Ideal Discussants, Real Food: Questioning the Applicability of Public Reason Approach in Healthy Eating Policies

Abstract

Healthy eating policies have become a hot and thorny domain of public concern because they affect people’s liberties, life prospects, and public expenditures. However, what policies state institutions may legitimately enforce is a controversial matter. Is state paternalism for the sake of public health permissible? Could people be incentivized to eat in a healthier manner? Barnhill and Bonotti’s recent book (Healthy Eating Policy and Political Philosophy) tackle these issues (and others) in a manner that seeks to combine the liberal values of state neutrality and antipaternalism, as well as the effectiveness and legitimacy of food policies. To do so, they rely on the accessibility model of public reason. Although Barnhill and Bonotti’s proposal fills an important gap in the field and the accessibility model of public reason overcomes some strictures of the Rawlsian account, their account of public reason faces some practical challenges. Indeed, the institutionalization of their framework seems to need the figure of a moderator of a deliberative panel. However, this figure would create a tension between the public reason framework and the common requirements of deliberative accounts.

Introduction

There are at least two possible strategies for applying public reason theories. One strategy may choose a specific issue in which extant disagreements about a certain issue may be solved if we could apply the principles of public reason. For instance, the issue may be religious diversity in the background culture (March 2009) or disagreement about the treatment of animals (Zuolo 2020). An alternative strategy may envisage an institutionalized procedure that could embody the spirit, if not the principles, of public reason. Such a procedure is usually thought to be a deliberative model of discussion and deliberation. The most important merit of Barnhill and Bonotti’s book is that it is an original contribution in both respects because it provides a novel application of public reason to the thorny domain of food policies, as well as a proposal concerning how the principles of public reason could be transformed into an ethics toolkit for both policymakers and deliberative bodies.

Healthy eating policies have become a thorny domain of public concern, particularly in Western countries. Although healthy eating policies may have a diverse set of meanings, they also have a common source of normative worry: the way people eat is a matter of public concern. Although people’s dietary habits are and should be a matter of free individual choice, they are also a matter of public interest. To be sure, the fact that individual dietary habits are a matter of public concern does not, per se, entail that they should be addressed by institutional intervention. However, this fact at least means that individual dietary habits could be discussed at the public level. If so, this issue may be analysed by a normative framework inspired by public reason, even though it might not necessarily be the case that it should be governed by public institutions. The need for this discussion stems from the fact that we should not understand this domain as including the perfect conditions for exclusively free private choice. First, this is because people’s possible choices are frequently conditioned and constrained by how food production and distribution shapes our dietary habits (Swinburn et al. 2019). Second, people’s choices concerning food are often less informed and autonomous than people’s standards of free choices would like them to be: matters of time, advertising, lack of viable alternatives and so on make people’s dietary habits less free than they could be (Cohen and Farley 2008). Third, food habits in particular, although this is by no means exclusive, pose epidemic problems insofar as they are the main cause of most widespread diseases in Western societies (WHO 2013). Fourth, the way food is distributed and produced sometimes negatively affects people’s opportunities and/or reinforces already disadvantaged groups (e.g., ethnic minorities, working mothers) (Bleich and Ard 2021).¹ In sum, we should not understand this domain as being purely characterized by people’s choices in an open market that provides alternatives fit for all. Hence, there is space for public intervention in healthy eating issues beyond the mere idea that the state should safeguard the space for the free market and individual choices.

However, what policies state institutions may legitimately enforce is a controversial matter. Controversies range from ethical issues (can state policies intervene in people’s choices concerning food?) to more empirical debates (should public institutions be satisfied with a purely quantitative measure of food acceptability—also called “nutritionism”, Barnhill and Bonotti 2022: 164-7—rather than with the consideration of the social value of food?). Hence, public policies may be controversial for different reasons (Barnhill and Bonotti 2022: Chap. 3–4). First, they may be perceived as paternalistic interventions in people’s choices. However, interventions may be justified not only in terms of people’s presumed best interests (being healthy), but also in terms of the costs that unhealthy lifestyles have on societies. Second, state interventions may be criticized for being perceived as non-neutral (and illiberal) impositions of some values. Healthy eating standards may be perceived as posing a challenge to traditional and religious food habits.²

Barnhill and Bonotti tackle these issues (and others) in a manner that seeks to combine the liberal values of state neutrality and antipaternalism, as well as the effectiveness and legitimacy of food policies. To do so, they rely on the accessibility model of public reason. In Barnhill and Bonotti’s (2022) view, the accessibility model of public reason has several advantages.³ Indeed, the accessibility model avoids the problems of the two other competing accounts (the shareability model and the intelligibility model)⁴, while maintaining a robust commitment to the principles of public reason. The first advantage of the accessibility model is that it is applicable to healthy eating policies because it employs accessible reasons (more open than shareable reasons, but more conclusive than merely intelligible reasons). Moreover, it does not presuppose too overly an idealized account of the participants and their epistemic and moral capacities, and is, thus, compatible with a moderate idealization.

