INTERVENTIONS The trouble with inequalities in global health partnerships An ethical assessment

In this essay, I offer a philosophical–ethical analysis of inequalities in global health partnerships. Using literature from medical anthropology and the health sciences as a basis, I begin by distinguishing two categories of concern. First, I identify the inequalities between partners, such as between research institutions in the United States and African countries, which can include resource, epistemic, and power inequalities, and, second, I highlight associated concerns such as the lack of acknowledgement of inequalities. I then focus on what might be ethically wrong with these inequalities, emphasizing that there can be significant instrumental and noninstrumental harms associated with them. By underscoring what may be ethically troubling about inequalities in global health partnerships, this essay provides preliminary guidance on how to create more equal and more equitable relationships between partners in the field of global health.

Global health partnerships often consist of highly unequal partners (Crane 2010(Crane , 2013Okwaro and Geissler 2015;Okeke 2016). The participating agents and institutions representing one partner, usually the partner in the global North, can have much greater decision-making power and access to resources and opportunities than those representing the other partner, often from the global South. Additionally, these inequalities are set within a context of postcolonialism and the dominance of neoliberal global economic policies and institutions, which only exacerbate them (Crane 2010;Gill and Bakker 2011;Crane 2013). Yet referring to these as 'partnerships' or 'collaborations' connotes equality and consensus, seemingly neglecting the inequalities that underlie them (Crane 2010, 90, 93;Geissler 2013). Medical anthropology and the health sciences are developing an expanded literature detailing these inequalities. What they do not focus on in depth, due to their disciplinary boundaries, is the 'trouble' with them or, in other words, the normative ethical problems with these inequalities. 1 In this essay, I examine the ethical problems with inequalities between global health partners, especially those between research units or universities in the United States and Africa, 2 from the disciplinary perspective of applied ethics in philosophy. I take concerns about inequalities in global health partnerships that have been identified by four authors working in medical anthropology and the health sciences -Johanna Crane (2010Crane ( , 2013Crane ( , 2018, P. Wenzel Geissler (2013), Geissler and Ferdinand Okwaro (2014;Okwaro and Geissler 2015), and Iruka Okeke (2016Okeke ( , 2018) -as my starting point. I first identify the kinds of inequalities relevant to many global health partnerships according to those authors, and parse them to show that there are distinct but interdependent inequalities. Furthermore, I indicate how some of the concerns highlighted in the current literature are not about inequalities per se but about associated concerns such as the lack of acknowledgement of the existence and significance of these inequalities. In the second part of the essay I focus on what is wrong with these inequalities and associated concerns. I argue that when it comes to inequalities per se, from a philosophical standpoint, they are instrumentally harmful as they are often detrimental to promoting health and to good science, but they should also be viewed as being inherently problematic. Additionally, the associated concerns are problematic partially in instrumental relationship to the inequalities -they appear to interfere with achieving greater equality -but they also have the potential for other problematic ethical features such as relational toxicity, exploitation, and epistemic injustice. I describe these problems toward the end of the essay. This analysis does not take the form of a comprehensive identification and full ethical assessment of global health partnerships; instead it offers an exploratory philosophical discussion of why global health scholars and practitioners should be troubled by what appears to be a lack of equality in many global health partnerships. 3 My focus is only on the harms associated with inequalities in global health partnerships. 4 Recognizing that there are harms, however, does not mean that these partnerships do not have benefits; they often do. For example, they can lead to the provision of free or more affordable health care resources, such as HIV medications, in low-income countries (see Crane this issue), or the enhancement of laboratories and their capacities for diagnostic testing (Okeke this issue). 5 Why is it important to conduct ethical assessments? Of course, aiming to act ethically whether in a private or professional capacity, aiming to design and implement ethical policies and practices, and aiming to achieve ethical (and by implication, equitable) social systems are in and of themselves self-evidently significant goals. However, we can also refer to the often self-stated aims of global health as a field as further reason to be concerned about the ethics of inequalities associated with global health partnerships. The field of global health aims at distinguishing itself from the paternalism of international health and tropical medicine, and appears to aspire to mutually beneficial relationships between collaborators and to achieving greater equity by, for example, improving the health of those in resource-poor settings (Koplan et al. 2009). Thus, according to what appears to be global health's own goals, it is important for those working in this field to mitigate or eliminate problematic inequalities. Global health practitioners and organizations could use ethical assessments such as this one as a basis to help guide and make more equitable their international collaborations.

