Journal of Practical Ethics

In the political philosophy of the present age, there is nearly universal agreement that democratic government should separate church and state—specifically, governmental and religious institutions. This paper sketches a framework for that separation widely acceptable by international standards. The paper will not argue for it beyond pointing to grounds of a kind that are commonly respected by writers in political philosophy. With the framework in view, the paper indicates why maintaining church-state separation tends to create difficulty in balancing two major democratic ideals: preserving equality before the law and protecting liberty, including religious liberty. The challenge is particularly complex where healthcare is either provided or regulated by government. There are at least two problems in balancing the imperatives of equal treatment and protection of liberty: one problem is raised by government’s requiring, for the well-being of the populace, healthcare of a kind that some religions prohibit; another is defining what constitutes healthcare in the first place.

With healthcare legislation by the Obama Administration as a case study, I will consider the issue of contraceptive coverage as a requirement on private employers who provide employee health insurance. I refer to the “preventive services” mandate of the Affordable Healthcare Act passed under the first Obama Administration. 1 The mandate has been challenged in the courts by (among others) Liberty University and the University of Notre Dame and is being widely debated in the U. S. My overall conclusions will bear on both that specific issue and the general question of how to balance considerations of democratic equality and freedom of religion.

The Separation of Church and State: Three Central Elements

On my view, democratic societies should be structured in keeping with three church-state principles as major standards for sound government. These largely rest on the premise that liberty and basic political equality, including one-person, one-vote and equality before the law, are default standards in democracies. Departures from them stand in need of justification, as where religious grounds are the only legal basis for exemptions from military conscription (Audi 2000; Audi 2011a). They no longer are in (for instance) the United States; and in many countries where religion has had a special legal status, that status has gradually diminished (see, e.g., the Supreme Court decision in United States v. Seeger, 380 U.S. 163,1965).

The Liberty Principle

The first standard is the liberty principle: Government should defend “maximal” freedom, including religious freedom (I assume that there is a moral right to such freedom). Determining such maximality is difficult. Here I suggest that we keep in mind something close to Mill’s famous harm principle: “the sole end for which mankind are warranted, individually or collectively, in interfering with the liberty of action of any of their number, is self-protection . . . to prevent harm to others.”2 Arguably, allowing a person to die by preventing or even withholding a transfusion or some other readily available medical treatment is doing a harm; but even if it is not (and is instead, e.g., allowing a harm), other principles proposed in this paper will justify government’s outlawing, as the U. S. Supreme Court has, Jehovah’s Witnesses’ refusing life-saving transfusions for their minor children. Consider, by contrast, sending messages announcing religious services by billboard postings on church property. The harm principle could not in any normal circumstances justify restricting these.

An intermediate case would be the use of loud speakers for public calls to prayer. Do loud announcements like this harm those whose concentration they break? And might it matter whether the call is only weekly or much more frequent? There is no simple answer, but the question does bring out that behavior that is not intrinsically harmful, such as low-volume monthly announcements with content of wide interest in a community, can rise to a harm if magnified or greatly increased in frequency.

The Equality Principle

The second standard is the equality principle: Government should treat different religions equally. If one thinks of churches as institutional citizens, this can be seen as a special case of the democratic commitment to equal treatment of citizens. The establishment clause in the U. S. Constitution, which prohibits government from establishing a church, accords with this principle. Similarly, the principle requires that church bells and religiously employed loud speakers be treated equally, so that, for instance, a level of annoyance created by church bells for citizens who dislike those is regulated comparably with loud speaker calls to Muslim prayer.

Granted, if churches outnumber mosques in a community, there might be more bell ringing than loud speaker calls to prayer. This disparity could nonetheless be an instance of proportionate equality. Unequal treatment is not entailed by unequal representation of a regulated behavior by different constituencies. Equality does not entail uniformity.

The Neutrality Principle

The third church-state standard I propose is the neutrality principle: Government should be neutral with respect to religion. This is not a consequence of the second principle, since equal treatment of different religions is compatible with preference for religious over non-religious institutions or citizens. There are other respects in which governments should be neutral. They should not, for instance, prefer the interests of athletes over those artists. The question at issue is structural preference, the kind built into a constitution, as opposed to legislation, passed by a democratic majority, which differentially benefits a given group. If, to respect majority preference, a city council votes to use limited funds to build a stadium rather than a concert hall—this does not violate the neutrality principle, whereas a constitutional preference of the same kind would.

