Introduction
Contemporary ethics and political philosophy have not enjoyed nearly as much influence upon discourse in philosophy of religion as has work from contemporary epistemology and metaphysics. It is difficult to surmise why this is so, but we speculate that part of the reason is that much of the work carried out by contemporary ethicists and political philosophers is, if not incompatible with religious ethics, much more difficult to harmonize with certain theological themes and doctrines than certain metaphysical and epistemological theses. Moreover, while rights-based accounts of morality have become much more prominent in the past several decades, minimal attention has been given to the question of whether God has any rights with respect to humanity. One aim of this article, then, is to fill this lacuna by exploring the moral relations between God and humans, focusing especially upon the question of whether God possesses any claim-rights with respect to humanity. Following the Hohfeldian schema, we conceive of obligations as correlative to rights, and so, on the framework pursued here, theorizing about God's rights would be part and parcel of theorizing about humanity's obligations to God.
Moving forward, according to Jesus, the first and greatest commandment, and so the one which is most central to Christian practice, is the command to love God. In this article, we argue that this commandment is best interpreted in aretaic rather than deontic terms. In brief, we argue that there is no obligation to love God. Although it may reflect poorly on one's character, and in the end spell spiritual and prudential disaster, failure to seek and enjoy a union of love with God is not in violation of any general moral requirement or duty.Footnote 1 The core argument is straightforward: relations of intimacy should not be morally imposed, via duties and obligations, upon autonomous beings. We contend that such reasoning applies to human beings' relationship to God. So, even if our ultimate end is to enjoy communion with God, God has no right to human beings' seeking out a relationship with him. We believe this thesis is interesting in and of itself, but moreover, if correct, it follows that the command to ‘love God’ is not the sort of moral principle that can be supported by threats of hellfire or other forms of punishment. Given this, we allege that traditional accounts of hell, which assume that rejection of God makes one vulnerable to divine punishment, are morally problematic.
We begin by laying out our understanding of moral rights and duties, drawing special attention to the concept of the suberogatory and the distinction between culpability and deplorability. It is worth emphasizing, at the start, that even if one does not carve up the deontological landscape exactly as we do (e.g. one might maintain that there are unenforceable duties), as long as one takes moral rights seriously, the core argument developed below should still hold water. Next, we argue that there is no divine right to receive worship or love from autonomous beings, and thus there is no corresponding obligation for human beings to love God. We then consider and respond to several objections and close with a discussion of what implications our thesis has for the doctrine of hell.Footnote 2
Rights, forfeiture, and the suberogatory
Not all moral considerations generate moral requirements or obligations. For example, there is a very clear sense in which, as a matter of courtesy, I ought to hold the door open for the person walking behind me. But the sort of ‘ought’ used here does not establish an enforceable moral claim upon me. While it might be rude of me to let the door slam shut in a person's face, I do not flout any obligation or duty by doing so. Among other things, we take this fact to mean that while my inaction might be morally deficient in some sense, it is not the sort of behaviour which licenses interference from others. I would be acting within my rights by choosing to let the door slam shut. On the other hand, if I attempt to rob a bank, others may permissibly interfere and use coercive measures in attempting to prevent my thievery. J. S. Mill (Reference Mill1979, 47–48) writes:
Duty is a thing which may be exacted from a person, as one exacts a debt. Unless we think that it may be exacted from him, we do not call it his duty . . . There are other things, on the contrary, which we wish that people should do, which we like or admire them for doing, perhaps dislike or despise them for not doing, but yet admit that we are not bound to do; it is not a case of moral obligation; we do not blame them, that is, we do not think they are the proper objects of punishment.
Obligations, then, are necessarily enforceable; failure to uphold one's duty licenses third-party interference and makes one liable to punishment.Footnote 3
Following Julia Driver (Reference Driver1992), we understand actions which merit negative evaluation, but which are nonetheless permissible, to be suberogatory. Although the concept of supererogation has been widely recognized and employed by ethicists, we believe its twin concept – that of the suberogatory – is of equal, if not greater, philosophical significance. A supererogatory action is one which, while being laudable, is not morally required. By contrast, a suberogatory action is one which, while bad, is not impermissible. To illustrate the distinction with a simple case, consider the following. Andrew is hosting dinner and invites his friend Niklas over, who has just finished a marathon. Andrew is sitting in his recliner – the only chair in the whole house – and sees that Niklas is exhausted, but fails to offer Niklas his seat. In this scenario, it seems that Andrew ought to give Niklas his comfy spot. However, if Andrew fails to do so, he does not violate Niklas's rights. While his actions might be suberogatory, Andrew is acting within his moral rights by keeping his spot, and so does nothing impermissible.
