We are grateful to Dan Hausman, John Kinuthia, Dheepa Rajan, Benjamin Rouffy-Ly, Sara Bennett, and Maria Merritt for their thoughtful and constructive engagement with Open and Inclusive: Fair Processes for Financing Universal Health Coverage (‘the Report’) (World Bank, 2023). We appreciate that there are many areas of agreement and have learned from their criticism and ideas. In what follows, we primarily focus on areas of apparent disagreement, as this is where discussion can make the most progress.
At the start, we must clarify our relationship to the Report. We are three of its seven co-authors. The Report also received extensive input from an expert panel, seven country case studies of health financing decisions (Gopinathan et al., Reference Gopinathan, Dale and Evans2023), and from the organisations that co-published it: the World Bank, the Norwegian Institute of Public Health, and Bergen University's Center for Ethics and Priority Setting. Here, we present our own views. While we put forward ideas in the Report's spirit, this response should not be taken to represent the views of our other co-authors or its institutional backers.
Since our critics mostly develop complementary lines of discussion, we proceed by addressing each critic's views separately in subsequent sections. However, where they raise similar concerns, we address them jointly.
1. The value of procedural fairness
Hausman's incisive and challenging comments (Hausman, Reference Hausman2025) afford an opportunity to clarify and defend the Report's perspective. His first key claim is that substantive fairness – equity in the distribution of benefits and burdens, rights and responsibilities – is the central value in financing Universal Health Coverage (UHC). Procedural fairness – equity in how decisions about who gets what and who pays are made – matters only when it contributes to substantive equity or when principles of substantive equity leave the precise design of arrangements for raising, pooling, and spending resources for health underdetermined.
In reply: Naturally, we agree that substantive fairness is of great importance. As the Report states (World Bank, 2023: 8), its discussion of fair processes in deciding how to finance UHC is intended to complement work that focuses on principles of distributive justice in health (e.g., World Health Organization, 2014). Moreover, one of the key reasons it puts forward for attending to procedural fairness is that doing so can promote substantive equity by ensuring that the voices of those who are often marginalised are heard and the interests of those who are often neglected receive due consideration (World Bank, 2023: 13).
But procedural fairness is valuable in ways beyond its usefulness in promoting substantive fairness and beyond its contribution to making decisions when substantive principles of distributive justice in health are indeterminate. In health financing, people's core interests in health and financial security are at stake. These interests often conflict. Moreover, in making these inter-personal trade-offs, different values or principles of justice may need to be balanced against each other. By way of illustration, providing coverage for dialysis in a low-income country may assist some of the worst off in terms of health and financial risk, but also require resources that could instead be used to improve average population health to a far greater extent (Voorhoeve et al., Reference Voorhoeve, Tan-Torres Edejer, Kapiriri, Norheim, Snowden, Basenya, Bayarsaikhan, Chentaf, Eyal, Folsom, Halina Tun Hussein, Morales, Ostmann, Ottersen, Prakongsai, Saenz, Saleh, Sommanustweechai, Wikler and Zakariah2017). There are often differences of opinion among the affected population on how to make these trade-offs (Baker et al., Reference Baker, Mason, McHugh and Donaldson2021). Furthermore, people's understanding of the values that are at stake and how they are promoted (or set back) by the decision is crucial for their acceptance of how burdens and benefits will be distributed and therefore for the sustainability of the system. For instance, a system of health financing is more likely to function well and endure if the public and actors within the system appreciate core elements of its rationale, such as the extent to which it provides prudentially valuable insurance and the extent to which it embodies solidarity between rich and poor, healthy and ill. Since so much is at stake for people (both in terms of their interests and values), and out of respect for their capacities as rational agents and social cooperators, they are owed a justification for how the system functions and have a claim to participate in decisions about the structure of their health system – a claim that is recognised as part of the human right to health (Office of the United Nations Commissioner for Human Rights and World Health Organization, 2008; World Bank, 2023: 14–16).
Because health financing has these characteristics, contra Hausman, procedural fairness matters even when decision-makers' own conception of equity determines choice by selecting a particular option as more equitable than all the other feasible options. One reason is that the public may not know why a decision is substantively fair, and so require an explanation and assurance that the decision is taken on impartial grounds rather than, say, to serve the interests of a particular group. Another reason is that a considerable part of the population may espouse different values or different principles of substantive equity than the decision-makers do or may assign different weights to some values and principles. In such cases, an open, dialogic process can improve the degree of mutual understanding and may allow parties to identify common ground. It also can allow for social learning about the nature of the trade-offs at stake and the extent of disagreement (Daniels and Sabin, Reference Daniels and Sabin2008: 51; Mazor, Reference Mazor2020: 146–147).
Procedural fairness can also enhance legitimacy. It is useful to distinguish between normative legitimacy – the degree to which the state (or public agent) is morally justified in its assertion of its authority and can create moral obligations to obey its commands – and descriptive legitimacy – the de facto acceptance both of a state's (or public agent's) authority and of the need to obey its edicts (Peter, Reference Peter, Zalta and Nodelman2023). While the Report does not articulate the distinction between these two types of legitimacy, it is concerned with the contribution that procedural fairness can make to both. There are, of course, many accounts of normative legitimacy. Given the Report's aim to advance a practical framework for procedural fairness that can be supported by a variety of perspectives, it is important that the Report's claim that open and inclusive decision-making contributes to normative legitimacy gains support from three types of account: (i) those that appeal to public reasoning, (ii) those that appeal to participation, and (iii) those that rely on the need to temper social hierarchy (Kolodny, Reference Kolodny2023; Peter, Reference Peter, Zalta and Nodelman2023: sec. 3.2).
On public reason-based accounts, public bodies' power and authority are legitimate just in case they are exercised in ways that can be accepted by all ‘reasonable’ citizens. In these accounts, being reasonable means being disposed to seek and respond rationally to evidence and being motivated to find agreement with fellow citizens, conceived of as free and equal to oneself, while recognising that citizens' interests and moral values will differ (Rawls, Reference Rawls1993; Daniels and Sabin, Reference Daniels and Sabin2008). The basis of common acceptance may be substantive reasons, such as that a particular health financing policy will promote population health and reduce inequality; it may also be procedural reasons, such as that it was the upshot of a method of decision-making that gathered enough evidence and weighed all pertinent interests impartially.
On participation-based accounts, what makes a political decision legitimate is that it was arrived at through a method that provides all relevant persons with an equal opportunity to participate (Peter, Reference Peter, Zalta and Nodelman2023: sec. 3.2).
Open and inclusive decision-making can contribute to legitimacy on both accounts. Public justification of health policies, especially when it takes a dialogic form in which deliberation aims to find consensus, can bring decisions closer to being based on shareable substantive reasons. Even when such consensus on the substance is absent, the fact that the decision was evidence-based, that people had an equal opportunity to voice their views and that all relevant interests and perspectives received consideration can make it the case that people have common reasons to endorse the process by which the decision is made. Fair procedures also recognise people's claims to contribute to health-related decision-making, thereby contributing to meeting the core requirement of participation-based accounts of legitimacy.
Procedural fairness also contributes to legitimacy by reducing the degree to which state officials' superior power and authority generate objectionable relations of inferiority (Kolodny, Reference Kolodny2023: 125–144). Objectionable relations of inferiority, Niko Kolodny argues, often involve one or more of the following factors: (a) the exercise greater power or authority to advance personal interests rather than the common good; (b) a disparity in opportunities to influence the decision; (c) a lack of accountability to those subject to power or authority; (d) unmerited differences in regard, with some, less powerful groups having their interests and perspectives given less weight than warranted; and (e) the arbitrary exercise of power and authority.
A fair process puts in place what Kolodny calls ‘tempering factors’ on each of these problematic elements of unequal power and authority. Ad (a), a fair procedure demands impartiality and impersonal justification to ensure that powers are exercised for reasons that are universalisable, rather than to serve the personal interests of the decision-maker or a select constituency. Impersonal justification makes it the case that those affected by state decisions are not so much subject to a particular individual with their personal aims or idiosyncratic opinions, but rather to the decision-maker qua office holder, who is required to act on shareable reasons (Kolodny, Reference Kolodny2023: 131–134). Ad (b), the Report's principle of equality (and associated criteria in the voice domain) contributes to what Kolodny calls ‘equal influence’. Full equality of influence requires that any citizen subject to a public decision-maker's power has as much of an opportunity to influence the decision as any other citizen (either directly by having the possibility to influence the decision, or indirectly by having a possibility of influencing a higher level in the decision-making hierarchy) (Kolodny, Reference Kolodny2023: 136–138). Inequality of opportunity for influence comes in degrees, and generally, the more a decision process reduces such inequality, the lesser the extent to which it generates problematic relations of inferiority. Ad (c), through the criterion of reason-giving, the Report's conception of a fair process promotes ‘downward accountability’ – the requirement that those who wield superior decision-making power must explain to those subjected to this power how the decision was arrived at and on what grounds it was taken (Kolodny, Reference Kolodny2023: 136). This elevates the status of the person who is entitled to a justification compared to a situation in which the decision-maker has no such obligation. Ad (d), the equal and respectful consideration of each person's interests and views that a fair process requires eliminates unmerited disparity of regard (Kolodny, Reference Kolodny2023: 140–141). Finally, ad (e), oversight and institutionalisation of fair procedures lessen the degree to which decision-makers can wield arbitrary power and ensure that there are avenues through which dubious decisions can be corrected.