Similarly, if majority preference leads to prohibiting new construction of tall structures in a certain region, this might interfere with plans to build minarets yet not affect church construction projects (since none need involve building new steeples), whereas a prohibition of the former as such but not the latter as such would be unequal treatment of religious institutions. These cases illustrate differences in treatment of religion that are intrinsic to governmental policy from differences in effect on religious institutions that are contingent on circumstances.3 This difference remains even if, because of such factors as governmental commitment to civil liberties, the de facto level of freedom for citizens, and indeed the level of support for religious institutions in general, are higher than they might be without an established church. In England, for instance, efforts in these directions could be partly motivated by a realization that, in terms of ideals for democracy as opposed to historical continuity, an established church is (other things equal) undesirable.

Protection of the Sense of Identity as a Normative Standard in Democracies

Why should religion be singled out in political philosophy in a way other voluntary commitments, such as artistic ones, are not? One answer concerns the history of certain democratic societies and the importance of religion therein. But there is another consideration, independent of contingencies of time and place. For the sake of the flourishing of citizens, democracies should observe a protection of identity principle: The deeper a set of commitments is in a person, and the closer it comes to determining that person’s sense of identity, the stronger the case for protecting the expression of those commitments.4 This principle is neutral with respect to how particular values and activities play this role for a given citizen. It is also normatively neutral regarding what those values and activities are: the democratic commitment is to the inherent value of protecting people’s freedom to realize their deepest desires, which include their “self-defining” ones; and it does not discriminate among these desires except insofar as protections of liberty (or comparably strong democratic standards) require it.

Although the protection of identity principle is religiously neutral in content, it has special significance for church-state issues. For as a matter of historical fact and perhaps of human psychology as well, religious commitments tend to be important for people in both ways: in depth and in determining their sense of identity. Other kinds of commitments can be comparably deep (in a sense implying both rootedness and a tendency to control a significant segment of behavior); this principle does not discriminate against those—nor does it presuppose any controversial metaphysical view regarding what determines a person’s actual identity. But few if any non-religious kinds of commitments combine the depth and contribution to the sense of identity that go with many—though not all—religious commitments. Patriotism is a good example here; it can run very deep in a person, and the protection of identity principle applies to it. It is an interesting question whether the deepest and most behaviorally controlling forms of patriotism tend to have properties akin to those of religious commitment.

Healthcare Policy as an Issue in Democratic Societies

This is not the place for a theory of the overall role of government in democracies. Here I simply assume something few political philosophers will contest: that a prosperous democracy should seek to guarantee (even if it does not itself provide) a suitable minimum standard of healthcare for citizens.5 In some cases, religious objections to a medical policy or procedure clash with requirements that governments or majorities of citizens or both take to be within the suitable minimum. These are the kind holding special interest for this paper. Let us consider some general points and then proceed to the Obama Administration’s contraceptive ruling.

Certain extremes may be clear, as noted above in relation to the liberty principle. Consider the Jehovah’s Witnesses, whose religion prohibits blood transfusions. The protests against the Obama mandate do not go to the extreme of seeking to exclude coverage for transfusions where they would be refused on religious grounds and might be imposed by force, say to save the lives of children. Here the state might require them (and their reimbursement by healthcare plans) despite religiously based parental protests. But the imperative to protect liberty could be invoked from another perspective—that of employees or students who might feel their liberty is affected by financial hardship. Consider a married female custodian who has three children and very strongly wants no more. If her healthcare plan does not cover contraceptive services, she may have to choose between expenses she cannot afford and abstinence that—perhaps because of factors beyond her control—she cannot achieve. One might argue that democratic governments need not be concerned with such matters or, more plausibly, that imposing the costs in question does not imply a restriction of liberty that government should prevent: surely, someone might argue, the custodian may privately purchase contraceptive services simply by being more economical in buying food or clothing.

The issue here is representative of many in ethics: it is nothing less than what level of cost or suffering renders its imposition a restriction of liberty or, more specifically, a restriction sufficient to justify government’s passing laws to prevent it. When do we lack freedom to do something, and when is doing it merely costly? Moreover, there are degrees of freedom both to act and in acting; and democratic governments may properly seek to support the highest degree of at least the former. Regarding freedom in action, there is no simple uncontroversial way to distinguish free from unfree action or to determine degrees of freedom in action. But, concerning freedom to act (roughly, of action), a good case can be made for a democratic society’s taking reproductive freedom—the freedom to reproduce or not6 —as both important and, especially in the contemporary world, easily undermined. This point applies particularly to women, many of whom are either dependent on men for at least a large portion of their economic resources or largely subject to the will of men, or both.