In sum, in Barnhill and Bonotti’s view, the accessibility model provides the framework for thinking about the applicability in the two senses mentioned previously (the application to a specific domain and deliberation). Nevertheless, how the accessibility model of public reason is applied in this framework needs further scrutiny. I will subsequently argue that the transition from the application of public reason in the first sense—namely the application of the accessibility model of public reason to the domain of healthy eating policies—to the transition in the second sense—namely in its application as an institutional framework—is not unproblematic. There are several features in Barnhill and Bonotti’s conception of the deliberative bodies that make the applicability of their account less convincing than it could be.

From Public Reason to Public Deliberation

To understand why, we need to more closely examine the specific form that Barnhill and Bonotti give to the institutionally consultative and deliberative bodies. The institutional application is conceived of in two steps. First, drawing on the principles of the accessibility conception of public reason, the authors outline an ethics tool, namely a set of questions that could and should be used by policymakers and policy advisors to check whether a certain proposed healthy eating policy respects the desiderata of public reason. “This framework consists of questions that public health officials and legislators should ask themselves when designing new healthy eating efforts” (Barnhill and Bonotti 2022: 182). Next, the authors seek to envisage how this framework could be institutionalized “by incorporating into policymaking a process of consultation and deliberation that includes public health officials, ordinary citizens, advocacy groups, and representatives of affected groups” (Barnhill and Bonotti 2022: 182).

Barnhill and Bonotti understand their ethics tool as a principlist approach, following other principlist approaches in the ethics framework. Principlism is understood as the method of admitting that diverse irreducible principles are at stake in most important normative decisions. Accordingly, the role of ethics frameworks is precisely that of balancing their relative importance in each context. The specificity of a principlist approach based on public reason is that it not only weighs different relevant values, but it also filters out “reasons that are grounded in values that are controversial and non-political” (Barnhill and Bonotti 2022: 186-7). In Barnhill and Bonotti’s approach, the procedure of public reason should also include “an epistemic dimension: public justification for healthy eating efforts must be grounded both in a reasonable balance of shared political values and in sound empirical evidence” (Barnhill and Bonotti 2022: 187).

Accordingly, the first way in which Barnhill and Bonotti’s approach is thought to be applicable is by providing an ethics tool to guide the actions of policymakers; in a complementary manner, such a tool may also be used by citizens to assess how a certain policy fares with respect to the requirements of public justification. The ethics tool includes the following main questions, which should be posed when discussing a food policy:

• (1) What are the public health-related aims of the policy? […]

• (2) Does the policy have other aims? […]

• (3) Is the policy likely to have any unintended positive or negative side effects? […]

• (4) Does the policy strike a reasonable balance of political values? […] (Barnhill and Bonotti 2020: 188)⁵

Next, Barnhill and Bonotti envisage the institutionalization of their framework by imagining a consultative and deliberative panel, which could embody the principles of their public justification approach and the ethics tool. Instead of imagining such a deliberative body in an abstract and general manner, they directly run a hypothetical scenario in which they represent a discussion about the admissibility of an actual policy proposal, following the real case of the New York State petition to exclude sugary drinks from SNAP (the Supplemental Nutrition Assistance Program). Participants in this imagined consultation exercise include the following:

• A local public health official who is considering the policy.

• A SNAP participant who supports the policy.

• A SNAP participant who opposes the policy.

• An anti-hunger and anti-poverty advocate who opposes the policy.

• An advocate who supports the policy.

• A fair-minded group of people trying to reach an agreement (Barnhill and Bonotti 2022: 201).

Hence, Barnhill and Bonotti reconstruct a possible discussion of this case in a mildly idealized manner. The first and most obvious, but nonetheless crucial, idealization is the implicit assumption that all participants are in good faith, reasonable and committed to finding a solution. The authors do not make this assumption explicit because it is—as it were—in the logic of the situation, insofar as only people’s reasons, not their persons, are being considered. It is likely inevitable to rely on this assumption in the process of passing from a public reason setting to a deliberative scenario. I do not want to put this assumption into question. Although it would hamper any concrete application of the model, let us assume it is true for argument’s sake.