Inequalities in global health partnerships
In this section I use published ethnographic material and literature reviews to delineate two categories of ethical concern in contemporary global health practice: (1) the actual inequalities that are systematically to the detriment of African partners, including resource, epistemic, and power inequalities; and (2) associated concerns, which include the lack of acknowledgement of existing inequalities, the denial of inequalities, and the taking advantage of inequalities. Geissler and Okwaro (2014, 303) refer to a number of resource inequalities between colleagues or peers working in the field of global health, for example, individual researchers at African universities are more likely to 'have lower salaries than their northern counterparts and smaller allowances for health, housing, retirement or their children's education'. Additionally, resource inequalities are not limited to those inequalities between individual members of the partnerships themselves. Global health partnerships can have consequences for society-wide resource inequalities. They can reinforce or worsen the existing health and health care inequalities they are supposed to be narrowing; for example, the international funding of HIV research in African countries can bypass national health care systems and contribute to the internal resource drain on them by enticing domestic physicians away from patient care in rural areas or from the national system entirely (Pfeiffer et al. 2008;Crane 2013, 128, 138).
Inequalities in resources between partners can have significant effects beyond the lack of resources per se. Consider, for example, 'young African researchers at an international conference, sleeping and eating outside the five-star venue and missing informal scientific exchanges to save their allowances' (Geissler and Okwaro 2014, 303). Excluded African researchers are thus unable to take up opportunities to network and form collaborations, and develop their ideas in the same way as their Northern peers. This can have additional knockon effects, for example for their opportunities to further their careers and secure grants. Such resource inequalities also mean that researchers are unable to be on an equal footing with their peers from the global North in terms of knowledge production. These latter inequalities can be described as 'epistemic inequalities', which are inequalities associated with knowledge; more specifically in this case they are inequalities in how individuals are treated as 'knowers', for example, as fellow scientists.
Epistemic inequalities leading from resource inequalities can also be seen in the lack of reciprocity in educational opportunities for African students in global health. While students from the United States are often afforded the opportunity to study abroad with their medical school's African partners, African students are seldom afforded similar opportunities to study at US medical schools (Erikson and Wendland 2014). We can refer to educational inequalities as resource inequalities, but they also constitute and can have consequences for the abilities of African students to achieve equal status in the scientific community, and thus they lead to epistemic inequalities.
In her review of international research on bacterial comparative genomics, Okeke (2016) emphasizes the lack of opportunities that are available for African researchers as epistemic equals in relation to their Northern counterparts. This includes a lack of opportunities to lead scientific projects, design studies, conduct experiments, and take on more than a peripheral role in writing papers (Okeke 2016, 462-68). Okeke shows that the contribution of African authors to scientific papers in this field is primarily in providing materials, for example biological specimens and demographic information (462, 464). Furthermore, as Okeke argues in this collection, highly trained African scientists often have to take on administrative tasks within partnerships that their international collaborators avoid, which distracts from their own scientific work and teaching. Similarly, Boum's essay in this collection also highlights inequalities in the processes of knowledge production that, among other problems, undermine opportunities for local African investigators to influence research protocols and to direct research to address health concerns relevant to their local communities.
Immense power inequalities are among the numerous inequalities in global health partnerships that Crane (2010Crane ( , 2013 has identified. These include direct inequalities in decision-making power between partners. One African interviewee refers to this inequality as being required to 'dance to the other person's tune' (Crane 2013, 134-35). She provides the example of a group of physicians in Uganda being asked to change patient care practices and protocols to fit their US partner's research requirements, even when benefit to the patient may not be immediately apparent and it requires uncompensated, additional work for the physicians. Based on their ethnographic research with scientists and employees at a medical research department in East Africa, Okwaro and Geissler (2015, 499-500) also found that global North collaborators were seen to exert decision-making control over their African counterparts. US institutions can also have tight administrative control over global health partnerships, often motivated by perceptions of African inefficiency and corruption (Crane this issue). This helps to ensure that decision-making power is maintained by US institutions and it can also lead to further resource inequalities as, for example, when US institutions receiving federal funding are reimbursed for administrative costs at much higher rates than African institutions on the same grant (Crane this issue).
Furthermore, Crane (2010, 112) has also emphasized the wider context of power dynamics: African countries are former colonies of countries of the global North and 'in a postcolonial context, the power dynamics and hierarchies of "normal science" take on additional meaning and complexity, since they are inevitably infused with the politics of national autonomy, "Western" political and economic hegemony, and (often) race'. Not only, then, are partnerships characterized and shaped by power inequalities within the partnerships themselves but they are also thus impacted by wider and historical power inequalities, which include the influences of racial discrimination and colonialism.