If we add that some people may think they have a religious obligation of stewardship to limit the size of their families, then religious liberty itself may be argued to be curtailed by excluding contraceptive coverage. Perhaps this much may be concluded here: in a society in which government abides by the liberty principle, where reasonable disagreement may occur regarding what constitutes a restriction of liberty in a given realm (as with contraceptive use), then even if governmental protection of that liberty is not constitutionally required, it may be imposed, in an appropriate way, by majority rule. This is in any case one route to defending the Obama mandate, though we will soon see that there is an alternative policy likely to be favored by religious institutions.

Institutional Resistance to the Mandate on Contraceptive Coverage

Consider the Notre Dame protest, as stated in a letter from the University of Notre Dame’s President, John Jenkins, to Kathleen Sebelius, Obama’s Secretary for Health and Human Services.7 A main point President Jenkins made is that the mandate treats institutions that by policy serve mainly co-religionists differently than those, such as Catholic universities, that do not, since both admissions and faculty appointments are not restricted to Catholics (even if, as a matter of statistical pattern, co-religionists are in the majority). This point is significant; but from the perspective of political philosophy, given the rationale for minimal healthcare standards, the point might be argued to favor extending the mandate to the former institutions rather than exempting the latter. The issue I want to concentrate on is not what exemptions there should be, if any, but how to approach the problem of balancing healthcare policy requirements against conflicting considerations raised by the right to free exercise of religion.

The Possible Bearing of a Principle of Double Effect

Here it may be instructive to consider the principle of double effect, which is, if not an element in much Roman Catholic moral teaching, at least respected by many ethicists writing in the Catholic tradition. I take this to be roughly the principle that if an action (such as adopting a healthcare plan) has two effects, one good and one bad, we may perform it in order to bring about the good effect, provided the bad effect is neither our (intended) means nor our (desired) end in doing the deed, and the good effect is sufficiently good to warrant permitting the bad one. The idea, as applied to the contraceptive mandate, would be that the intention of the Catholic institution is, e.g., to provide good, egalitarian healthcare without discrimination on the basis of religious conviction, and that covering of contraceptives is only a foreseen and regretted collateral consequence of adopting this healthcare plan.8

Even if the principle is both sound and applicable, it is not clear that it solves the problem for Catholic institutions. Granted, covering contraceptives is not a means to adopting the overall healthcare plan. Even granting, too, that covering contraceptives is an effect of, and not an element in, adopting the plan—which might be plausibly argued to be a conjunctive action with that as a component—the argument overlooks what seems presupposed by plausible appeals to double effect: that if there is an appropriate way to produce the good effect without the bad one, then producing the former does not warrant permitting the latter. Consider collateral damage in just war theory: if the only way to defeat the enemy requires bombing that will kill 100 civilians and defeating the enemy is morally important enough to justify the bombing, that is one thing; but suppose one could defeat the enemy by means equally deadly to combatants yet with far less collateral killing. Then the fact that bombing will do it does not warrant the collateral damage.9 In the contraceptive case, government can apparently provide an alternative: direct funding of contraceptive services by, for instance, vouchers. Thus, an institution opposing the mandate could conceive its alternative to including contraceptive coverage in its policy as (by legal action) causing government to realize such an alternative. Minimally, if the principle of double effect is to justify an action with a bad effect, it must be formulated so as to entail that the good the action does cannot be realized with lesser undesirable consequences than bringing about the relevant bad effect.

The matter now becomes more complicated. How is the good effect of adopting and maintaining a healthcare provision to be determined? If, as in the case of the mandate, contraceptive coverage includes more than birth-control devices and drugs that prevent conception, one might have to consider the badness of terminating what many Catholics consider pregnancies as opposed to preventing pregnancies by contraceptives. If the coverage is more limited, this factor may be eliminable. Above I referred to the good of providing “good, egalitarian healthcare without discrimination on the basis of religious conviction” but did not assign any particular value to, e.g., being egalitarian in this way. This value is arguably great, but it cannot be quantified. Nor can we quantify the badness of unwanted pregnancies reasonably taken to be prevented by contraceptive coverage. These are only a sample of points suggesting that the comparative weighting required by the double effect principle will be difficult and may be inconclusive. That is not to suggest one should not attempt the weighting; it is relevant to any plausible moral appraisal of the issue, and even inconclusive weightings reflecting the many values raised by an issue can facilitate understanding of it and increase the probability of a negotiated settlement.

Governmental Funding of Healthcare

The difficulty of arriving at a solution that, even using the double effect principle, is satisfactory to all Roman Catholic institutions should lead us to consider more carefully the alternative of direct governmental funding of contraception. One alternative is a voucher system; another is simply reimbursing healthcare providers such as physicians and pharmacies. There are at least two important points here. Both illustrate that what a principle calls for differs in different circumstances. The first point is that the principle applies differently if government will fund contraception given their arguing for it than if it will not. If it will, then, for Catholic institutions, the good effects of adding contraceptive coverage would apparently not be supported by the principle as supplemented in the way proposed here, where the institutions can cause a policy change that will achieve the good result without the bad. If it will not, then (as where a suitable healthcare plan cannot be provided through government funds), the principle might support incorporating the Obama plan.