People tend to recognise these ways in which procedural fairness enhances normative legitimacy. This makes them more willing to accept and abide by decisions, thereby promoting descriptive legitimacy. As Tom Tyler puts it in a review of the social scientific literature:
When third-party decisions are fairly made, people are more willing to accept them voluntarily. (…) The procedural justice effects are found in studies of real disputes, in real settings, involving actual disputants. (…) Research suggests that people voluntarily cooperate with groups when they judge that group decisions are being made fairly (Tyler, Reference Tyler2000: 119).
In contrast, when a decision – even one that decision-makers and expert observers have reason to regard as substantially just – is simply imposed, this tends to generate mistrust and opposition.
Two case studies that informed the Report illustrate these points. The first involves the 2017 legislation that established Ukraine's Programme of Medical Guarantees, a unified, tax-financed health benefit package for the full population administered by a central purchasing agency (Verkhovna Rada, 2017). This legislation was in line with key principles of substantive fairness for financing UHC proposed by many experts and endorsed by organisations such as the World Health Organization (WHO) and the World Bank. It would, once implemented, help ensure that care for the poor and ill would be subsidised by the rich and healthy. In this respect, it was arguably superior in terms of substantive equity to both the flawed status quo (in which there were very high out-of-pocket payments for services that were nominally free) and an alternative payroll-based insurance system that was proposed by some, in which people's coverage would track their financial contributions. In developing and passing the legislation, important aspects of procedural fairness were followed (e.g., legal requirements on transparency, consultation, and the public provision of a rationale for the policy were met and the legislation was passed in a democratically elected parliament). Nonetheless, consultation with the public and key stakeholders (including medical professionals and academics) fell short of ideals of procedural fairness. Due to a perceived short window of opportunity, reformers aimed to push through the legislation quickly. This meant that dialogue between the band of reform-minded technocrats and the public, civil society organisations, and academics was limited. There was little engagement with the value that some opponents of the reform and parts of the public saw in a more contributory system of insurance. Moreover, associations of health professionals were not consulted because they were regarded by the reformers as being too invested in the deeply flawed status quo (Dzhygyr et al., Reference Dzhygyr, Dale, Voorhoeve, Gopinathan and Maynzyuk2023). Several experts believe that, as a consequence, the reform faced strong resistance from those who felt their views and interests were not considered. Yuriy Dzhygyr, a lead advisor to the Minister of Health at the time of the reform (later Deputy Minister of Finance and, subsequently, Deputy Minister of Defence) puts it as follows in personal communication:
I was on the side of the proponents to deliver as soon as possible (…). We were struggling to involve people in a meaningful conversation over a predominantly payroll-based system versus a system based on general taxes. I [now] see that the dilemma was not about technical choices, but about whether to have a system based on a personal link to entitlements or on solidarity. That is what we should have communicated. Some disagreements would have persisted, but the fact that we ignored them and sort of forced the decision on them resulted in a much higher resentment and backlash.
To see how a procedure that is more open and inclusive can generate more constructive attitudes, consider the case study of the decision process in Thailand on whether to include pre-exposure prophylaxis (PreP) for populations at high risk of contracting HIV in the package of services covered under their UHC programme. Even though, during the study period, Thailand fell short of meeting the requirements of a well-functioning democracy overall (Freedom House, 2024), the system for deciding which services to cover was transparent and inclusive, with substantial, institutionalised efforts made to hear the voices of many stakeholders, including civil society organisations and patient groups. This process was judged to be of value even in those instances in which Thailand's substantive criteria for inclusion (which include severity and cost-effectiveness) clearly required coverage – as was the case for PreP. The dialogic public process of gathering evidence and providing reasons was perceived by those surveyed in the study as contributing to public trust (Viriyathorn et al., Reference Viriyathorn, Sachdev, Suwanwela, Wangbanjongkun, Patcharanarumol and Tangcharoensathien2023: i41). Both the Ukraine and Thailand case studies illustrate ways in which open and inclusive decision-making can be important even when principles of substantive justice widely accepted by the policy-making community are clear in their recommendations.
Hausman's next point is to question the Report's three proposed foundational principles (equality, impartiality, and consistency across time) and seven more concrete criteria (reason-giving, transparency, accuracy and completeness of information, inclusiveness, participation, revisability, and enforcement) for procedural fairness. (For a visual representation and brief account of these principles and criteria, see the opening article of this symposium, Voorhoeve et al., Reference Voorhoeve, Dale and Gopinathan2025, Figure 1.) In the Report's view, the principle of equality requires equal representation and consideration regardless of status, gender, ethnicity, religion, income, or power. It also requires equal access to information and opportunity to articulate views, which are to be considered with equal respect (World Bank, 2023: 11). Hausman rightly points out that this leaves room for interpretation and debate, e.g., about the extent of the population entitled to equal consideration and voice.
In reply: This need for further specification of the principle of equality does not sap the principle of content. Even an incompletely specified egalitarian principle can be of use, e.g., because all reasonable ways of spelling it out will condemn some common inequalities. Moreover, once the importance of this standard of equality is accepted, the debate narrows to which ways of considering people's interests and which opportunities to voice their views are compatible with it.
Hausman also questions the principle of consistency over time in how decisions are made. In this, he is joined by Rajan and Rouffy-Ly (Reference Rajan and Rouffy-Ly2025: sec. 3). Their criticism can be distilled into two points. The first is that, in contrast to equality and impartiality, which are values, it is difficult to discern the value of consistency over time. The second is that the value of consistency, where it is discernible, is conditional on the satisfaction of the principles of equality and impartiality.
In reply: We emphasise that this principle does not require completely static procedures; it demands merely that any changes in the ways decisions are made must not be too frequent and must occur in accordance with fair procedures, rather than being ad hoc or in response to pressure from special interests. So understood, we wonder whether our disagreement runs deep. Hausman, Rajan, and Rouffy-Ly acknowledge that stability in procedures guards against bias and ensures that ‘like cases are treated alike’. Hausman further notes that consistency over time helps orient stakeholders and gives them a sense of what they can expect – no small matter when it comes to the interests at issue for both citizens and health service providers. (The importance of such dependability is illustrated by Bennett and Merritt's discussion of inconsistency over time in country funding allocations of the US President's Emergency Plan for AIDS Relief [PEPFAR] [Bennett and Merritt, Reference Bennett and Merritt2025].) We would merely add that consistency over time can offer assurance to stakeholders and the public of equal consideration and that stability in decision processes allows for learning how to run complex participatory and evaluative systems, such as the ones described in the case study from Thailand (Viriyathorn et al., Reference Viriyathorn, Sachdev, Suwanwela, Wangbanjongkun, Patcharanarumol and Tangcharoensathien2023). Together, these establish the value of consistency.
Hausman's question whether ‘rapid and large change is always unfair, regardless of its sources’ (Hausman, Reference Hausman2025: sec. 2), and Rajan and Rouffy-Ly's comments suggest the following interesting further challenge to the principle of consistency (Rajan and Rouffy-Ly, Reference Rajan and Rouffy-Ly2025: sec. 3). If a decision procedure fails to meet the demands of the principles of equality and impartiality in important ways, then consistency in the use of this procedure over time is of doubtful value, since it would bias us towards keeping in place an unfair system. Indeed, fast, substantial changes in such a status quo may be welcome insofar they represent a move towards greater equality and impartiality in decision-making. This suggests that the value of consistency is at least partly determined by the extent to which equality and impartiality are satisfied. Rajan and Rouffy-Ly further posit that once equality and impartiality are sufficiently respected, changes in decision procedures would happen only when the country context makes them appropriate. In sum, when a decision procedure lacks equality and impartiality, the value of consistency over time is attenuated, at best; when it respects equality and impartiality, consistency over time will naturally occur to the right degree. It would follow that attending to the degree to which a system of decision-making satisfies equality and impartiality might be adequate; there is little need for an independent principle of consistency over time.
We agree that consistency over time in a decision process is of greater value when this process also satisfies equality and impartiality. Indeed, the values we mention above – treating like cases alike, allowing parties to plan, not disappointing expectations, providing assurance of equal and impartial treatment, and institutional learning – are at least partly conditional in this way. For example, one typically has less of a claim against having one's expectations disappointed when these are based on the operation of an unfair decision procedure than when they are based on the operation of a fair procedure. However, we see no basis for accepting Rajan and Rouffy-Ly's further claim that once decision systems meet a threshold of adherence to equality and impartiality, any changes that take place will be fully justified. This is partly an empirical claim, for which evidence is required. Moreover, it seems conceivable that a system could change in ways that threaten the goods of consistency over time without violating equality and impartiality. For example, one could imagine an open and participatory system for deciding which interventions to cover under a country's UHC plan which kept vacillating about which among a broad family of reasonable criteria to appeal to, or about the weight assigned to these criteria. Such vacillation would fail to treat like cases alike, make it more difficult for parties to plan, inhibit learning, and might understandably raise suspicion of a lack of impartiality or equal consideration. So there remains a need for a principle of consistency over time.