While inequalities themselves, and their effects of creating or reinforcing further inequalities, are the starting point that motivates this article, they are accompanied by additional concerns that are not well-characterized as being about inequalities per se. Rather, we can refer to 'associated concerns' to describe Northern partners' implicit and explicit attitudes and practices in relation to inequalities. The first of these associated concerns is a lack of acknowledgement of the existence or import of inequalities. Geissler and Okwaro (2014, 303) refer to this as 'unknowing', which is 'the tendency to avoid straightforward talk about obvious inequalities' (see also Geissler 2013). Not only do inequalities often go unacknowledged but they also seem to be denied and obfuscated by the pattern of referring to partnerships or collaborations in the first place, and by employing concepts like 'equity' to describe the quality and aims of research collaborations. As Okwaro and Geissler (2015, 495) argue, 'the phraseology of collaboration might . . . camouflage underlying asymmetry and northern dominance . . . and hinder open discussions and negotiations'.
A further associated concern is that partners from the North take advantage of the inequalities in global health partnerships or those underlying them. Inequalities between the global North and African countries, such as those in health care resources and in health outcomes, are used for gain, for example, to get a paper published or to win a grant. Okeke (2016, 457-58) highlights how research in global health is often justified in publications in terms of how it will be used to combat the impact of disease in Africa, although often without taking accountability for actually making effort to translate this research into these consequences. We should not conclude that individual researchers are necessarily to blame, but questions should be asked, for example, about whether research bodies are funding the necessary translational work. 6 Furthermore, global health as a discipline is dependent on global health inequalities. It owes its existence in part to the lack of health care resources and the high burden of disease in many low-income countries; as Crane (2013, 8) writes: 'poverty and inequality are invoked as both the enemy and paradoxically, the fuel of global health'. The discipline itself could arguably be seen to take advantage of low-income countries.
Okeke and Crane do not claim that scientists and staff from the global North are malicious or necessarily motivated by their own gain; indeed, they explicitly emphasize the compassionate and humanitarian motivations behind much global health research conducted in the North (Okeke 2016, 457-58;Crane 2010, 83-84). Global health researchers and project administrators from the US have also been known to try to work around their Southern partners' deficit in resources by finding informal ways of redirecting more money to them (Crane this issue). Recognizing that there may be something harmful about the inequalities and associated concerns of partnerships does not mean that these harms are necessarily due to malice or are intended. Additionally, as I emphasized in the introduction to this essay, recognizing that there may be harms associated with them should not obscure the good that may be brought about by many of these partnerships.
However, if we want to consider how we might improve these kinds of partnerships, preliminarily we need to understand both the harms and the benefits so that moving forward the benefits can be maintained and the harms reduced or eliminated. But what precisely are these harms?
The instrumental and inherent ethical problems with inequalities in global health partnerships From a philosophical standpoint, part of what is ethically troubling is that many of the inequalities in global health partnerships are instrumentally problematic. To say that something is instrumentally problematic means that it is problematic only insofar as it brings about certain problematic consequences. The harm lies entirely in its consequences; if it did not lead to those consequences it would not be harmful.
There are at least two ways that the inequalities in global health partnerships can be considered to be instrumentally problematic. First, where these inequalities are a symptom of, or are contributing to, a community or a country's inability to provide a good standard of health care and to do so equitably, then the impact on health care and on health is problematic. An example of this is the concern discussed in the previous section that the international funding of HIV-research in Africa contributes to draining resources from health care systems or to the inequitable distributions of them between rural and urban areas. In turn, this could be leading to poor or poorer health care being offered through the national system and in rural areas. Here we can invoke a minimal standard of social justice to defend the claim that this is ethically problematic. As I use it, 'social justice' delineates what kinds of social, economic, and political structures are needed and how significant social goods, such as health care resources, should be distributed, in order for a society to be considered fair. Although there are different theories of social justice, to be cogent such theories should, at least, require that the basic needs of citizens be fulfilled. 7 Although global health partnerships are not the primary cause of longstanding resource deficits and inequitable distributions in many African countries, the fact they contribute to these deficits is enough to indicate that they are harmful in as much as they are helping to violate the minimal standards of justice.
When we refer to this reason as articulating what is ethically troubling about inequalities, equality is treated as only instrumentally valuable: it is valuable in as much as it helps to promote a minimal standard of justice. The problem is not inequality itself, whether global inequality in health or inequality in global health partnerships, but in the actual standards of health care provided to and levels of health experienced by individuals in African countries. This means that if we achieved the standards of health care and health that were required with the inequalities associated with global health partnerships in place, then we could not say that these inequalities are problematic, or at least not for this reason of a minimal standard of justice.
