I have known Kevin and his work for five years, and I am glad that so many great people have joined in this reading group. Special thanks go to Simon May for supporting our discussion. I will here keep my summary remarks to just the most central issues of the chapter for, despite the many things Kevin and I agree about, I have a number of issues to raise.
Chapter 6 draws out the implications of convergence liberalism in two areas. First, Kevin is concerned to bring out the implications for the political production of law, including issues of public advocacy for policies and judicial interpretation of laws. Second, Kevin considers the issue of legal accommodation, illustrating the convergence analysis of accommodation with illustrations drawn from case law regarding the free exercise of religion.
Considering the crafting of law, Kevin defends a general principle of convergent restrain. After some important refinement (181-8), we get the Principle of Convergence Restraint (PCR): “A should not publicly advocate law L in order to contribute to M’s becoming or remaining law (where L may be equivalent to M) if A justifiably believes (a) that members of the public lack sufficient reason Rn to endorse M and (b) that A’s public advocacy effectively contributes to M’s becoming law.” (188)
The locution about L and M is there to accommodate strategic maneuvering and the simple fact that citizens often have to advocate a very rough policy though they really want to more nuanced and refined one to be implemented. What this principle essentially does is prevent people from bringing about policies they believe will not be justified. The fundamental issue, for Kevin, is not their advocacy, which may be futile, but the bringing about of unjustified coercion. Not all advocacy by citizens is equally efficacious in bringing about laws, so Kevin argues that this general principles has different implications when applied to at least three distinct roles a citizen may be in.
The first role is that of the general citizen, who is just one of millions and whose vote is likely to never change the outcome of even a single election. Kevin holds that public reason gives us no sufficient reason to restrict the general deliberations or even the votes of typical citizens.
Second, Kevin considers the role of legislators. He finds that their dialogue is much like that of the typical citizen in that it is generally lost to history and has negligible impact on the law. In contrast to the typical citizen’s vote, however, the legislator’s vote has tremendous power. A legislator may be only one of hundreds, and vote many times a year, and thus gains a significant chance of actually affecting policy through her vote. Convergence liberalism thus supports Convergence Restraint for Legislators (PCRL): “A legislator should not vote for law L in order to contribute to M’s becoming or remaining law (where L may be equivalent to M) if he justifiably believes that members of the public lack sufficient reason Rn to endorse M” (191)
Those in the role of judge have the greatest impact on the law. There are but few judges, and they both interpret the laws and may even determine the constitutionality of the laws. What most distinguishes the judges from legislators, however, is that the reasons judges provide for their decisions often play a crucial role in the way future cases are determined. Through this, then, the reasons judges present significantly affect the laws, or interpretations thereof, that will be implemented in the future (194-5). The reasons themselves, Kevin argues, must also be relevant to convergence liberalism. He thus defends Convergent Restraint for Judges: “Judges should not issue decisions concerning the constitutionality of coercive laws that appeal to reasons they justifiably believe are not shareable (or accessible) for members of the public.” (195)
Let us now turn to legal accommodation. Kevin focuses on cases of accommodation, or exemption from requirements to comply with certain laws, involving religious minorities like the Amish or members of the Native American Church. The religious nature of the objections, however, should not detain us, for Kevin argues that non-religious objections should be treated likewise (217-9). The key then is that some citizens have defeaters for laws that could be addressed with accommodations, and we see many real legal cases that can be considered through the convergence liberal lens. The structure of the argument for accommodation is roughly that for some law, L, a minority may have a relevant defeating reason such as that it imposes upon their core religious practices without providing sufficient compensatory benefit to warrant their endorsement over “no law.” L is not justified, but there may be a simple modification of L, L*, that exempts the objecting minority (either in a targeted or neutral way, with the latter including an option to opt out open to all citizens). L* may be justified, because it provides sufficient benefit to those citizens it applies to even if its scope does not include all citizens. For instance, the exclusion of the Amish from social security does not seem to significantly affect the eligibility of the social security program for non-Amish, even if they would like the policy to include the Amish as well. I would treat Kevin’s discussion of granting an exemption as equivalent to replacing the broad-scoped law with a relevantly similar narrow-scoped law.