The second point here concerns an issue that is too easily ignored: the justification of taxation. Governmental funding of contraception entails that taxpayers cover an expense that—in an employer-centered system—arguably employers should shoulder. Even citizens who agree on the appropriateness of the Obama mandate in the first place might complain of being taxed for such a purpose. Still, protecting religious liberty is something all can agree is important; and here, as elsewhere when citizens are taxed to support things they disapprove of, the complaint is understandable but not decisive. If it were decisive, the freedom protected by democratic governments would include the option to pay either no taxes or pay them selectively. To be sure, a well-functioning democracy is designed to allow public protest before major policies are instituted and to use the ballot box to change governmental priorities.

An alternative to vouchers is for government to exempt religious employers from mandatory inclusion of contraceptive coverage and provide for all women wishing it to obtain it, if not from their own healthcare plan, then through government’s requiring insurance companies to pay the costs.10 This, however, seems objectionable on at least three counts. First, it violates the equality principle in favoring church-affiliated (“religious”) institutions over private employers who may have the same religious objections to paying for contraceptives, thus privileging one kind of religious citizen or group of citizens over another religious kind of citizen or group. Second, it violates the neutrality principle in requiring coverage to be paid for even if it is as deeply disapproved of for secular reasons as it may be for religious ones. There is, e.g., no conscientious objector status for insurance companies that (however unlikely) have deep moral objections. Third, as so far described, it does not cover contraceptive services for men.

As things now stand in the U. S., it is not clear that taxpayers—or at least the administration currently representing them—will approve a voucher system of a kind that would satisfy the demands of many employees in both religious institutions and certain private organizations. I will assume, then, that the question whether the mandate should be imposed on those institutions must (at present) be approached on the assumption that government will not fulfill those demands by vouchers or other “direct” means.

Levels of Coercion and their Normative Significance

These problems must be acknowledged to be serious, but here the main point that emerges from considering the ethics of taxation concerns several different kinds of coercion of citizens. Governmental coercion is our chief concern, but coercion may of course be perpetrated by non-governmental agents.

First-Order Versus Higher-Order Coercion

Consider the difference between requiring citizens to pay taxes to support, say, conducting the Vietnam War—which many Americans opposed—and subjecting them to conscription to fight it, with no exceptions for conscientious objection. Requiring people to fight a war against their conscience might be called first-order coercion: it is a case of forcing them to do the basic deeds,11 such as killing people, that they consider wrong. Requiring them to support someone else’s forcing others to do the deeds in question, as where a government taxes citizens partly to pay for military conscription, is, by-contrast, plausibly considered either complicity in the doing of those deeds or even higher-order coercion. It would not be second-order coercion, since citizens are not themselves coercers of the conscripts forced to kill; but it is, by forced taxation, at least complicity in government’s exercising first-order coercion, and it could be conceived as higher-order coercion of the deeds insofar as it is like empowering officials already intending (and able) to cause those deeds actually to cause them.

As this case shows, both the notions of complicity and of higher-order coercion need analysis and may in some cases both apply to an action. Moreover the order of coercion, as conceived here, is determined by how many levels of coerced or potentially coerced decision (as distinct from the number of individual decisions) are required above the level of the act-type (such as killing civilians) that is the basic object of disapproval or resistance on the part of the coerced. The matter is not as simple as coercing someone, at gunpoint, to coerce a third party to relinquish funds (a case of second-order coercion). Paying taxes that, for instance, support a war does not yield (militarily) killing people except as decided by those who order it or do it, or both; and there might be still other levels of decision. At each level, abstention from the deeds in question is at least commonly a possibility and, if so, the support of the actions the taxpayer disapproves of goes through the agency of someone else. This is morally significant, though by no means the only morally significant element in higher-order cases.12

The Moral Significance of the Order of Coercion

Democracies seem generally—and properly—to presuppose that the case for first-order coercion on the part of government must be stronger than the case for governmental higher-order coercion (though beyond the second-order case there may be no automatic diminution in the governmental responsibility to justify coercion). This presupposition is supported by a number of considerations, including the points that (a) one’s moral responsibility for what one is coerced to support is at least less great than for voluntarily doing the thing(s) in question, and (b) at least commonly, the secondary agent(s)—those supported by the higher-order coercion, may themselves ultimately refuse to do the relevant deed(s).13 A manifestation of the operation of this presupposition in some democracies may be a policy of allowing conscientious objector status for military conscription, which is first-order coercion, but not for the portion of taxation that supports the practice of conscription and military use of conscripts. To be sure, taxation need not be coercive for those who approve of it and of the use of the funds in question.14 But it may be coercion at some higher level, depending on whether government does things to which the taxpayers in question are forced to acquiesce. The kind of difference indicated here between first- and higher-order conscription seems to provide some support for democratic presupposition that—other things equal—the need for governmental justification of coercion diminishes with increases in its order.