While the above example is rather mundane, it is worth highlighting just how deplorable some suberogatory acts can be, even if such acts are ultimately deemed permissible. For example, Driver (Reference Driver1992, 287) illustrates the distinction between the suberogatory and the impermissible by considering a case in which a brother refuses to donate his kidney to a sibling dying of severe renal disease. Assuming the only reason for failure to donate in this case is mere inconvenience, the brother's actions plausibly reveal selfish attitudes which merit disapproval. Yet, refusing to donate his kidney does not violate any general moral requirement or obligation. The ill sibling has no moral entitlement to their brother's kidney. Or consider a more controversial but real-life case that occurred several years ago. A lawyer was caught speeding and was pulled over by a law enforcement officer. As the officer wrote the lawyer a traffic ticket, another driver lost control of their vehicle and careened into the officer, immediately killing him. When the lawyer saw this, he promptly jumped on social media where he made a derisive post gleefully mocking the officer's death. How should the lawyer's actions be evaluated? Without a doubt, the lawyer behaved terribly and revealed himself to have a loathsome and odious character. Yet this aretaic evaluation of the lawyer's character is distinct from a deontic evaluation of the lawyer's action. Although revealing of a vicious character, the lawyer's actions (apart from the initial traffic violation) seem to fall short of being truly obligation-failing. There is no moral duty to refrain from mocking another's misfortune and so the officer had no right against being mocked.
The distinction between the suberogatory and prohibited naturally raises the question: how are we to demarcate the boundary between actions which are bad-but-still-permissible from those which are truly verboten? On our view, only a rights-based account can provide an adequate answer. We believe our moral obligations perfectly correspond to people's moral rights. Thus, an agent is only guilty of failing to uphold some duty when she violates the rights of another. Furthermore, we believe that when a person culpably violates the rights of another, she forfeits some of her own rights against the imposition of harms.Footnote 4 Importantly, the notion of forfeiture is employed in two different contexts – those of defensive force and punishment. We believe these two domains identify two distinct ways in which moral rights are enforceable. Typically, then, when an agent culpably violates the rights of another, she will forfeit her own rights against the imposition of both defensive and punitive harms.Footnote 5 In other words, we take it that the use of coercive measures to compel or punish those who fail to do what they morally ought to do is only justified when a person intentionally and culpably transgresses the moral boundaries of another, thus forfeiting some of her own rights. If Sally burns down Jim's house, violating the latter's property rights, we believe Sally forfeits her rights against the infliction of proportionate punishment.Footnote 6 However, we do not think that the same can be said of the schadenfreude-filled lawyer above; while despicable, unless one thinks the lawyer's social media posting constituted a rights violation, the lawyer forfeited no rights. To reiterate, we believe that claims of enforceability track rights and obligations, but not aretaic considerations.Footnote 7
Of course, it would be silly to insist that this is how these moral terms are always used. There are different ways of carving up the conceptual landscape and many philosophers endorse the possibility of there being unenforceable duties.Footnote 8 However, because of (i) the importance of keeping track of what sorts of actions result in rights forfeiture, and (ii) the fact that we have at our disposal concepts, such as the suberogatory, which allow us to speak about morally-bad-though-permissible actions, we believe it is theoretically fruitful to reserve talk about our moral duties to those claims which are essentially enforceable – that is, those actions which license interference via defensive action and/or punishment. Indeed, beyond mere terminological dispute, it is difficult to understand what would differentiate a suberogatory act from one which violates an unenforceable duty.Footnote 9 Furthermore, even if one disagrees with this picture and adopts a more expansive conception of rights and duties – one that grants the conceptual possibility of there being unenforceable duties – our main thesis and argument will be little effected, for the main claim advanced below is that there is no enforceable obligation to love God.
To summarize: persons are bound by obligations – obligations grounded in the rights of other persons. A person only forfeits rights against punishment when she culpably attempts to violate the moral rights of another. Importantly, an agent can engage in non-rights violating behaviour that may still be deemed morally bad and/or vicious, but if there is no rights-related culpability involved, such suberogatory behaviour does not open one to third-party interference via defensive force or punishment. The lawyer is of course susceptible to criticism, but he cannot be tossed in prison or subjected to other forms of hard treatment characteristic of punishment because his gleeful reaction to the officer's death violates no rights.