Moreover, there is a certain level of idealization because the participants are assumed to be perfectly representative of a certain stance, and only the reasons backing a certain position are being considered. However, Barnhill and Bonotti do not assume that the participants are perfectly rational. The kind of idealization assumed in this scenario concerns the relevance of the reasons that are discussed: neither the participants are considered in their subjectivity, nor are they merely bearers of their interests. Rather, they are thought to be the bearers of relevant reasons, and as bearers of relevant reasons they contribute to the discussion which, therefore, aims to reach a shared point of view. However, this mildly idealized picture does not assume that all passages are flawless or conclusive, because even Barnhill and Bonotti admit that we should accept that a discussion might be inconclusive when relating to certain issues.

I do not want to criticize this idealized setting per se. To some extent, it is inevitable to opt for a certain level of idealization when imagining a possible application of an admittedly idealized reasoning such as public justification. What I find less convincing is the way in which the deliberative panel is constructed. Subsequently, I will question the plausibility of the application of the framework in light of some standard features and requirements of deliberative approaches. I will argue that the example of a consultative and deliberative panel could not withstand common requirements of deliberative approaches. This criticism is appropriate even though Barnhill and Bonotti’s theory is not devised as a fully deliberative one. Indeed, the author themselves conceive of their approach as very close to deliberative theories and the whole effort to imagine the application is in deliberative terms.

Deliberation, but how?

Although the authors follow some requirements of deliberative approaches—for instance, they strive to make the example of the panel as representative as possible—it is majorly lacking. What seems to be missing in this setting is a moderator. In deliberative bodies—which, following Brown (2006), we may call the citizens panels—moderators have the following roles: they seek to ensure that all are given an equal chance to speak, that a part does not monopolize the discussion or impose its view on others, that the parts share the common rules, that the argumentative passages are accepted, and in general that any obstacle to discussion and deliberation are removed (Ackerman and Fishkin 2004). Barnhill and Bonotti do not seek to represent real deliberative citizens panels, but the function of their envisaged consultative and deliberative groups is analogous to the role of citizens panels: instructing a policy, checking its acceptability and making it representative.⁶ Moreover, Barnhill and Bonotti explicitly mention deliberative institutions as one of the most important sites of application of their framework. As to the kind of representativeness, they do not opt for a statistical representation but rather for a cross-selection of the types of categories interested. Hence, albeit in a peculiar manner, Barnhill and Bonotti’s framework shares many features of deliberative approaches and, therefore, can be assessed as such.

In light of this, I claim that the lack of a moderator seems to be a flaw in their conception. Such absence is an issue because it is not obvious that the members of the public would accept the specific conclusion of each argument that the authors suggest. Barnhill and Bonotti are well aware of this problem, and they do not seek to hide the fact that real discussions are likely to be characterized by inconclusiveness or stubborn disagreements. To remedy this problem, the authors introduce the figure of a “fair-minded group of people trying to reach agreement”. How should we understand this figure? There are two possible ways to understand it.

In the first interpretation, the “fair-minded group of people” should be understood as the good will of all the participants⁷, namely as the capacity to obtain the best outcome of the deliberation on a certain issue, given the available evidence. In this sense, the fair-minded group of people is a normative figure, namely the personification of truth or reasonability in the debate. Barnhill and Bonotti seem to lend support to this idea when they characterize the fair-minded group of people as those who are “listening to everyone’s point of view, but also trying to reach agreement”. This figure “is meant to replicate the process of deliberation that the group would engage in” (Barnhill and Bonotti 2022: 200‒1). However, this interpretation has a problem. It seems at odds with the presentation of the consultation. As seen above, Barnhill and Bonotti imagine the consultation being comprised of a SNAP participant supporting it, a SNAP participant who opposes it, an anti-poverty advocate opposing the policy, and a SNAP supporter (Barnhill and Bonotti 2022: 201). To these figures Barnhill and Bonotti add the fair-minded group of people. But, in this list, the fair-minded group of people either is different from the other parties, insofar as its members do not have a position, or is part of them. If they do not have a position of their own, they should be understood as something different from the other participants. (We will pick up this point shortly in discussing the second interpretation.) If the fair-minded group of people is not another separate participant, then the fair-minded group of people are the other participants, who are “fair-minded” only when, and to the extent that, they, on a certain issue, are right. If so, it is not clear to me why creating this figure makes sense, given that it only postulates that some might be wrong and some might be right (because they represent the best understanding of the deliberation in a certain moment). In this sense, then, the fair-minded group of people is just the correct application of the procedure, not a figure of its own representing a specific stance.