A second way that we may be instrumentally concerned about these inequalities is that they are harmful for science. Students and researchers from African universities are marginalized and excluded from being full collaborators in the epistemic endeavor. This is not only an interference with the 'credit and career development' of individuals (Okeke 2016, 471). Scientific research is also being hampered: 'they create a knowledge environment in which the voices of researchers in the North are more likely to be heard' (Okeke 2016, 471). One of a number of reasons why this exclusion is detrimental to science is that biomedical tools need to be context-specific. For example, we need 'adapted reference values for nutrition or toxicity' and 'surgical-safety procedures in understaffed theatres' (Geissler and Okwaro 2014, 303). How they can be made to be context-specific will require input from local staff. As diversity of knowledge and greater inclusion from partners in the South is likely to be required in order to produce good research, ensuring that greater equality in partnerships can be considered necessary for reasons of science. When we refer to this reason though, equality is again not sought for itself but only for achieving a further end. 7 Even the harshest theories of justice, such as libertarian theories, are likely to include some basic social benefits. See, for example, Fried 2011. I don't mean to elide major differences between theories of justice nor do I deny that there is likely to be debate about what 'basic needs' are. I aim only to emphasize that it is not particularly controversial to claim that states are expected to help provide for their citizens' basic needs. The more controversial points are what is required beyond these minimal requirements and especially how much equality we should be striving for.
Considering the potential goals of many global health partnerships -promoting knowledge and health -helps us discern good instrumental reasons for greater equality between partners. While I agree that we should indeed be concerned about the instrumental effects associated with inequalities in global health partnerships, I want to emphasize that there are also inherent or noninstrumental reasons to promote equality. A productive way of understanding this is to examine inequalities in global health partnerships against a framework of social-relational egalitarianism. Simply put, according to this theory, individuals are equals in a fundamental sense, but they are often treated as if they are not (Fourie, Schuppert, and Wallimann-Helmer 2015;Anderson 1999). Individuals are all equals in an important normative sense: they all have equal moral worth. This means that no matter how many inequalities of other kinds exist between these individuals, such as inequalities in power or resources, they remain, in this fundamental sense, equals.
On the basis of this notion of equal moral worth, individuals should be treated as equals and this includes having equal social standing, but they are often denied this standing. Precisely what constitutes equal social standing (and violations of it) is open to some debate. However, treating someone as an inferior type of human being is necessarily a violation of equal social standing and any attempt to sort individuals by hierarchies of worth according to, for example, gender, sexuality, race, ethnicity, and so on, would violate equal social standing. Furthermore, vast differences in resources or power would undermine equal social standing because they can interfere with the mutual recognition, trust, and social cohesion necessary for social-relational equality.
This framework -assuming it can be justified and explained plausibly; I can only provide a basic sketch here -could provide us with further reason as to why some inequalities are ethically problematic. That is, they are problematic when they violate equal social standing or are caused by such violations. What is particularly interesting about this framework is it provides us with a reason for caring about inequality that goes beyond instrumental reasons. In fact, it provides us with an egalitarian reason why we should pursue equality. Instead of making the instrumental claim that we should pursue equality only because it helps us to achieve something else -for example, better science, greater knowledge, better health -it tells us that equality is worth pursing because it is required for equal social standing and to respect equal moral worth. If we understand equal social standing as necessary per se, in cases where it is not clear that equality will lead to additional beneficial consequences, such as the promotion of better science, we can say that having equal standing between partners should remain a goal to pursue.
Using this framework can help us to understand the equalities and inequalities in global health partnerships in particular ways. For example, we can say that the inequalities in resources and power, as well as the epistemic inequalities discussed in the previous section, are often problematic not only for instrumental reasons but also because they do not allow partners from the global South to function with equal social standing to their partners from the global North. By parsing out various forms of significant equality and inequality, this framework also helps explain why partners in the global South and North are equals in certain ways -they have equal moral worth -but are not equals in many others. Furthermore, it shows the ethical connection between these two claims; the fact that partners in the global South are not equals epistemically, for example, means that they are not equals in social standing, and this in turn violates a fundamental equality of moral worth.