The distinction between first- and higher-order coercion for the healthcare issue has an important implication: it can explain why government may require actions at some higher level that it may not require at the basic level. Thus, supposing it cannot properly require contraceptive use by those who disapprove of it, it might still require contraceptive coverage to be funded by disapproving employers or, at a still higher level, by vouchers. Consider an institution that, like some Catholic universities, is self-insured. Requiring it to pay for contraceptive services could be, in certain special cases, roughly higher-order coercion. It would be forced to order lower-order behavior of a kind it disapproves of, such as imposing reimbursement obligations on those who actually pay out the funds. Here the insuring institution may consider itself complicit, even if involuntarily, in wrongdoing. Given the coercion, we might call this complicitious coercion; for institutions that do not disapprove, we would have cooperative coercion. The same distinction would apply to an institution’s being forced to pay taxes to support vouchers that fund the relevant services, but the coerced support would be at least one level higher in the coercive framework: paying government to pay providers such as physicians or pharmacists, versus paying providers directly through funding one’s healthcare plan that compensates them.

Moreover, where, as in the U. S., universities are tax-exempt, it is individuals who would be forced to support the services under a government voucher plan. These taxpayers, even if Catholic, may or may not disapprove of contraceptive use, and all are free to protest its use or to press for its restriction by law. Tax exemption does not apply to businesses, but they are also not religious institutions and so not my main focus here. Their proprietors as individuals, of course, should have the religious liberties justified by the church-state separation principles proposed above.15

The significance of levels of coercion also bears on the prospect—which may be realized in the future—of religious institutions refusing to cover medical services of any kind by physicians or others who provide contraceptive services or certain others, even where those using contraceptives pay the costs. Government’s requiring coverage of this comprehensive kind would be at worst coercion two levels above the one at which the basically objectionable action occurs: coercion to support programs that are required to support providers who support users. Universities, for instance, would be required to support, although though intermediaries, medical activities of physicians who might prescribe contraceptives, thereby exercising presumably free agency, to someone who might use them, thereby exercising free agency at the “basic” level. If some of these physicians also provide legal abortions, the matter is more serious. The difference between levels of coercion, however, is still relevant. Thus, to the points that have emerged so far, we should add that, other things equal, governmental coercion to support voluntary doing of deeds against one’s conscience requires less justification than governmental higher-order coercion whose object is mandating those same deeds. This need not be weak justification; the principle does not concern absolute levels of justification. With all this in mind, let us consider abortion as a foil for the case concerning contraception.

The Special Problem of Abortion

The Obama Administration’s mandate does not require private employers to pay for legal “elective” abortions, and these are the main cases of abortion we must consider in relation to the issue of healthcare policy. Here I have two points. First, this exemption reflects governmental appreciation of the point that the moral case against elective abortions is—or is at least is plausibly judged to be—stronger than the moral case against contraceptive use. On most views of the matter, this applies both to the force of available arguments against contraceptive use as opposed to abortion and to the moral gravity of the act in question relative to that of contraceptive use.

One plausible general principle applicable here is this. Other things equal, where killing a kind of being would be wrong, preventing the creation of one of the same kind, if wrong, would be less objectionably so. A second point is that the scope of the term ‘abortion’ is disputed. Some consider the morning after pill potentially abortifacient; others reject this view. For some people, the issue turns on when, in the period between fertilization of an ovum and its implantation in the uterus, the pill preventively acts and when, during that process or later in normal prenatal development, personhood may be properly ascribed to what might be generically called the union of sperm and egg. This point raises the question whether, in a democratic society, the scope of healthcare and indeed of personhood should be legally defined in a way that is religiously neutral.

Under the neutrality principle proposed here, the answer is affirmative. It should be added that if, as seems plausible, no actual government could allow religious considerations to figure in defining personhood without favoring some religions over others, then the equality principle would also be violated. The plausibility of that conclusion is enhanced by the point that, between and even within different religious groups even in the U. S. alone, there are disagreements regarding abortion itself, and consequentially regarding what counts as desirable healthcare.16 It would seem, in any case, that the liberty principle yields a similar conclusion. Recall that some people may have religious reasons for wanting to limit the size of their families—or, perhaps, to avoid having to tolerate bearing a child as a result of rape. The religious liberty of the latter, like that of women subject to other kinds of coercion by husbands or others, would be abridged by prohibiting use of the morning after pill.