The second interpretation of the fair-minded group of people understands this figure as the functional equivalent of a moderator because its role is particularly invoked when there is a disagreement. This interpretation is supported by the many references in the book (Barnhill and Bonotti 2022: 203, 204, 206, 207, 209, 212, 2013, 215) where the fair-minded group of people does not seem to represent any specific stance and its role is that of drawing conclusions, when conclusions are possible given the available evidence and arguments, or of tilting the balance in favor of the best solution in case there is a disagreement. But creating this figure amounts to introducing a sort of moderator into the consultation. Indeed, it seems to play the role of moderator without being a separate figure. If so, it introduces a strange and instable form of idealization in an already idealized model. As previously stated, idealization is not wrong per se, for it is rather necessary to think of a possible application of the framework. However, the “fair-minded group of people trying to reach agreement” is idealized both at an epistemic and at an ethical level. Indeed, these people are supposed to perfectly understand the arguments at stake and draw the implications of the discussion. Lacking any specific agenda, they deliberate and decide for the sake of the best. In this sense, they seem close to some fictional figures in some classical ethical theories. They are assumed to have such a function in virtue of their being “fair-minded” without specific powers attached to it.

Why not instead think of a real moderator? The difference might seem negligible but it is not. Unlike the “fair-minded group of people trying to reach agreement”, the moderator is institutionally given a certain role and capacities. There can be some sort of idealization, requiring an actual moderator to act in the proper manner according to the best interpretation of the responsibilities of the role. However, that idealization is incorporated into the role of the moderator, which means that if a specific moderator does not act according to the required criteria, s/he can be criticized or removed. This cannot be possible in the case of a “fair-minded group of people trying to reach agreement”, because this ideal figure is only meant to act in this manner. The difference looks tiny and merely verbal but is not. If we were to actually put in practice a deliberative panel on healthy eating policies by following the example given in the volume, it is not clear how we could recruit the “fair-minded group of people trying to reach agreement”. Randomly? On what basis could we assign them to this role and hope that they will act accordingly? Moreover, it is unclear what we could do if the selected people do not act as expected. Could they be reproached for not being fair-minded or for not being epistemically irreprehensible? And even though it is intended that they be criticized for their failure, on what basis could this be achieved? Insofar as the “fair-minded group of people trying to reach agreement” does not have a specific institutional task, they could hardly be reproachable, and any criticism levelled against their behavior could likewise be directed against other parties for failing to behave irreprehensibly or for failing to live up to a certain epistemic performance.

The authors might protest that these remarks do not constitute a criticism because their framework is positioned at a certain level of idealization, which works as such before the specific instantiation in actual deliberative panels. That is true, but the other features of their framework are thought to work as a normative model directly applicable in specific exercises of deliberation, and that is what the authors want to do.

The figure of the moderator is also necessary in order to make the communicative and deliberative work proceed smoothly. The authors indeed understand this framework as a pure exchange of reasons and points of view. Accordingly, they assume that all the parties express their legitimate but different perspectives faultlessly, understand each other seamlessly and draw the implications of the argument properly. But sometimes they get stuck because there is, for instance, insufficient data to reach a certain specific conclusion or the arguments in favor or against a specific issue are not fully settled.

As discussed, idealization is not a problem per se. The question is the kind of idealization needed in this framework and whether the type of idealization fits the purpose it is set for. How are we to interpret the idealization of the representatives? On the one hand, they are quite idealized in the epistemic sense in order to ease the deliberative framework; however, if fully understood as idealized in the epistemic sense, they would lose their representativeness of different positions, and they would get closer to fair-minded people who just want to reach an agreement but do not resemble real positions. If, on the other hand, such representatives were understood as a shorthand placeholder of people as they are, it would be unlikely that representatives could easily process all the data and implications of the deliberative panel.

Of course, choosing the appropriate level of idealization is always debatable and tricky. But the problem is that this choice is not solved by the role played by fair-minded groups of people trying to reach an agreement, because such a group does not work as an institutionalizing tool, either in a real setting or in an idealized scenario.

Barnhill and Bonotti’s Possible Reply and a Rejoinder

At this point, Barnhill and Bonotti might reply that my critical remarks, if valid, could be solved pretty easily: by introducing in their procedure the figure of a moderator in addition to the “fair-minded group of people trying to reach agreement” or in substitution to them.