To illustrate the reasoning and implications of convergence reasoning in the area of accommodation, Kevin considers three important U.S. Supreme Court cases. To be clear, Kevin’s discussion is not about the decision of the Court, for the Court was not attempting to implement convergence liberalism and is under various institutional constraints. Kevin’s aim, then, is not to assess the court decisions or legal reasoning, but to consider these as real cases for which we have considerable relevant information and can see the reasoning of the individuals involved. Moreover, the legal process has likely lead to a refinement of the reasoning of those involved, and thus serves as an indication of what the idealized reasoning of similar citizens would likely come to.
Kevin has some discussion that he claims is “developing Intelligible Exclusion-based interpretations of the two principles of religious exclusion in the U.S. Constitution, the Free Exercise and Establishment clauses[, and applying] the resulting principles to three prominent legal cases concerned with religious accommodation.” (182) He also indicates that he sees these clauses as cases of “principles of exclusion embedded in the American constitution…[,]” and that they “specify how state coercion satisfies” public justification requirements (197). And yet further he discusses a “publicly justified establishment clause….” (203) With these diverse things being said, I am not confident in my interpretation, but I read the discussion of the constitutional clauses as indicating that they are a useful indicator for potential cases of laws for which citizens have defeaters. More precisely, the fact that someone raises a free exercise complaint indicates a potential defeater for the public justification of the law. The clauses themselves are not, as I read it, really part of the apparatus of public justification, but instead provide a useful heuristic for finding a class of cases. Of course, the fact that some law does not raise issues of free exercise or establishment does not indicate that there are not defeaters. Let us now lay out the cases and Kevin’s findings.
Wisconsin v. Yoder: Wisconin law required children to attend secondary school, but a number of Amish families argued that this burdened them because such schooling would detract from the integration of the children into the Amish community and its religious traditions. According to Kevin, the sincere and well-reasoned (from their prospective) objection of the Amish families shows that a universal compulsory secondary education is not justified. Since such education is justified for the rest of the public (or at least this is assumed in this discussion), and the public would prefer a law accommodating law to no law at all on this issue, the appropriate convergence liberal policy is to have accommodation for the Amish (as the court in fact found).
Employment Division v. Smith: Smith and Black were members of the Native American Church and took peyote in accordance with its rites, but in violation of both the law and the policies of the drug rehab facility at which they worked. As a result, they were fired and, because they were fired for violation of the law, denied unemployment benefits. Kevin argues that “Black and Smith merited exemptions because they possessed the relevant intelligible defeaters for the denial of unemployment benefits because peyote serves a sacramental role in their church.” (210)
Mozert v. Hawkins: In this case, a group of families objected to a text book used in a public sixth-grade class that included many readings the families claimed are hostile to their faith. Here, Kevin finds that “a convergence approach supports the Mozert parents on the grounds that they have intelligible reason to reject the subjection of their children to ideas they sincerely believe are dangerous.” (213)
1) On Judicial Restraint
Kevin argues that most people should be unrestricted in the reasons they present for laws, and he certainly gets a lot of resistance from people about that permissiveness (including from participants in this reading group). It will thus seem eccentric, but I will here raise a worry for Kevin’s claim that judges must be restricted to shareable reasons in their official decisions. That is, I think Kevin has actually been too restrictive.
Kevin’s main point is that the reasons judges give in their cases affect the future coercive laws in that those reasons may affect the future interpretation of laws. Kevin writes as if the reasons themselves must be shared when it is always the end laws that must be justified. As I am sure Kevin knows, the use of shared reasons is neither necessary nor sufficient for securing convergently-justified laws. Suppose Justice Jon makes a decision appealing only to some presumed shareable reasons like equality and finds for, say, a very strict interpretation of a law prohibiting gender-based discrimination in employment. Though the case Justice Jon considered involved a major publicly traded corporation, but other justices will be considering cases involving the Catholic Church’s selection of priests. The judges in those future cases may well take Justice Jon’s decision and his appeals to equality as reasons to interpret the law as requiring that Catholics not discriminate in their priestly appointments (which I assume Catholics have sufficient reason to reject, even if they too value equality). Moreover, Justice Jon might foresee this result, so use of shared reasons seems at best to be an imperfect strategy for promoting justified laws.
