, should prepare for consideration by the UN the adoption of health equity as a core global development goal, with appropriate indicators to monitor progress both within and between countries” [16]; the World Health Assembly’s resolution WHA62.14, where the WHO was encouraged to politically commit to policies with a focus on health inequities [31]; and broader Sustainable Development Goals’ (SDGs) focus on equity [34], more attention is needed to better understand the concept of health equity, which has been sought for the past several decades. Thus, the nuances and considerations raised through this study should be of value for guiding future work in this domain, including spurring investigation into conceptual approaches to combat health inequity beyond the work of the WHO [45].