However, the solution is not so easy. Or at least I will argue so. Making this change to their deliberative panel would make it closer to real deliberative experiences. But I contend that this would be compatible with Barnhill and Bonotti’s commitment to public reason only under certain conditions. Why do I say so given that previously I have explicitly claimed that there is a close connection between public reason and deliberative theories, one that Barnhill and Bonotti rightly rely on?

To understand this, we need to go back to the basics of public reason and deliberative democracy. At the basis of public reason we may find the overarching principle of liberal legitimacy: laws or institutions are legitimate to the extent that those who are subject to them have good reasons to accept them (or no reasons to reject them). From this principle, public reason theories argue that we should use certain procedures to accept certain kinds of arguments (non-controversial, reasonable ones) and to discuss accordingly, at least in the domains about which the liberal principle of legitimacy should apply. Deliberative theories argue that we can reconcile the democratic principle of people’s participation with an epistemic requirement of good deliberation. Whether this should be made via deliberative polls or mini-publics does not concern us here. What is relevant here is that from deliberative theorists’ point of view, actual deliberation represents people’s views and improves the quality of deliberation. This improvement is achieved thanks to the epistemic virtues of deliberation, which “forces” people to be exposed to different views, reconsider their position, give and accept reasons and so on. Additionally, a moderator is usually thought to be an inevitable ingredient to get closer in practice to these epistemic desiderata.⁸

However, the inclusion of a moderator, as a separate figure and not as a hypothetical group of participants who are right, supports the view that the deliberative procedure is really open-ended and that the procedure is an epistemic practice of looking for the best solution. But the procedure envisaged by Barnhill and Bonotti seems to be more faithful to the spirit of public reason. Accordingly, the procedure is not meant to be a procedure to look for the truth. Although Barnhill and Bonotti’s account of public reason does include robust epistemic desiderata, it is by and large a procedure to filter out what is publicly non-acceptable and what all the parties have reasons to accept. Of course, what is publicly non-acceptable (insofar as it is sectarian or poorly supported) is frequently epistemically poor too. Reasonableness is not necessarily detached from truth, even in a post-Rawlsian account, which does not subscribe to the Rawlsian “squeamishness” about truth.

However, the scope of solutions of discussions informed by public reason principles seems to differ slightly from the scope of solutions that a deliberative body might in principle take. In the former case, indeed, all solutions are to be compliant with the normative principle of public reason, mutual acceptability and scientific facts (at least in the framework of public reason endorsed by Barnhill and Bonotti). In the latter, the conclusions of a deliberative panel are likely to overlap with the outcomes of public reasons. In principle, however, they are not the same. Perhaps being accepted by members of the public and being deliberatively stronger than other democratic decisions is not necessarily coincident with the requirements of being reasonable and acceptable. At the least, we would need further reasons to suppose that this coincidence should actually occur. This is not a specific problem of Barnhill and Bonotti’s framework, and rather points to the dispute between different accounts concerning the relation between public reason and deliberation.

There is a further possibility. We could require that the moderator be instructed by the principles of public reason. In other words, the moderator could be instituted not as a figure facilitating the epistemic elaboration of the discussion with a view to reaching the best solution, whatever it is, but rather as a figure whose purpose is that of keeping the discussion within the premises of public reason.⁹ In this case, the possible tension between public reason and deliberative consultation would dissolve. However, this solution would come at the cost of transforming the possible activity of facilitation and improvement of the discussion into a policing of the discussion. If so, public reason, instead of being a set of principles that could inspire fair dialogic relations, would become a set of rules to be enforced accompanied by sanctions for any lack of compliance. I assume that this would not be a desirable option for Barnhill and Bonotti.

Perhaps it is overdemanding to require that Barnhill and Bonotti’s book solve the complex relation between public reason and deliberative democracy. In a more constructive manner, the troubles in interpreting the function of the fair-minded group of people and in viewing the institutionalization of their framework as a possibly open-ended procedure could be understood as a challenge, rather than as an objection. Accordingly, we could ask under what conditions real deliberation could coincide with public reason. It seems plausible to assume that this could be so only to the extent that the parties and the moderator have certain epistemic capacities (knowledge of the relevant empirical facts) as well as a real commitment to finding the best solution. That these conditions actually occur is certainly not impossible. However, these conditions seem significantly demanding and, again, show the level of idealization required for this framework to be applicable. These remarks should not be understood as a fault of Barnhill and Bonotti’s approach. However, they show the difficulty of conceiving the applicability of their framework.