It should be clear, then, that if some religions endorse—on religious grounds such as divine ensoulment of human eggs immediately upon fertilization—the early personhood of those entities, and other religions reject this timing of initial personhood, whether on religious grounds or not, then governmental prohibitions or restrictions of the use of the morning after pill would be de facto unequal treatment of different religions. It does not follow, though it is certainly arguable, that such rulings would be a manifestation, perhaps unconscious, of preference for one or more religions I say ‘certainly arguable’ because, regarding the time or temporal period during which personhood is first present, and particularly regarding the view that it coincides with conception, it is at best difficult to find arguments of a completely religiously neutral kind that carry the conviction of even a near majority of the leading thinkers who have studied the problem without relying on religious considerations or presuppositions.

Are there religiously neutral arguments that might justify law-making that restricts either the morning after pill or at least uncontroversial cases of abortion? It is noteworthy that many in the Catholic tradition who oppose abortion (among others who oppose it) appeal to natural law arguments or other arguments presented in secular terms. But, to a good majority of secular moral and political philosophers, as well as to a large proportion of reflective people in many religious traditions, those arguments do not seem cogent. This suggests that the secular arguments do not justify governmental prohibitions of all elective abortions; and although it certainly does not follow that the arguments dependent on religious considerations are not sound, they are, at least on the church-state separation principles presented here, the wrong kind to serve as a basis for definitions of personhood, or the associated restrictions of liberty, in a democratic society.17

What Counts as Healthcare?

The religious realm is not the only area in which what counts as healthcare is controversial. Where the malady is “emotional,” as with anxiety, there are differences over what is normal, what counts as health, and what should be covered by insurance. With contraception, both emotional and other psychological variables are relevant to coverage, as well as biological factors. People who think that contraception is morally wrong will tend to believe that even for those who disagree on this, it is not a healthcare need. People who believe this might divide over whether, at least for women, contraception counts as a preventive healthcare need in living conditions in which rape or other kinds of sexual coercion are difficult to prevent, as in parts of India and Africa (HIV infection remains a problem in this connection, particularly in certain regions of Africa). In any case, by uncontroversial standards for preventing physically dangerous and psychologically trying or even traumatizing conditions, the grounds for regarding contraceptives as preventive healthcare in those situations—which exist in many parts of the world—are considerably stronger than in countries where women can freely control their potentially reproductive behavior. Here I will assume that these grounds are morally sufficient to make contraceptive coverage a high priority for certain governments, even non-democratic ones.

In relation to the contraceptive coverage issue as understood by private employers facing the Obama Administration mandate, however, even in the case (as with many marriages) in which women can successfully refuse to reproduce—at least after bearing a number of children acceptable to them—some people may doubt whether contraceptives as such are a healthcare need.18 Appraising this doubt is impossible here, but it may be instructive to compare circumcision, which is covered by some healthcare plans. There is disagreement in the medical community over whether this is desirable from a healthcare point of view, but even apart from how that issue may be resolved, two points are pertinent. First, non-circumcision normally does not affect the life and well-being of the males in question to anywhere near the extent to which the unavailability of contraception usually affects the life and well-being of sexually active women. Second, given this, and given how much of the strong sentiment favoring circumcision is religiously grounded, full or partial coverage for it but not for contraception creates a presumption—not irrefutable but difficult to defeat—that the differential treatment of the two by government would violate the neutrality principle.19 Some would likely hold that it also shows a preference for the interests of males over those of females.

The main issue before us, however, is not what actually constitutes healthcare but whether government may require what it considers necessary for adequate healthcare against the religious principles or convictions of those who must provide it, at least in paying for it. If, as I have argued, religious considerations may not, for public policy purposes, be used to define healthcare, then the central question is whether a democratically proper, religiously neutral definition of it may be imposed on private employers whose religious liberty is thereby reduced. The next section will suggest a positive answer for cases in which the burden on religious liberty is not sufficiently great relative to the healthcare benefits to outweigh the government’s case.

The Comparison of Burdens on the Free Exercise of Religion

If, in democracies, what constitutes healthcare cannot be properly defined by religious criteria (or by taking their satisfaction as a necessary condition), then there is no question that democratic governments may count terminating a pregnancy that will kill the pregnant woman as healthcare toward her. But suppose the pregnancy is due to rape and the woman strongly desires to terminate it. This case is more complicated. Even some who believe that termination would in at least some such cases be healthcare would hold that the rights of the conceptus or fetus still preclude the moral permissibility of termination and (some of these people might also argue) should preclude legalizing termination or certainly should prevent requiring private employers to cover it in their plans. Here I would reiterate that government should not be required to define healthcare by religious criteria; but we must grant that forcing certain private employers, at least those who hold traditional Roman Catholic views, to include in their healthcare plans even the kinds of abortions described would be a more serious abridgment of their religious freedom than simply requiring contraceptive coverage. How is the comparison between governmental responsibility and religious liberty to be understood here?