Consider the reverse case. Judge Judy could appeal to many values that are not shareable without this producing any unjustified laws. Perhaps the reasons she appeals to regard longstanding traditions that not all citizens accept, but that turn out to be fairly innocuous in their legal interpretations. In particular, in the midst of other considerations, these unshared reasons may only be tipping the scales within the eligible set, as may happen if the reason seems primarily relevant to certain kinds of property disputes. While some citizens may not like that that reason has entered into the determination of the laws, they may not have sufficient reason to reject the laws themselves. In another way, the reasons could, while not shared, be essentially pretty good proxies for the concerns of others. Lastly, I will point out that at least some unshared reasons might systematically not lead to unjustified legal laws, such as unshared reasons against coercion or laws generally – perhaps appealing to the general concerns about the potential threat of government overreach and concentrations of power to strike down laws, without tending to generate many new ones. So, the restriction Kevin promotes seems to go too far in that it restricts the appeal to many unshared reasons that we should not think will tend to produce unjustified laws.
It seems to me that the best Kevin could get is that judges should not appeal to reasons in their decisions if they justifiably believe that such appeal will lead to unjustified laws in the future. Such a restriction, it seems, would not systematically call for or restrict shareable or unshared reasons.
2) On Legislators
Kevin holds that citizen A should not advocate to bring into office politician Q if A believes that “will contribute no less than the alternatives to the imposition of proposals” lacking public justification. Note that the citizen is allowed to make tradeoffs here, supporting Q even though she believes Q will impose some unjustified laws, but this is permissible because A believes Q will do less than alternatives. There are multiple ways these tradeoffs can be made, such as merely counting the number of laws, or considering the number of people who have defeaters, or even somehow trying to weigh the magnitude of the defeaters, but let me set that aside. I think the tradeoffs of the voters point to the need to allow tradeoffs for politicians. Suppose politician Q knows that if she does not vote, perhaps decisively, for unjustified yet widely supported law L, then she will lose in the next election to her opponent who will implement many more unjustified laws. It seems that Kevin’s account gives us reason to say that Q should vote for L (and thus prevent the many other unjustified laws), but this goes against Kevin’s Convergent Restraint for Legislators (191). I wonder if Kevin has reason to think politicians must be more uncompromising than voters or if he would amend his principle to allow these tradeoffs.
In a related vein, Kevin argues that legislators should not vote for policies they believe are unjustified, but should not considerations of efficacy make at least some of these votes permissible? After all, Kevin thinks that the votes of the general citizenry should not be constrained because such votes, considered individually, should not be expected to be efficacious. At least sometimes, though, legislators know that they will not be efficacious in their votes either. Consider, in parallel to a case Kevin gives, the last legislator to vote on a bill that has already received enough votes to fail. Can’t the legislator vote yes on such a bill, perhaps for strategic or symbolic reasons? What objection could Kevin have to, say, Bernie Sanders introducing and voting for a bill, presumably unjustified if passed, to implement socialism knowing that he is the only avowed socialist in Congress and that the bill is doomed to fail?
3) On Accommodation Generally
I think it will help to clarify Kevin’s claims about accommodation. The first point is that nearly every citizen could find laws that they find burdensome and for which they would prefer some sort of exemption for themselves, members of their group, or people engaged in relevantly similar activities, but most of these cases do not present a serious challenge for the un-accommodating law remains eligible for all. The case for accommodation, at least for accommodation seeming necessary for public justification, depends upon the law being defeated if it has a universal scope. For each of these cases, then, what those seeking accommodation must argue is simply that they have relevant defeaters for the law, and demonstrating that there are such defeaters provides reason to eliminate the law. Kevin’s key insight at this point is that those favoring the law may, in effect, create a new law with the same requirements but a more limited scope such that those with the defeaters are not under the law. I take it that Kevin holds that the law need only be justified to those who are under the law, so the law does not need to be justified to those who are exempt from compliance. As long as the accommodating law is eligible for those who would be under it, we can see reason to pass the accommodating law in place of the universally applied law rather than simply have no law at all.
Note, then, that though Kevin writes as if the accommodated citizens rank the accommodating law above no law (205), I think Kevin really holds that accommodating laws need only be justified to those under the law. For instance, a law applying to all those within a city need only be justified to those within the city, and not the larger region. There is nothing I can see in Kevin’s account that requires “accommodation” only for minorities, though he does think that the typical case does involve a small group. Consider a case, though, like the U. S. requirement of male citizens to register for selective service, and previous laws to conscript males for military service. I think Kevin has to hold that such enslavement laws need only be justified to males, or males of the relevant age group or otherwise eligible. One may likewise think that some laws regulating abortion need only be justified to women (or some subset thereof?) as the de facto (even if not stated in the law) targets of the law, and there is the potential to pass other laws that regard only certain religious, cultural, or other communities so long as those laws are justified to all members of the relevant community. An example of the latter case is the requirements in German law taxing registered members of certain faiths to financially support the institutions of that faith. Targeted laws seems structurally equivalent to accommodating laws, so I expect Kevin will say the same about them.