Finally, Hausman writes that the Report's seven practical criteria strike him as ‘having little connection to fairness, but a great deal to do with (…) legitimate decision-making and the appearance of fairness’ (Hausman, Reference Hausman2025: sec. 2).
In reply: We reject Hausman's suggestion that there is a disconnect between procedural fairness, legitimacy, and trust. The Report's criteria embody the ideas that citizens and other stakeholders should have a voice in key aspects of health financing decisions, that decision-makers should enter into a public exchange and assessment of reasons, and that such efforts should not be at the discretion of policymakers but should be institutionalised. These things are required by procedural fairness; they also contribute to decision-making that people rightly recognise as legitimate and worthy of trust.
2. The importance of education and facilitation
Kinuthia's account of what he takes to be the Report's ‘blind spots’ (Kinuthia, Reference Kinuthia2025) offers an opportunity to highlight some pre-conditions for good public deliberation. His first claim is that the Report implicitly assumes that the mere availability of information will contribute to a well-informed citizenry and civil society. In his view, the Report thereby overlooks that information can be grasped only if there is a capacity to process and use it. To remedy this lack, Kinuthia emphasises the need for civic education.
In reply: We agree that the knowledge and capacities of the public and of civil society determine whether they can be effective interlocutors on policy and can hold decision-makers to account. We also acknowledge that some passages in the Report (e.g., on the importance of transparency [World Bank, 2023: 25]), could have benefited from making this explicit. Still, the Report does mention the need for the kind of education that Kinuthia emphasises. For example, it writes:
achieving greater inclusiveness (…) depends on investing resources to strengthen knowledge among marginalised and vulnerable populations. Developing critical thinking, communication, research, and analytical skills among these groups can enable them to more effectively engage in decision-making processes (World Bank, 2023: 35).
The Report goes on to emphasise that investment is required to raise public awareness and that budgetary information must be presented in an understandable way (World Bank, 2023: 35–36). Moreover, it discusses how mechanisms of public involvement, such as citizen panels or participatory budgeting, can create a learning environment for participants (World Bank, 2023: 27–28). The need for such learning and capacity-building also emerges in our case studies. For example, the case study on Ukraine's 2017 health financing reforms notes that one impediment to inclusive decision-making was that reformers believed that the public, local academics, and civil society organisations lacked the expertise to engage in productive dialogue about key aspects of the reforms. It concludes that investment in such knowledge would help overcome this barrier (Dzhygyr et al., Reference Dzhygyr, Dale, Voorhoeve, Gopinathan and Maynzyuk2023). The case study from Thailand discusses in detail one example of how such social learning can be facilitated (Viriyathorn et al., Reference Viriyathorn, Sachdev, Suwanwela, Wangbanjongkun, Patcharanarumol and Tangcharoensathien2023). Further discussion of how such educative and capacity-building processes can succeed is provided by the WHO in its report on social participation for UHC (World Health Organization, 2021). We agree with Kinuthia that, to allow the public to make up its own mind, the promotion of such civic learning and strengthening of civil society's capabilities should not be left solely to governments.
Kinuthia's second criticism is that that the Report lacks a discussion of the role of facilitators in discussion and how decisions are to be made when deliberations do not reach consensus.
In reply: The Report does discuss the importance of facilitation (World Bank, 2023: 27 and 36). Kinuthia's comments have, however, made us realise that it would be useful to supplement the Report with an account of what makes for good facilitation. While the matter requires further thought, a promising account, due to Afsoun Afsahi, focuses on helping participants develop the attitudes and skills that constitute ‘deliberative capital’, including civility, open-mindedness, assurance of others' willingness to contribute, as well as the ability to analyse others' arguments and find points of agreement as well as dissensus (Afsahi, Reference Afsahi2022).
We also agree with Kinuthia that deliberation cannot be assumed to lead to consensus and that fair procedures should involve clear rules on how decisions are made in the face of whatever disagreement remains after deliberation (see also Baker et al., Reference Baker, Mason, McHugh and Donaldson2021). We admit that the Report is silent on which rules might be used (e.g., decision-making by consensus where available and then by majority voting on areas of remaining disagreement, along with a publication of reasons for both the majority decision and minority dissent). Our sense is that the right approach will depend on context. Again, we recognise that it would be valuable to supplement the Report's discussion with an account of possible rules and their impacts and would welcome research into this.
Kinuthia's third point is that decision processes in health are part of overall budgeting decisions, and hence, the articulated procedural fairness principles and criteria should apply to the entire public financial management system of a country and not merely to health financing. He believes that the Report fails to appreciate this.
In reply: Contrary to Kinuthia's interpretation, the Report does not assume that its fairness framework applies only to decisions in the health sector. Indeed, it explicitly states that the process around decisions on taxes and transfers is to be evaluated using its framework (World Bank, 2023: 18). One example the Report provides is a decision to increase a wealth tax in Norway; another is Tanzania's electricity subsidies. It also discusses the decision whether to allocate resources to health or other sectors. Furthermore, it highlights the International Budget Partnership's Open Budget Survey and Public Expenditure and Financial Accountability assessments, which examine procedural aspects of a country's public finances (World Bank, 2023: 36–37). Lastly, in articulating its conception of fair decision-making, the Report draws on a review of many disciplines and fields of application, including areas of budgeting unrelated to health (Dale et al., Reference Dale, Peacocke, Movik, Voorhoeve, Ottersen, Kurowski, Evans, Norheim and Gopinathan2023). One of the ways it aims to improve on established frameworks for procedural fairness in health, such as Accountability for Reasonableness (Daniels and Sabin, Reference Daniels and Sabin2008), is precisely that it examines revenue raising, pooling, and spending decisions in health as part of the overall budget cycle, in just the way Kinuthia proposes.
3. The importance of voice
Rajan and Rouffy-Ly's first point is that the omission of the term ‘accountability’ in the framework is a missed opportunity to establish a strong connection between accountability and procedural fairness (Rajan and Rouffy-Ly, Reference Rajan and Rouffy-Ly2025). A consequence of this omission, they contend, is that the Report fails to highlight the ways in which its proposed procedural fairness framework is more valuable than traditional accountability frameworks.
In reply: In our view, in global health, the term ‘accountability’ is overused to the point of losing clarity. It has multiple analyses, including interpretations that are unconnected to citizen engagement. For example, in one paper, it is understood as mainly a ‘financial term’ (Robinson and Adams, Reference Robinson and Adams2022: 9) concerned with monitoring budgets and how money is spent in the health care system, rather than with the dialogical exchange of ideas, justification of positions, and respectful engagement with differing views that the Report aims to promote. The Report's authors therefore made a deliberate choice to refrain from using the term ‘accountability’ and instead emphasise two core elements of accountability: reason-giving (which is akin to the commonly used term ‘answerability’) and enforcement. This choice was explained in the scoping review that forms the basis for the Report's framework (Dale et al., Reference Dale, Peacocke, Movik, Voorhoeve, Ottersen, Kurowski, Evans, Norheim and Gopinathan2023: i17), but we acknowledge that it would have been useful to also clarify it in the Report.
Second, Rajan and Rouffy-Ly advocate regarding the Report's ‘voice’ domain (which encompasses the criteria of inclusiveness and participation) as the ‘linchpin’ of procedural fairness. Their motivation is that, to ensure fairness in health financing processes, power imbalances among stakeholders must be addressed. They argue that meaningful engagement with people, communities, and civil society is central to achieving such equity in influence. In their view, the Report's other domains (‘information’ – encompassing reason-giving, transparency and accuracy, and completeness of information – and ‘oversight’ – encompassing revisability and enforcement) then serve as prerequisites for such rebalancing of power in decision-making.
In reply: We agree (and the Report recognises) that procedural fairness involves rebalancing influence and power within decision processes. For example, better representation can advance the interests of marginalised groups and oversight mechanisms can hold decision-makers accountable and mitigate imbalances in the ability to exercise influence. These measures can also enhance substantive equity, because they moderate the inequality in consideration and forms of partiality that are common sources of unjust outcomes (World Bank, 2023: 13).
However, if Rajan and Rouffy-Ly intend to suggest that ‘voice’ is generally to be prioritised, with ‘information’ and ‘oversight’ primarily serving supporting roles, then we disagree. We see no such general hierarchy among the three domains. Although in some conditions, prioritising one domain may be justified, the prioritised domain need not always be ‘voice’.