An Incommensurability Problem

There is surely no one value, such as the badness of pain, in terms of which alone we can make the comparison. To be sure, some ethical thinkers might advocate a utilitarian approach. But even if democratic governments could make a commitment to some one kind of ethical theory, there are at least two normative problems that go beyond the difficulty of determining the relevant probabilities for positive and negative outcomes. The first is that on any serious utilitarian view there are at least two kinds of value, roughly the positively and the negatively hedonic, say pleasure and happiness and, on the other side, pain and suffering. The second is that there seem to be qualitative differences in the worth of these. How can we weight, say, the intrinsic value of aesthetic versus culinary pleasures or the intrinsic disvalue of physical pains versus such psychological suffering as acute anxiety and severe depression? Mill and other utilitarians have sketched ways to represent differences of quality as differences in quantity for purposes of moral decision (Mill 1869/1978, ch2). But even if these sketches can be made adequate for utilitarian purposes, there is too much resilient disagreement among morally reflective people to allow democratic governments to determine benefits and burdens entirely by utilitarian standards.20

If utilitarian considerations alone do not suffice to determine when a healthcare requirement unwarrantedly restricts religious liberty, neither do considerations from any comparably simple ethical view (if there is such a view among the most plausible candidates, including Kantian ethics, Aristotelian virtue theory, and Rossian intuitionism). But on any plausible view, considerations of pleasure and pain are important. Economic factors can roughly indicate these, but the incommensurability problem cannot be solved using economic criteria to compare the value of a healthcare requirement against that of a religious liberty it abridges. This is especially so where citizens differ greatly in wealth. One person’s pin money is another lifeline.

If political philosophy is to guide such governmental and institutional decisions as the contents and scope of healthcare in a society, it must respond to what seems an irreducible plurality of values. Among these are the central default values for morally sound democracies: liberty, limited only by considerations of harm, and basic political equality, requiring one-person, one-vote and equal treatment before the law. More specifically, in church-state matters we have identified at least six standards for public policy.

Standards for Guiding Religiously Controversial Healthcare Policy

The first three standards are the liberty, equality, and neutrality principles. Governmental adherence to these may require policies that differentially benefit religious people and institutions, depending mainly on the religious composition of the citizenry. But differential benefit does not necessarily indicate preferential treatment. Sociopolitical injustice may not be inferred simply from differences in benefits or prosperity.

Fourth, with these three principles governing church state-relations in mind, and given that governments should seek to reduce the alienation and resentment that can result from differential benefits, I have proposed the protection of identity principle, on which the deeper a set of commitments is in a person, and the closer it comes to determining that person’s sense of identity, the stronger the case for protecting the expression of those commitments. This bears on the difference between requiring coverage for ordinary contraceptive services and requiring it for what are conscientiously believed to be abortifacients.

Fifth, beyond these points, we have seen that in determining what limitations of religious liberty are permissible in the framework described, governments should distinguish orders and kinds of coercion. Other things equal, the need for justification of governmental coercion is inversely proportional to its order.21 A related principle (which supports the former) is that, other things equal, coercion to do something requires stronger justification than coercion to give indirect support, as by paying taxes, to someone else’s doing it. For reasons indicated above, this applies to individuals and non-governmental organizations such as universities, as well as to governments.

Sixth, a lesson of our discussion of double effect indicates that, negatively, a limitation of liberty , such as requiring religious employers to adopt a healthcare plan they object to, is not necessarily justified when it is a collateral consequence of doing something whose value outweighs its disvalue, say guaranteeing adequate healthcarfe for all citizens. A plausible principle of double effect would apply only where there is no preferable way to achieve the greater value—such as using vouchers to guarantee adequate healthcare to all employees—without the bad consequence. A major problem here is to determine what alternatives are preferable. For instance, how far should governments go in using tax revenues to avoid burdening free exercise? In such cases preferability may be taken to be in part a matter of majority vote: in democracies, majority vote is a prima facie normative reason for government to realize the preferred state of affairs.