4) On the Cases
When discussing particular cases, it is absolutely essential to be clear about what law exactly is in question. Sometimes we have some wiggle room because the objectors would reject a wide range of plausible laws in question, but this is not always the case. Take first an easy case like the Yoder family’s objection to compulsory attendance to secondary school. There are various ways we might think the laws disaggregate. We could think that there are two laws, one of which requires attendance to primary school, the other of which requires attendance to secondary school, with the Yoders rejecting only the second law. On the other hand, we could further disaggregate, thinking of each grade having its own relevant law, and the Yoders thus raising the same basic defeater to a number of laws. The importance case is that the law might actually be broader, as by being a single law requiring school attendance inclusive of the primary and secondary levels. If this last law is taken as the relevant law in question, then it is not enough to show that the Yoders have an objection to secondary school requirements. The Yoders, as Kevin reports, did not object to primary education, and it seems they saw real benefit to primary education. For them to have a sufficient objection to the general compulsory education requirement, then, it must be that they have sufficient reason to defeat even the requirement of primary education. As it happens, the Yoders probably do have a sufficient defeater for that law too, given the willingness of Amish communities to educate their own members and general lack of interest in imposing upon outsiders. It is not at all obvious, however, that other potential objectors would be able to defeat the more general law. In particular, here, we should consider the Mozert family.
As Kevin, sometimes gaining support from John Tomasi, reports, “the Mozert parents’ complaints were very limited” and they did not object to public education, the teaching of critical thinking skills, or the like (212). The only thing they rejected was the use of one particular text used in the schools, though they did not object to any particular part of the text outside of a supposed repetitiveness of a message they rejected. It here seems to matter greatly what law is in question. Suppose the law is that requiring public education at all levels, considered as a single law. Though I believe the Yoders would defeat that law, it does not seem that the Mozerts would bring down all of public education over one book’s repetitions. They may have a complaint, from their own perspectives, but as I noted above we probably all have complaints about some law or other. Moreover, I doubt that the Mozerts even have suffient defeaters for the sixth-grade – would they really rather there be no public sixth-grade then their daughter have to read a bit more secular rationalism than they like. (Again, note that they did not object to their children being taught some of the material, just to the amount of it in the book in question.) So, it seems to me that Kevin’s support for the Mozerts depends upon thinking that the law in question truly is narrow. Kevin seems to even treat the question as really about the very book, as if the Mozerts could have a defeater for use of the book. While I am not sure exactly how we should specify the relevant law, it seems totally implausible to me that it really is a law about that one book. So it would really help me if Kevin could clarify what law, specifically, he takes to be in question in that case. If it is a law broader than the book, I think some more defense is necessary to show that the Mozerts have a sufficient defeater; if it is just the book, I think Kevin needs to say much more about why we should go that narrow and what the implications will be of trying to think of other laws in similarly narrow terms.
I will note that I find the Smith case even more perplexing. Smith and Black were denied unemployment because they were fired from their jobs for illegal activity. If they have defeaters for the drug prohibiting laws, then we might think that they should receive unemployment since they at least did not violate a justified law. But Kevin states that he assumes the law is justified, which has to mean justified to Smith and Black themselves. If so, then I simply cannot discern what law they have a defeater for. Is it the law that prevents people from receiving unemployment when fired for a crime? That would be rather odd, and seem to entail that they think people should receive unemployment after being caught stealing from their companies, engaging in workplace harassment, or even (to be parallel to the Smith case) that a child molester fired from working at a day care should receive compensation. Perhaps that is the relevant law, and perhaps Smith and Black really would reject it, but if so it does not seem to have much to do with their religious use of peyote. The point is, in any case, that I cannot really discern from Kevin’s discussion what law, precisely, Smith and Black reject and the grounds for it. I hope it is not that they reject to the very narrow case of them personally being denied benefits, but it not that and not drug prohibition, then what could it be?