One consideration in this regard is the degree to which the decisions in question are ‘directional’ or ‘technical’ (World Bank, 2023: 28–30). Directional decisions establish the value orientation of health system financing, e.g., by determining the extent to which solidarity should guide contributions to a publicly financed insurance scheme. In directional decisions, public participation is important for reasons outlined in our reply to Hausman. In contrast, technical decisions: (i) require the application of expert knowledge that lay people cannot be expected to acquire; and (ii) do not themselves determine the aims, principles, and values of (the relevant part of) the health financing system but are instead in the service of pre-specified aims and/or guided by pre-specified principles and values. In such instances, the Report posits, it can make sense to delegate decisions entirely to experts, with minimal or no public participation. The Report cites the example of the National Health Service of Ukraine, which is a body with autonomy over technical and operational decisions, including specifying services under the Programme of Medical Guarantees, selecting providers, and developing payment methods and rates (World Bank, 2023: 29).
Still, even for technical decisions, non-dialogical (unidirectional) public reason-giving is required for a fair process, because the legitimacy of technical bodies depends on the quality of their public reasoning and the public's acceptance of their justifications (Eriksen, Reference Eriksen2021). One important component of such justifications is the aims, values, and principles that have been set for these bodies. What these should be is a directional decision, and so there is reason to have these parameters for expert decision set through a participatory process.
There is, of course, a risk that decision-makers label certain decisions as ‘technical’ to avoid the implementation of mechanisms for public participation, even when the nature of the decisions calls for such participation. Moreover, many decisions that have appeared to people as purely technical in fact contain important value-driven, or directional, elements. This may be increasingly recognised; there is a trend in policy-making to employ methods for public participation, such as citizen panels, to inform decisions in health, social, and environmental policy that were previously dominated by technical experts and government officials (Mitton et al., Reference Mitton, Smith, Peacock, Evoy and Abelson2009; Street et al., Reference Street, Duszynski, Krawczyk and Braunack-Maye2014; Dryzek et al., Reference Dryzek, Bächtiger, Chambers, Cohen, Druckman, Felicetti, Fishkin, Farrell, Fung, Gutmann, Landemore, Mansbridge, Marien, Neblo, Niemeyer, Setälä, Slothuus, Suiter, Thompson and Warren2019; Alemanno, Reference Alemanno2022).
However, even when voice is important because decisions are directional, there may be reasons to improve other domains of procedural fairness first. Meaningful public engagement is demanding in terms of money, institutional capability, and time. In low-resource settings, particularly where democratic institutions are maturing, it is important to have a realistic assessment of the degree to which effective improvements in voice are feasible, at least in the short term. This is illustrated by a case study from The Gambia that informed the Report. Following the country's democratic transition starting in 2016, the process to enact the National Health Insurance scheme (which was passed in 2021) incorporated laudable ambitions for stakeholder participation. While the process ended up being more open than was common before the country's democratic transition, limited resources, time, and inadequate administrative capacity proved to be barriers to consultation with a sufficiently wide range of stakeholders. Consultation therefore fell short of aspirations (Njie et al., Reference Njie, Dale and Gopinathan2023). While, under such circumstances, it remains important to improve voice, it is possible that improvements in other domains, such as the information domain (encompassing transparency, reason-giving, and accuracy and completeness of information) may be worth focusing on first, because these may be more readily feasible and would make a substantial contribution to enhancing procedural fairness. For example, in the study period, The Gambia showed a significant improvement in budgeting and fiscal transparency, primarily due to increased public access to budget information and decisions (International Budget Partnership, Reference International Budget Partnership2021).
4. Procedural fairness in development assistance for health
Bennett and Merritt's application of the Report's framework to decision processes in development assistance for health (Bennett and Merritt, Reference Bennett and Merritt2025) represents a departure from the Report's focus on the national level. Even where the Report suggests that development partners might use the proposed framework ‘to examine their own processes’, the examples it offers involve supporting recipient countries in improving their own institutions and capacity to satisfy the procedural fairness criteria that apply to them (World Bank, 2023: 11 and 39). Bennett and Merritt's analysis of the applicability of the Report's principles of equality, impartiality, and consistency across time to development partners' own decision-making – with application to PEPFAR – therefore involves an exciting extension of the Report's ideas.
Bennett and Merritt first ask whether fairness principles for public decision-making within a country also apply to the decision processes of internationally operating development partners. There is certainly room for doubt that the answer is affirmative. After all, there are many differences between the two types of actors. National health financing decisions by state agencies typically concern the use of resources that come from its citizens (such as tax income) and involve the use of the state's coercive power (such as threatening penalties for non-payment of tax or mandatory insurance contributions). They also typically purport to be done in the name of its citizens and in service of their interests. It follows that there is a clear core constituency – the citizenry – who are owed a justification for decisions and who have a claim to an equal opportunity to participate in decision-making. Consider, in contrast, the situation of development partners that are government agencies, such as the US Department of State's Bureau of Global Health Security and Diplomacy, which runs PEPFAR. The people who supply its resources through their taxes and in whose name the agency operates are predominantly US citizens, while the people whose interests these decisions purport to advance are primarily those affected by, or at risk of developing, HIV in countries that face substantial challenges in addressing the HIV/AIDS pandemic.
Bennett and Merritt argue that these differences do not stand in the way of applying the Report's fundamental principles to decision-making by a development partner. Their first reason is that development partners should be concerned not merely with supplying health-related benefits to recipients, but also with empowering them by taking seriously their perspectives on the ways they might be affected. Their second reason is that development partners often work with or through country governments, so that the way they make decisions impacts the way recipient country governments are perceived by their populations, thereby affecting their descriptive legitimacy. One might add that where development partners work closely with recipient governments, the latter's normative legitimacy may be at stake when development partners' influence is exercised in ways that fail to meet locals' claims for participation in how their state bodies make decisions. A case study that informed the Report provides evidence of the latter idea. In Tanzania in 2017–2018, the Ministry of Health established a technical working group on reforms of its Community Health Insurance Schemes that included representatives of ministries, civil society organisations, development partners, and some of the private organisations supporting the implementation of community health fund schemes. However, beneficiaries and health service providers were not included, and there is some evidence that this lack of inclusion created a lack of trust in the scheme (Binyaruka et al., Reference Binyaruka, Mtei, Maiba, Gopinathan and Dale2023).
In reply: We agree that respect for the agency of recipients is a reason to ensure that those directly involved in and impacted by the development partner's actions have access to information about and rationales for its decisions. It is also a reason for their views (or the views of their representatives) to be sought out and considered. We further agree that where development partners work with governments, they should aim to contribute to more open and inclusive local decision-making in the areas in which they operate. However, we are not yet convinced that these reasons warrant the application of the full principle of equality as specified by the Report, which calls for ‘equal access to information, equal capacity to express one's views, and equal opportunity to influence decisions’ (World Bank, 2023: 10). After all, there are many cases in which, out of respect for individuals' agency, decision-makers must offer them some access to information about and rationales for decisions that affect them, as well as some voice, without being required to offer equal access and voice. An everyday example is the role of students in university decision-making. Students are owed information about and an explanation for policies that affect them, as well as fora in which their views are heard and respectfully engaged with. But this does not imply that their influence needs to be fully equal to that of other stakeholders, such as permanent staff. Bennett and Merritt's argument for the applicability of the equality criterion to development partners' decisions therefore strikes us as incomplete.
This point need not threaten Bennett and Merritt's criticism of PEPFAR's prioritisation policy. For their argument need not rest on a violation of the principle of equality. Instead, it can rely on the observation that this policy was developed with too little input by recipient country governments and/or other genuine representatives of affected populations, despite the substantial implications of these decisions for their access to life-saving services. Bennett and Merritt's conclusion that existing processes ‘[do] not adequately reflect the voices of in-country stakeholders who ultimately may be most affected by these policies’ can stand, since what is ‘adequate’ might merely be substantial, rather than fully equal, voice.
The question whether the Report's principles of impartiality and consistency can be straightforwardly extended to development partners requires further thought for similar reasons.
Notwithstanding these doubts, we hope that Bennett and Merritt's contribution sparks a discussion about the fairness of decision processes that shape global health financing. Many of the Report's principles and criteria for procedural fairness align with proposals for more inclusive and equitable approaches to determining development assistance to low- and middle-income countries, such as the Lusaka Agenda on the Future of Global Health Initiatives (Mwangangi and Røttingen, Reference Mwangangi and Røttingen2023). Bennett and Merritt's paper has inspired us to consider in which form the framework might be applied to evaluate processes in global health financing, such as the Global Fund's country coordination mechanisms (Sands, Reference Sands2019) and GAVI's vaccine funding allocation processes (Gandhi, Reference Gandhi2015; Nunes et al., Reference Nunes, McKee and Howard2024).
In closing, we note that a fundamental premise of the Report is that critical and open-minded exchange of ideas between people with different perspectives can improve our thinking about health financing. We are grateful to our critics for supplying evidence for the truth of this premise through their engaging contributions.