One further consideration should be brought to the fore. No adequate set of standards to guide public policy can be so precise and so clear in its requirements that morally responsible conduct by individual citizens—especially if they are legislators, judges, or executives—is not needed for the flourishing of the society as a whole. With this in mind, I have proposed, especially but not solely for the domain of church-state relations and matters of religious liberty, a principle of secular rationale—alternatively (in ideologically neutral terms) the principle of natural reason: Citizens in a free democracy have a prima facie obligation not to advocate or support any law or public policy that restricts human conduct, unless they have, and are willing to offer, adequate secular reason for this advocacy or support (e.g. for a vote).22 This principle is in no way anti-religious; it simply states a (defeasible) necessary condition for justifying coercion. The condition is one that even religious people should accept insofar as they consider impartially the alienating repugnance of being compelled to do something for reasons tied to someone else’s religion.

It should be obvious that citizens internalizing this principle will tend to support government’s adhering to the other principles proposed above. It is of course not obvious what counts as an adequate reason, but this is a general problem for normative decision-making and needs no special treatment here. The principle is one that many religious people seem guided by even if only at the level of presupposition. Many, especially in the Roman Catholic tradition, try to find good arguments not dependent on theology at least where they burden other citizens. Natural law arguments are often thought—controversially, to be sure—to have this status.

The principle of secular rationale may seem to imply that religious reasons have no normative force or at any rate may be ignored in the ethics of citizenship. This is not so, and a plausible companion principle addressed to religious citizens is the principle of religious rationale: Religious citizens in a free democracy have a prima facie obligation not to advocate or support any law or public policy that restricts human conduct, unless they have, and are willing to offer, adequate religious reason for this advocacy or support. This principle admittedly might burden some of the political activities of some religious people; but the obligation is prima facie, and where a religion does not bear on an envisaged law or public policy, either the prima facie obligation is overridden or the principle may be considered inapplicable. The principle would, from some religious perspectives, support considering unequal healthcare coverage invidious. That conclusion could, for instance, be considered implicit in “Do unto others as you would have them do unto you.” Similarly, suppose the morning after pill is considered an abortifacient on religious grounds such as clerical pronouncements. Would rejecting its inclusion in a comprehensive healthcare policy on those grounds violate the Do-unto-others rule, at least for those who would resent burdens on their exercise of freedom owing to pronouncements of clergy in some religion not their own? Whatever the answer, those abiding by the secular rationale principle would tend not to reject its inclusion at least if governmentally funded. This leaves open, of course, whether, for government, it is (as I have suggested) better to cover such cases directly rather than requiring employers to do so, even if the relevant funds are given to them for distribution. If healthcare is nationalized, however—a policy change that, for the U. S., at least, raises issues not addressed here—there is little question that the framework of this paper would indicate the desirability of including that pill along with other contraceptive services.

Democracy is a negotiatory framework. The preservation of liberty and equality are essential if it is to realize the ideal of government of, by, and for the people. Coercion by laws and institutional policies should be minimal. Where standards of healthcare or other elements of the well-being of the populace must be guaranteed, persuasion is better than coercion. This paper presents a framework for guiding, and indeed for minimizing, coercion in church-state matters when it is necessary and for engendering persuasion in those matters where persuasion is possible. Toward these ends, I have proposed a number of connected principles: for the institutional realm, three principles of separation of church and state, a principle concerning the protection of identity, and another concerning the justification of coercion at different levels; for the realm of individual citizenship, principles of secular rationale and religious rationale. My hope is that, taken together, these principles may guide governments and institutions and enhance both the liberties and the moral standards of individual citizens.

Acknowledgments: For helpful comments on an earlier draft I thank Nevin Climenhaga, Michael Deem, Shannon Gilreath, Kathryn Pogin, Benjamin Cohen Rossi, and an anonymous reader for the Journal.

References

Audi, Robert. 1974. Moral Responsibility, Freedom, and Compulsion. American Philosophical Quarterly.

———2001. Religious Commitment and Secular Reason. Cambridge:Cmbridge University Press.

———2011a. Democratic Authority and the Separation of Church and State. Oxford: Oxford University Press.

Bronner, Ethan. 2013. “A Floor of Suits Fights Coverage of Birth Control. New York Times, January 27, 1 and 11.

Cullity, Garrett. 2004. The Moral Demands of Affluence. Oxford: Oxford University Press.

Eberle Christopher J. 2002. Religious Convictions in Liberal Politics. Cambridge: Cambridge University Press.

Lichtenberg, Judith. 2010. Negative Duties, Positive Duties, and the ‘New Harms.’ Ethics 120, 557—578.

Mill, John Stuart. 1869/1978. On Liberty. Indianapolis. Hackett.

Rawls John. 1971. A Theory of Justice. Cambridge, MA. Harvard University Press.

———1993. Political Liberalism. New York. Columbia University Press.