Thus far I have focused on actual inequalities and claimed that we have both instrumental and inherently egalitarian reasons as to why we should be troubled by them. I have not yet discussed what may be problematic about the associated concerns also identified in the previous section: a lack of acknowledgement of the existence or import of inequalities, the denial or obfuscation of inequalities, and the taking advantage of inequalities. Close examination shows us, first, that there are also instrumental ethical problems with these associated concerns. Not recognizing inequality is instrumentally problematic because it is likely to stand in the way of addressing the actual inequalities; we cannot aim to mitigate inequalities if their existence is not acknowledged nor if their ethical significance is ignored. A reason to address the associated concerns is then subsumed under our reasons for addressing the inequalities. As long as the inequalities are ethically problematic, with the assumption that we need to try to find ways of making partnerships more equal, then it seems that the associated concerns are ethically problematic too because they may contribute to keeping these ethically problematic inequalities in place. Geissler and Okwaro (2014, 303) appear to be motivated by this kind of concern, among other concerns, when they admit that 'articulating inequalities between unequal parties is difficult' and can lead to defensiveness on the part of Northern partners, and they suggest that motivating research group leaders to initiate more explicit conversations about inequalities is one way that partners can start to eliminate 'unknowing' and increase the likelihood of improving communications and research.
Is the only ethical problem with associated concerns that they could stand in the way of mitigating resource, epistemic, and power inequalities? No, I think we also have further reasons to object. There are at least three potential further harms.
First, when Northern collaborators refuse to recognize the significance of inequalities between themselves and their Southern partners, they create conditions that are toxic to good relationships. Good relationships are built on features like trust and mutual understanding; when inequalities are not properly acknowledged and, as much as possible, addressed, these features of relationships are undermined.
Second, when it comes to Northern partners taking advantage of inequalities, we should be concerned if this constitutes exploitation. Exploitation is the wrongful use of a person or people (Logar 2010). A particular example of wrongful use is when an exchange of goods takes place but the value of the goods is disproportionate and greatly favors one party. The advantaged agent is exploiting the disadvantaged agent. When exchanges are set within a context of structural injustices, disproportionality can be built into these transactions, and even routine and consensual exchanges that seem to be fair become exploitative (Ganguli Mitra 2017). In global health partnerships, while ostensibly the exchanges are between peers -for example, between biomedical researchers -and may seem fair in many usual respects, it could be appropriate to describe the relationships as structurally exploitative, because the resource, epistemic, and power inequalities are likely to systematically skew the value of exchanges in favor of the Northern partners.
A final concern is that the lack of acknowledgment and the denial of inequalities that can characterize global health partnerships constitute epistemic injustice. This means that it is a kind of injustice perpetrated against people in their capacity as knowers (Dotson 2011;Fricker 2009). The kind of epistemic injustice applicable here is known as 'testimonial injustice'; we commit testimonial injustice when we do not give speakers' testimony its due, for example when we dismiss their reports of their experiences. If Northern partners ignore or deny the testimony of their African counterparts -such as testimony about the inequalities within global health partnerships or testimony about the necessity of developing context-specific diagnostic tools -testimonial injustice can be indicated.

Conclusion
In this essay, I have identified two categories of concern that underlie recent literature on the inequalities in global health partnerships. Parsing out the inequalities in global health partnerships and the attitudes or practices associated with them can be useful in terms of mitigating them. If we want to create more equal relationships between partners, and I have argued this is both an instrumentally and inherently important aim, we need to ensure we understand them and the precise nature of what is ethically troubling about them.
Relationships between health care workers, such as nurses and physicians, and their patients, or between scientists and their research subjects, are systematically subject to ethical assessment and are regulated by codes of ethics. However, these are not the only kinds of relationships in health and health care that can be harmful. It's strange that the highly unequal relationships among scientists, administrators, and other employees from Northern and Southern institutions would not also be subject to such ethical scrutiny. Consider, for example, the vast literature on medical ethics, which analyzes and assesses how to make relationships between physicians and their patients ethical. Clinical codes of ethics rely on this body of literature for their justification. In contrast, the literature that could be used similarly for developing and justifying codes of ethics between global collaborators is sparse. Ethical assessments, such as this one, can help to provide guidance for how to make relationships between institutions and members of global health partnerships more ethical and lay a basis for developing codes of ethics that will mitigate the harms of inequalities.
Of course, what is required is not merely ethical assessment; knowing what is problematic does not itself lead to improvement. We require the will and the commitment of Northern partners to recognize inequalities and their ethical significance, and to address them, using the guidance of their Southern collaborators to identify which inequalities are relevant to that specific relationship and how best they can be resolved or mitigated.