You have
Access
Open access
We are grateful to Dan Hausman, John Kinuthia, Dheepa Rajan, Benjamin Rouffy-Ly, Sara Bennett, and Maria Merritt for their thoughtful and constructive engagement with Open and Inclusive: Fair Processes for Financing Universal Health Coverage (‘the Report’) (World Bank, 2023). We appreciate that there are many areas of agreement and have learned from their criticism and ideas. In what follows, we primarily focus on areas of apparent disagreement, as this is where discussion can make the most progress.
At the start, we must clarify our relationship to the Report. We are three of its seven co-authors. The Report also received extensive input from an expert panel, seven country case studies of health financing decisions (Gopinathan et al., Reference Gopinathan, Dale and Evans2023), and from the organisations that co-published it: the World Bank, the Norwegian Institute of Public Health, and Bergen University's Center for Ethics and Priority Setting. Here, we present our own views. While we put forward ideas in the Report's spirit, this response should not be taken to represent the views of our other co-authors or its institutional backers.
Since our critics mostly develop complementary lines of discussion, we proceed by addressing each critic's views separately in subsequent sections. However, where they raise similar concerns, we address them jointly.
1. The value of procedural fairness
Hausman's incisive and challenging comments (Hausman, Reference Hausman2025) afford an opportunity to clarify and defend the Report's perspective. His first key claim is that substantive fairness – equity in the distribution of benefits and burdens, rights and responsibilities – is the central value in financing Universal Health Coverage (UHC). Procedural fairness – equity in how decisions about who gets what and who pays are made – matters only when it contributes to substantive equity or when principles of substantive equity leave the precise design of arrangements for raising, pooling, and spending resources for health underdetermined.
In reply: Naturally, we agree that substantive fairness is of great importance. As the Report states (World Bank, 2023: 8), its discussion of fair processes in deciding how to finance UHC is intended to complement work that focuses on principles of distributive justice in health (e.g., World Health Organization, 2014). Moreover, one of the key reasons it puts forward for attending to procedural fairness is that doing so can promote substantive equity by ensuring that the voices of those who are often marginalised are heard and the interests of those who are often neglected receive due consideration (World Bank, 2023: 13).
But procedural fairness is valuable in ways beyond its usefulness in promoting substantive fairness and beyond its contribution to making decisions when substantive principles of distributive justice in health are indeterminate. In health financing, people's core interests in health and financial security are at stake. These interests often conflict. Moreover, in making these inter-personal trade-offs, different values or principles of justice may need to be balanced against each other. By way of illustration, providing coverage for dialysis in a low-income country may assist some of the worst off in terms of health and financial risk, but also require resources that could instead be used to improve average population health to a far greater extent (Voorhoeve et al., Reference Voorhoeve, Tan-Torres Edejer, Kapiriri, Norheim, Snowden, Basenya, Bayarsaikhan, Chentaf, Eyal, Folsom, Halina Tun Hussein, Morales, Ostmann, Ottersen, Prakongsai, Saenz, Saleh, Sommanustweechai, Wikler and Zakariah2017). There are often differences of opinion among the affected population on how to make these trade-offs (Baker et al., Reference Baker, Mason, McHugh and Donaldson2021). Furthermore, people's understanding of the values that are at stake and how they are promoted (or set back) by the decision is crucial for their acceptance of how burdens and benefits will be distributed and therefore for the sustainability of the system. For instance, a system of health financing is more likely to function well and endure if the public and actors within the system appreciate core elements of its rationale, such as the extent to which it provides prudentially valuable insurance and the extent to which it embodies solidarity between rich and poor, healthy and ill. Since so much is at stake for people (both in terms of their interests and values), and out of respect for their capacities as rational agents and social cooperators, they are owed a justification for how the system functions and have a claim to participate in decisions about the structure of their health system – a claim that is recognised as part of the human right to health (Office of the United Nations Commissioner for Human Rights and World Health Organization, 2008; World Bank, 2023: 14–16).
Because health financing has these characteristics, contra Hausman, procedural fairness matters even when decision-makers' own conception of equity determines choice by selecting a particular option as more equitable than all the other feasible options. One reason is that the public may not know why a decision is substantively fair, and so require an explanation and assurance that the decision is taken on impartial grounds rather than, say, to serve the interests of a particular group. Another reason is that a considerable part of the population may espouse different values or different principles of substantive equity than the decision-makers do or may assign different weights to some values and principles. In such cases, an open, dialogic process can improve the degree of mutual understanding and may allow parties to identify common ground. It also can allow for social learning about the nature of the trade-offs at stake and the extent of disagreement (Daniels and Sabin, Reference Daniels and Sabin2008: 51; Mazor, Reference Mazor2020: 146–147).
Procedural fairness can also enhance legitimacy. It is useful to distinguish between normative legitimacy – the degree to which the state (or public agent) is morally justified in its assertion of its authority and can create moral obligations to obey its commands – and descriptive legitimacy – the de facto acceptance both of a state's (or public agent's) authority and of the need to obey its edicts (Peter, Reference Peter, Zalta and Nodelman2023). While the Report does not articulate the distinction between these two types of legitimacy, it is concerned with the contribution that procedural fairness can make to both. There are, of course, many accounts of normative legitimacy. Given the Report's aim to advance a practical framework for procedural fairness that can be supported by a variety of perspectives, it is important that the Report's claim that open and inclusive decision-making contributes to normative legitimacy gains support from three types of account: (i) those that appeal to public reasoning, (ii) those that appeal to participation, and (iii) those that rely on the need to temper social hierarchy (Kolodny, Reference Kolodny2023; Peter, Reference Peter, Zalta and Nodelman2023: sec. 3.2).
On public reason-based accounts, public bodies' power and authority are legitimate just in case they are exercised in ways that can be accepted by all ‘reasonable’ citizens. In these accounts, being reasonable means being disposed to seek and respond rationally to evidence and being motivated to find agreement with fellow citizens, conceived of as free and equal to oneself, while recognising that citizens' interests and moral values will differ (Rawls, Reference Rawls1993; Daniels and Sabin, Reference Daniels and Sabin2008). The basis of common acceptance may be substantive reasons, such as that a particular health financing policy will promote population health and reduce inequality; it may also be procedural reasons, such as that it was the upshot of a method of decision-making that gathered enough evidence and weighed all pertinent interests impartially.
On participation-based accounts, what makes a political decision legitimate is that it was arrived at through a method that provides all relevant persons with an equal opportunity to participate (Peter, Reference Peter, Zalta and Nodelman2023: sec. 3.2).
Open and inclusive decision-making can contribute to legitimacy on both accounts. Public justification of health policies, especially when it takes a dialogic form in which deliberation aims to find consensus, can bring decisions closer to being based on shareable substantive reasons. Even when such consensus on the substance is absent, the fact that the decision was evidence-based, that people had an equal opportunity to voice their views and that all relevant interests and perspectives received consideration can make it the case that people have common reasons to endorse the process by which the decision is made. Fair procedures also recognise people's claims to contribute to health-related decision-making, thereby contributing to meeting the core requirement of participation-based accounts of legitimacy.
Procedural fairness also contributes to legitimacy by reducing the degree to which state officials' superior power and authority generate objectionable relations of inferiority (Kolodny, Reference Kolodny2023: 125–144). Objectionable relations of inferiority, Niko Kolodny argues, often involve one or more of the following factors: (a) the exercise greater power or authority to advance personal interests rather than the common good; (b) a disparity in opportunities to influence the decision; (c) a lack of accountability to those subject to power or authority; (d) unmerited differences in regard, with some, less powerful groups having their interests and perspectives given less weight than warranted; and (e) the arbitrary exercise of power and authority.
A fair process puts in place what Kolodny calls ‘tempering factors’ on each of these problematic elements of unequal power and authority. Ad (a), a fair procedure demands impartiality and impersonal justification to ensure that powers are exercised for reasons that are universalisable, rather than to serve the personal interests of the decision-maker or a select constituency. Impersonal justification makes it the case that those affected by state decisions are not so much subject to a particular individual with their personal aims or idiosyncratic opinions, but rather to the decision-maker qua office holder, who is required to act on shareable reasons (Kolodny, Reference Kolodny2023: 131–134). Ad (b), the Report's principle of equality (and associated criteria in the voice domain) contributes to what Kolodny calls ‘equal influence’. Full equality of influence requires that any citizen subject to a public decision-maker's power has as much of an opportunity to influence the decision as any other citizen (either directly by having the possibility to influence the decision, or indirectly by having a possibility of influencing a higher level in the decision-making hierarchy) (Kolodny, Reference Kolodny2023: 136–138). Inequality of opportunity for influence comes in degrees, and generally, the more a decision process reduces such inequality, the lesser the extent to which it generates problematic relations of inferiority. Ad (c), through the criterion of reason-giving, the Report's conception of a fair process promotes ‘downward accountability’ – the requirement that those who wield superior decision-making power must explain to those subjected to this power how the decision was arrived at and on what grounds it was taken (Kolodny, Reference Kolodny2023: 136). This elevates the status of the person who is entitled to a justification compared to a situation in which the decision-maker has no such obligation. Ad (d), the equal and respectful consideration of each person's interests and views that a fair process requires eliminates unmerited disparity of regard (Kolodny, Reference Kolodny2023: 140–141). Finally, ad (e), oversight and institutionalisation of fair procedures lessen the degree to which decision-makers can wield arbitrary power and ensure that there are avenues through which dubious decisions can be corrected.
People tend to recognise these ways in which procedural fairness enhances normative legitimacy. This makes them more willing to accept and abide by decisions, thereby promoting descriptive legitimacy. As Tom Tyler puts it in a review of the social scientific literature:
When third-party decisions are fairly made, people are more willing to accept them voluntarily. (…) The procedural justice effects are found in studies of real disputes, in real settings, involving actual disputants. (…) Research suggests that people voluntarily cooperate with groups when they judge that group decisions are being made fairly (Tyler, Reference Tyler2000: 119).
In contrast, when a decision – even one that decision-makers and expert observers have reason to regard as substantially just – is simply imposed, this tends to generate mistrust and opposition.
Two case studies that informed the Report illustrate these points. The first involves the 2017 legislation that established Ukraine's Programme of Medical Guarantees, a unified, tax-financed health benefit package for the full population administered by a central purchasing agency (Verkhovna Rada, 2017). This legislation was in line with key principles of substantive fairness for financing UHC proposed by many experts and endorsed by organisations such as the World Health Organization (WHO) and the World Bank. It would, once implemented, help ensure that care for the poor and ill would be subsidised by the rich and healthy. In this respect, it was arguably superior in terms of substantive equity to both the flawed status quo (in which there were very high out-of-pocket payments for services that were nominally free) and an alternative payroll-based insurance system that was proposed by some, in which people's coverage would track their financial contributions. In developing and passing the legislation, important aspects of procedural fairness were followed (e.g., legal requirements on transparency, consultation, and the public provision of a rationale for the policy were met and the legislation was passed in a democratically elected parliament). Nonetheless, consultation with the public and key stakeholders (including medical professionals and academics) fell short of ideals of procedural fairness. Due to a perceived short window of opportunity, reformers aimed to push through the legislation quickly. This meant that dialogue between the band of reform-minded technocrats and the public, civil society organisations, and academics was limited. There was little engagement with the value that some opponents of the reform and parts of the public saw in a more contributory system of insurance. Moreover, associations of health professionals were not consulted because they were regarded by the reformers as being too invested in the deeply flawed status quo (Dzhygyr et al., Reference Dzhygyr, Dale, Voorhoeve, Gopinathan and Maynzyuk2023). Several experts believe that, as a consequence, the reform faced strong resistance from those who felt their views and interests were not considered. Yuriy Dzhygyr, a lead advisor to the Minister of Health at the time of the reform (later Deputy Minister of Finance and, subsequently, Deputy Minister of Defence) puts it as follows in personal communication:
I was on the side of the proponents to deliver as soon as possible (…). We were struggling to involve people in a meaningful conversation over a predominantly payroll-based system versus a system based on general taxes. I [now] see that the dilemma was not about technical choices, but about whether to have a system based on a personal link to entitlements or on solidarity. That is what we should have communicated. Some disagreements would have persisted, but the fact that we ignored them and sort of forced the decision on them resulted in a much higher resentment and backlash.
To see how a procedure that is more open and inclusive can generate more constructive attitudes, consider the case study of the decision process in Thailand on whether to include pre-exposure prophylaxis (PreP) for populations at high risk of contracting HIV in the package of services covered under their UHC programme. Even though, during the study period, Thailand fell short of meeting the requirements of a well-functioning democracy overall (Freedom House, 2024), the system for deciding which services to cover was transparent and inclusive, with substantial, institutionalised efforts made to hear the voices of many stakeholders, including civil society organisations and patient groups. This process was judged to be of value even in those instances in which Thailand's substantive criteria for inclusion (which include severity and cost-effectiveness) clearly required coverage – as was the case for PreP. The dialogic public process of gathering evidence and providing reasons was perceived by those surveyed in the study as contributing to public trust (Viriyathorn et al., Reference Viriyathorn, Sachdev, Suwanwela, Wangbanjongkun, Patcharanarumol and Tangcharoensathien2023: i41). Both the Ukraine and Thailand case studies illustrate ways in which open and inclusive decision-making can be important even when principles of substantive justice widely accepted by the policy-making community are clear in their recommendations.
Hausman's next point is to question the Report's three proposed foundational principles (equality, impartiality, and consistency across time) and seven more concrete criteria (reason-giving, transparency, accuracy and completeness of information, inclusiveness, participation, revisability, and enforcement) for procedural fairness. (For a visual representation and brief account of these principles and criteria, see the opening article of this symposium, Voorhoeve et al., Reference Voorhoeve, Dale and Gopinathan2025, Figure 1.) In the Report's view, the principle of equality requires equal representation and consideration regardless of status, gender, ethnicity, religion, income, or power. It also requires equal access to information and opportunity to articulate views, which are to be considered with equal respect (World Bank, 2023: 11). Hausman rightly points out that this leaves room for interpretation and debate, e.g., about the extent of the population entitled to equal consideration and voice.
In reply: This need for further specification of the principle of equality does not sap the principle of content. Even an incompletely specified egalitarian principle can be of use, e.g., because all reasonable ways of spelling it out will condemn some common inequalities. Moreover, once the importance of this standard of equality is accepted, the debate narrows to which ways of considering people's interests and which opportunities to voice their views are compatible with it.
Hausman also questions the principle of consistency over time in how decisions are made. In this, he is joined by Rajan and Rouffy-Ly (Reference Rajan and Rouffy-Ly2025: sec. 3). Their criticism can be distilled into two points. The first is that, in contrast to equality and impartiality, which are values, it is difficult to discern the value of consistency over time. The second is that the value of consistency, where it is discernible, is conditional on the satisfaction of the principles of equality and impartiality.
In reply: We emphasise that this principle does not require completely static procedures; it demands merely that any changes in the ways decisions are made must not be too frequent and must occur in accordance with fair procedures, rather than being ad hoc or in response to pressure from special interests. So understood, we wonder whether our disagreement runs deep. Hausman, Rajan, and Rouffy-Ly acknowledge that stability in procedures guards against bias and ensures that ‘like cases are treated alike’. Hausman further notes that consistency over time helps orient stakeholders and gives them a sense of what they can expect – no small matter when it comes to the interests at issue for both citizens and health service providers. (The importance of such dependability is illustrated by Bennett and Merritt's discussion of inconsistency over time in country funding allocations of the US President's Emergency Plan for AIDS Relief [PEPFAR] [Bennett and Merritt, Reference Bennett and Merritt2025].) We would merely add that consistency over time can offer assurance to stakeholders and the public of equal consideration and that stability in decision processes allows for learning how to run complex participatory and evaluative systems, such as the ones described in the case study from Thailand (Viriyathorn et al., Reference Viriyathorn, Sachdev, Suwanwela, Wangbanjongkun, Patcharanarumol and Tangcharoensathien2023). Together, these establish the value of consistency.
Hausman's question whether ‘rapid and large change is always unfair, regardless of its sources’ (Hausman, Reference Hausman2025: sec. 2), and Rajan and Rouffy-Ly's comments suggest the following interesting further challenge to the principle of consistency (Rajan and Rouffy-Ly, Reference Rajan and Rouffy-Ly2025: sec. 3). If a decision procedure fails to meet the demands of the principles of equality and impartiality in important ways, then consistency in the use of this procedure over time is of doubtful value, since it would bias us towards keeping in place an unfair system. Indeed, fast, substantial changes in such a status quo may be welcome insofar they represent a move towards greater equality and impartiality in decision-making. This suggests that the value of consistency is at least partly determined by the extent to which equality and impartiality are satisfied. Rajan and Rouffy-Ly further posit that once equality and impartiality are sufficiently respected, changes in decision procedures would happen only when the country context makes them appropriate. In sum, when a decision procedure lacks equality and impartiality, the value of consistency over time is attenuated, at best; when it respects equality and impartiality, consistency over time will naturally occur to the right degree. It would follow that attending to the degree to which a system of decision-making satisfies equality and impartiality might be adequate; there is little need for an independent principle of consistency over time.
We agree that consistency over time in a decision process is of greater value when this process also satisfies equality and impartiality. Indeed, the values we mention above – treating like cases alike, allowing parties to plan, not disappointing expectations, providing assurance of equal and impartial treatment, and institutional learning – are at least partly conditional in this way. For example, one typically has less of a claim against having one's expectations disappointed when these are based on the operation of an unfair decision procedure than when they are based on the operation of a fair procedure. However, we see no basis for accepting Rajan and Rouffy-Ly's further claim that once decision systems meet a threshold of adherence to equality and impartiality, any changes that take place will be fully justified. This is partly an empirical claim, for which evidence is required. Moreover, it seems conceivable that a system could change in ways that threaten the goods of consistency over time without violating equality and impartiality. For example, one could imagine an open and participatory system for deciding which interventions to cover under a country's UHC plan which kept vacillating about which among a broad family of reasonable criteria to appeal to, or about the weight assigned to these criteria. Such vacillation would fail to treat like cases alike, make it more difficult for parties to plan, inhibit learning, and might understandably raise suspicion of a lack of impartiality or equal consideration. So there remains a need for a principle of consistency over time.
Finally, Hausman writes that the Report's seven practical criteria strike him as ‘having little connection to fairness, but a great deal to do with (…) legitimate decision-making and the appearance of fairness’ (Hausman, Reference Hausman2025: sec. 2).
In reply: We reject Hausman's suggestion that there is a disconnect between procedural fairness, legitimacy, and trust. The Report's criteria embody the ideas that citizens and other stakeholders should have a voice in key aspects of health financing decisions, that decision-makers should enter into a public exchange and assessment of reasons, and that such efforts should not be at the discretion of policymakers but should be institutionalised. These things are required by procedural fairness; they also contribute to decision-making that people rightly recognise as legitimate and worthy of trust.
2. The importance of education and facilitation
Kinuthia's account of what he takes to be the Report's ‘blind spots’ (Kinuthia, Reference Kinuthia2025) offers an opportunity to highlight some pre-conditions for good public deliberation. His first claim is that the Report implicitly assumes that the mere availability of information will contribute to a well-informed citizenry and civil society. In his view, the Report thereby overlooks that information can be grasped only if there is a capacity to process and use it. To remedy this lack, Kinuthia emphasises the need for civic education.
In reply: We agree that the knowledge and capacities of the public and of civil society determine whether they can be effective interlocutors on policy and can hold decision-makers to account. We also acknowledge that some passages in the Report (e.g., on the importance of transparency [World Bank, 2023: 25]), could have benefited from making this explicit. Still, the Report does mention the need for the kind of education that Kinuthia emphasises. For example, it writes:
achieving greater inclusiveness (…) depends on investing resources to strengthen knowledge among marginalised and vulnerable populations. Developing critical thinking, communication, research, and analytical skills among these groups can enable them to more effectively engage in decision-making processes (World Bank, 2023: 35).
The Report goes on to emphasise that investment is required to raise public awareness and that budgetary information must be presented in an understandable way (World Bank, 2023: 35–36). Moreover, it discusses how mechanisms of public involvement, such as citizen panels or participatory budgeting, can create a learning environment for participants (World Bank, 2023: 27–28). The need for such learning and capacity-building also emerges in our case studies. For example, the case study on Ukraine's 2017 health financing reforms notes that one impediment to inclusive decision-making was that reformers believed that the public, local academics, and civil society organisations lacked the expertise to engage in productive dialogue about key aspects of the reforms. It concludes that investment in such knowledge would help overcome this barrier (Dzhygyr et al., Reference Dzhygyr, Dale, Voorhoeve, Gopinathan and Maynzyuk2023). The case study from Thailand discusses in detail one example of how such social learning can be facilitated (Viriyathorn et al., Reference Viriyathorn, Sachdev, Suwanwela, Wangbanjongkun, Patcharanarumol and Tangcharoensathien2023). Further discussion of how such educative and capacity-building processes can succeed is provided by the WHO in its report on social participation for UHC (World Health Organization, 2021). We agree with Kinuthia that, to allow the public to make up its own mind, the promotion of such civic learning and strengthening of civil society's capabilities should not be left solely to governments.
Kinuthia's second criticism is that that the Report lacks a discussion of the role of facilitators in discussion and how decisions are to be made when deliberations do not reach consensus.
In reply: The Report does discuss the importance of facilitation (World Bank, 2023: 27 and 36). Kinuthia's comments have, however, made us realise that it would be useful to supplement the Report with an account of what makes for good facilitation. While the matter requires further thought, a promising account, due to Afsoun Afsahi, focuses on helping participants develop the attitudes and skills that constitute ‘deliberative capital’, including civility, open-mindedness, assurance of others' willingness to contribute, as well as the ability to analyse others' arguments and find points of agreement as well as dissensus (Afsahi, Reference Afsahi2022).
We also agree with Kinuthia that deliberation cannot be assumed to lead to consensus and that fair procedures should involve clear rules on how decisions are made in the face of whatever disagreement remains after deliberation (see also Baker et al., Reference Baker, Mason, McHugh and Donaldson2021). We admit that the Report is silent on which rules might be used (e.g., decision-making by consensus where available and then by majority voting on areas of remaining disagreement, along with a publication of reasons for both the majority decision and minority dissent). Our sense is that the right approach will depend on context. Again, we recognise that it would be valuable to supplement the Report's discussion with an account of possible rules and their impacts and would welcome research into this.
Kinuthia's third point is that decision processes in health are part of overall budgeting decisions, and hence, the articulated procedural fairness principles and criteria should apply to the entire public financial management system of a country and not merely to health financing. He believes that the Report fails to appreciate this.
In reply: Contrary to Kinuthia's interpretation, the Report does not assume that its fairness framework applies only to decisions in the health sector. Indeed, it explicitly states that the process around decisions on taxes and transfers is to be evaluated using its framework (World Bank, 2023: 18). One example the Report provides is a decision to increase a wealth tax in Norway; another is Tanzania's electricity subsidies. It also discusses the decision whether to allocate resources to health or other sectors. Furthermore, it highlights the International Budget Partnership's Open Budget Survey and Public Expenditure and Financial Accountability assessments, which examine procedural aspects of a country's public finances (World Bank, 2023: 36–37). Lastly, in articulating its conception of fair decision-making, the Report draws on a review of many disciplines and fields of application, including areas of budgeting unrelated to health (Dale et al., Reference Dale, Peacocke, Movik, Voorhoeve, Ottersen, Kurowski, Evans, Norheim and Gopinathan2023). One of the ways it aims to improve on established frameworks for procedural fairness in health, such as Accountability for Reasonableness (Daniels and Sabin, Reference Daniels and Sabin2008), is precisely that it examines revenue raising, pooling, and spending decisions in health as part of the overall budget cycle, in just the way Kinuthia proposes.
3. The importance of voice
Rajan and Rouffy-Ly's first point is that the omission of the term ‘accountability’ in the framework is a missed opportunity to establish a strong connection between accountability and procedural fairness (Rajan and Rouffy-Ly, Reference Rajan and Rouffy-Ly2025). A consequence of this omission, they contend, is that the Report fails to highlight the ways in which its proposed procedural fairness framework is more valuable than traditional accountability frameworks.
In reply: In our view, in global health, the term ‘accountability’ is overused to the point of losing clarity. It has multiple analyses, including interpretations that are unconnected to citizen engagement. For example, in one paper, it is understood as mainly a ‘financial term’ (Robinson and Adams, Reference Robinson and Adams2022: 9) concerned with monitoring budgets and how money is spent in the health care system, rather than with the dialogical exchange of ideas, justification of positions, and respectful engagement with differing views that the Report aims to promote. The Report's authors therefore made a deliberate choice to refrain from using the term ‘accountability’ and instead emphasise two core elements of accountability: reason-giving (which is akin to the commonly used term ‘answerability’) and enforcement. This choice was explained in the scoping review that forms the basis for the Report's framework (Dale et al., Reference Dale, Peacocke, Movik, Voorhoeve, Ottersen, Kurowski, Evans, Norheim and Gopinathan2023: i17), but we acknowledge that it would have been useful to also clarify it in the Report.
Second, Rajan and Rouffy-Ly advocate regarding the Report's ‘voice’ domain (which encompasses the criteria of inclusiveness and participation) as the ‘linchpin’ of procedural fairness. Their motivation is that, to ensure fairness in health financing processes, power imbalances among stakeholders must be addressed. They argue that meaningful engagement with people, communities, and civil society is central to achieving such equity in influence. In their view, the Report's other domains (‘information’ – encompassing reason-giving, transparency and accuracy, and completeness of information – and ‘oversight’ – encompassing revisability and enforcement) then serve as prerequisites for such rebalancing of power in decision-making.
In reply: We agree (and the Report recognises) that procedural fairness involves rebalancing influence and power within decision processes. For example, better representation can advance the interests of marginalised groups and oversight mechanisms can hold decision-makers accountable and mitigate imbalances in the ability to exercise influence. These measures can also enhance substantive equity, because they moderate the inequality in consideration and forms of partiality that are common sources of unjust outcomes (World Bank, 2023: 13).
However, if Rajan and Rouffy-Ly intend to suggest that ‘voice’ is generally to be prioritised, with ‘information’ and ‘oversight’ primarily serving supporting roles, then we disagree. We see no such general hierarchy among the three domains. Although in some conditions, prioritising one domain may be justified, the prioritised domain need not always be ‘voice’.
One consideration in this regard is the degree to which the decisions in question are ‘directional’ or ‘technical’ (World Bank, 2023: 28–30). Directional decisions establish the value orientation of health system financing, e.g., by determining the extent to which solidarity should guide contributions to a publicly financed insurance scheme. In directional decisions, public participation is important for reasons outlined in our reply to Hausman. In contrast, technical decisions: (i) require the application of expert knowledge that lay people cannot be expected to acquire; and (ii) do not themselves determine the aims, principles, and values of (the relevant part of) the health financing system but are instead in the service of pre-specified aims and/or guided by pre-specified principles and values. In such instances, the Report posits, it can make sense to delegate decisions entirely to experts, with minimal or no public participation. The Report cites the example of the National Health Service of Ukraine, which is a body with autonomy over technical and operational decisions, including specifying services under the Programme of Medical Guarantees, selecting providers, and developing payment methods and rates (World Bank, 2023: 29).
Still, even for technical decisions, non-dialogical (unidirectional) public reason-giving is required for a fair process, because the legitimacy of technical bodies depends on the quality of their public reasoning and the public's acceptance of their justifications (Eriksen, Reference Eriksen2021). One important component of such justifications is the aims, values, and principles that have been set for these bodies. What these should be is a directional decision, and so there is reason to have these parameters for expert decision set through a participatory process.
There is, of course, a risk that decision-makers label certain decisions as ‘technical’ to avoid the implementation of mechanisms for public participation, even when the nature of the decisions calls for such participation. Moreover, many decisions that have appeared to people as purely technical in fact contain important value-driven, or directional, elements. This may be increasingly recognised; there is a trend in policy-making to employ methods for public participation, such as citizen panels, to inform decisions in health, social, and environmental policy that were previously dominated by technical experts and government officials (Mitton et al., Reference Mitton, Smith, Peacock, Evoy and Abelson2009; Street et al., Reference Street, Duszynski, Krawczyk and Braunack-Maye2014; Dryzek et al., Reference Dryzek, Bächtiger, Chambers, Cohen, Druckman, Felicetti, Fishkin, Farrell, Fung, Gutmann, Landemore, Mansbridge, Marien, Neblo, Niemeyer, Setälä, Slothuus, Suiter, Thompson and Warren2019; Alemanno, Reference Alemanno2022).
However, even when voice is important because decisions are directional, there may be reasons to improve other domains of procedural fairness first. Meaningful public engagement is demanding in terms of money, institutional capability, and time. In low-resource settings, particularly where democratic institutions are maturing, it is important to have a realistic assessment of the degree to which effective improvements in voice are feasible, at least in the short term. This is illustrated by a case study from The Gambia that informed the Report. Following the country's democratic transition starting in 2016, the process to enact the National Health Insurance scheme (which was passed in 2021) incorporated laudable ambitions for stakeholder participation. While the process ended up being more open than was common before the country's democratic transition, limited resources, time, and inadequate administrative capacity proved to be barriers to consultation with a sufficiently wide range of stakeholders. Consultation therefore fell short of aspirations (Njie et al., Reference Njie, Dale and Gopinathan2023). While, under such circumstances, it remains important to improve voice, it is possible that improvements in other domains, such as the information domain (encompassing transparency, reason-giving, and accuracy and completeness of information) may be worth focusing on first, because these may be more readily feasible and would make a substantial contribution to enhancing procedural fairness. For example, in the study period, The Gambia showed a significant improvement in budgeting and fiscal transparency, primarily due to increased public access to budget information and decisions (International Budget Partnership, Reference International Budget Partnership2021).
4. Procedural fairness in development assistance for health
Bennett and Merritt's application of the Report's framework to decision processes in development assistance for health (Bennett and Merritt, Reference Bennett and Merritt2025) represents a departure from the Report's focus on the national level. Even where the Report suggests that development partners might use the proposed framework ‘to examine their own processes’, the examples it offers involve supporting recipient countries in improving their own institutions and capacity to satisfy the procedural fairness criteria that apply to them (World Bank, 2023: 11 and 39). Bennett and Merritt's analysis of the applicability of the Report's principles of equality, impartiality, and consistency across time to development partners' own decision-making – with application to PEPFAR – therefore involves an exciting extension of the Report's ideas.
Bennett and Merritt first ask whether fairness principles for public decision-making within a country also apply to the decision processes of internationally operating development partners. There is certainly room for doubt that the answer is affirmative. After all, there are many differences between the two types of actors. National health financing decisions by state agencies typically concern the use of resources that come from its citizens (such as tax income) and involve the use of the state's coercive power (such as threatening penalties for non-payment of tax or mandatory insurance contributions). They also typically purport to be done in the name of its citizens and in service of their interests. It follows that there is a clear core constituency – the citizenry – who are owed a justification for decisions and who have a claim to an equal opportunity to participate in decision-making. Consider, in contrast, the situation of development partners that are government agencies, such as the US Department of State's Bureau of Global Health Security and Diplomacy, which runs PEPFAR. The people who supply its resources through their taxes and in whose name the agency operates are predominantly US citizens, while the people whose interests these decisions purport to advance are primarily those affected by, or at risk of developing, HIV in countries that face substantial challenges in addressing the HIV/AIDS pandemic.
Bennett and Merritt argue that these differences do not stand in the way of applying the Report's fundamental principles to decision-making by a development partner. Their first reason is that development partners should be concerned not merely with supplying health-related benefits to recipients, but also with empowering them by taking seriously their perspectives on the ways they might be affected. Their second reason is that development partners often work with or through country governments, so that the way they make decisions impacts the way recipient country governments are perceived by their populations, thereby affecting their descriptive legitimacy. One might add that where development partners work closely with recipient governments, the latter's normative legitimacy may be at stake when development partners' influence is exercised in ways that fail to meet locals' claims for participation in how their state bodies make decisions. A case study that informed the Report provides evidence of the latter idea. In Tanzania in 2017–2018, the Ministry of Health established a technical working group on reforms of its Community Health Insurance Schemes that included representatives of ministries, civil society organisations, development partners, and some of the private organisations supporting the implementation of community health fund schemes. However, beneficiaries and health service providers were not included, and there is some evidence that this lack of inclusion created a lack of trust in the scheme (Binyaruka et al., Reference Binyaruka, Mtei, Maiba, Gopinathan and Dale2023).
In reply: We agree that respect for the agency of recipients is a reason to ensure that those directly involved in and impacted by the development partner's actions have access to information about and rationales for its decisions. It is also a reason for their views (or the views of their representatives) to be sought out and considered. We further agree that where development partners work with governments, they should aim to contribute to more open and inclusive local decision-making in the areas in which they operate. However, we are not yet convinced that these reasons warrant the application of the full principle of equality as specified by the Report, which calls for ‘equal access to information, equal capacity to express one's views, and equal opportunity to influence decisions’ (World Bank, 2023: 10). After all, there are many cases in which, out of respect for individuals' agency, decision-makers must offer them some access to information about and rationales for decisions that affect them, as well as some voice, without being required to offer equal access and voice. An everyday example is the role of students in university decision-making. Students are owed information about and an explanation for policies that affect them, as well as fora in which their views are heard and respectfully engaged with. But this does not imply that their influence needs to be fully equal to that of other stakeholders, such as permanent staff. Bennett and Merritt's argument for the applicability of the equality criterion to development partners' decisions therefore strikes us as incomplete.
This point need not threaten Bennett and Merritt's criticism of PEPFAR's prioritisation policy. For their argument need not rest on a violation of the principle of equality. Instead, it can rely on the observation that this policy was developed with too little input by recipient country governments and/or other genuine representatives of affected populations, despite the substantial implications of these decisions for their access to life-saving services. Bennett and Merritt's conclusion that existing processes ‘[do] not adequately reflect the voices of in-country stakeholders who ultimately may be most affected by these policies’ can stand, since what is ‘adequate’ might merely be substantial, rather than fully equal, voice.
The question whether the Report's principles of impartiality and consistency can be straightforwardly extended to development partners requires further thought for similar reasons.
Notwithstanding these doubts, we hope that Bennett and Merritt's contribution sparks a discussion about the fairness of decision processes that shape global health financing. Many of the Report's principles and criteria for procedural fairness align with proposals for more inclusive and equitable approaches to determining development assistance to low- and middle-income countries, such as the Lusaka Agenda on the Future of Global Health Initiatives (Mwangangi and Røttingen, Reference Mwangangi and Røttingen2023). Bennett and Merritt's paper has inspired us to consider in which form the framework might be applied to evaluate processes in global health financing, such as the Global Fund's country coordination mechanisms (Sands, Reference Sands2019) and GAVI's vaccine funding allocation processes (Gandhi, Reference Gandhi2015; Nunes et al., Reference Nunes, McKee and Howard2024).
In closing, we note that a fundamental premise of the Report is that critical and open-minded exchange of ideas between people with different perspectives can improve our thinking about health financing. We are grateful to our critics for supplying evidence for the truth of this premise through their engaging contributions.
Acknowledgements
This work was presented at the LSE. We are grateful to Shira Ahissar, Suzanne Bloks, Richard Bradley, Rocco Friebel, John Kinuthia, and Kai Spiekermann for comments.
Financial support
Alex Voorhoeve's work in this paper was supported by the ‘Deliberation and Cohesion’ project, part of the ‘Cohesive Capitalism’ programme at the London School of Economics and Political Science funded by the Open Society Foundations.
Competing interests
None.