Why there is no obligation to love God
What implications might the above moral framework have for our understanding of the divine–human relationship? Some might plausibly say ‘none’. It is not uncommon for philosophers of religion to contend, for instance, that God has no moral obligations to human beings, and thus we have no rights against how God may treat us. For example, Marilyn Adams (Reference Adams1999, 94) argues that given the metaphysical gap between God and humans – namely, the fact that divine and human natures are so radically different as to be incommensurate and belonging to wholly separate ontological categories – it would be foolish to think that our system of moral rights and duties could be mapped onto the God–human relationship as though God and humans occupy the same social world. It is beyond the scope of this article to take up such arguments. Rather, we take as a theoretical starting point that God and human beings do not occupy different moral networks. Indeed, we believe it makes little sense to speak of God's justice towards us if we cannot demand that God treat us in certain ways.
This article aims to explore a different but parallel question to the one about God's obligations, namely: what sort of rights does God have with respect to humanity? The question is significant because, on our account, if there are no good reasons for thinking that God possesses certain rights in relation to human beings, then human beings have no correlative duties in relation to God. To put it another way: if there are no divine rights, humans would have no obligations to God, and so, it would be impossible to directly and culpably wrong God in non-aretaic terms. This would be a surprising conclusion to reach.
We suspect that talk of divine rights will strike many as unnatural, or, at least, as rather awkward. The reason for this is straightforward enough: despite being a welfare agent who has interests, an infinite and omnipotent God is simply not vulnerable or susceptible to harms in the same way that frail and finite human beings are. According to the two most dominant theories of rights, the function of rights is either (i) to protect people's freedom and autonomy (the will-based theory of rights), or (ii) to protect people's fundamental interests (the interests-based theory of rights). With regard to the former, because human beings cannot do anything to contravene or violate a sovereign God's freedom, talk of divine rights might seem superfluous or wrong-headed. Typical candidates of negative rights, such as the right against being maimed or killed, obviously need not be posited in order to protect God's being and will. Human beings (or any other non-divine beings) simply cannot harm God in such ways.Footnote 10
But what of God's interests? We take it as a given that, according to the most plausible versions of Christian doctrine, God is omnibenevolent and thus, as a matter of logical necessity, God has a fundamental interest or desire that all people come to know and enjoy a relation of love with him. Indeed, the creative and salvific plans of God are rooted in the singular desire for creatures to discover and return to God. Given this, God's interests are undoubtedly hindered (and God is, in some sense, harmed) when people use their creaturely freedom to reject and turn away from God. However, our main contention is that it would be a leap of logic to conclude from this fact that God has a right to human beings' seeking a union of love with him.Footnote 11
To illustrate the point, consider an example: Abby is desperately in love with Olivia, and desires to marry her. Suppose further that if Olivia were to respond positively to Abby's professed love and make a life with her, she would experience true and long-lasting happiness. Nevertheless, we take as an unobjectionable moral truth that it is within Olivia's rights to reject Abby's proposal. Even if both parties would be better off if Olivia agreed to marry Abby, we think that most would agree that Olivia has no moral duty to accept the marriage proposal. Moreover, it would be egregiously unjust of Abby to, upon rejection, kidnap and imprison Olivia in the basement of her house or punish her in some other fashion.Footnote 12 This is because relations of intimacy – even those that would be tremendously mutually beneficial – should not be imposed upon morally autonomous agents. No one can legitimately use coercion, violence, or punishment to procure love from autonomous beings.
There are at least two reasons in support of this basic moral truth. First, and most importantly, we believe that the value of autonomy has deontological significance in the context of intimate attachments such that a person should be able to exert control over her self-regarding affairs in this realm, regardless of any countervailing prudential considerations. Part of what it means to be an autonomous agent is to have morally privileged control over certain core aspects of one's identity and choices. We are our own individuals with our own life plans. Moreover, because relationships of intimacy are so tightly connected to our sense of self and practical identity, the moral boundaries which protect oneself and the ability to shape one's life according to one's priorities in this realm deserve especially strong moral protection. Second, we would contend that relations of intimacy, such as friendships or romantic relationships, are often more valuable when freely chosen rather than forced upon us – that the exercise of autonomy in the domain of intimacy can augment the comparative value of our relationships. For example, we believe that an arranged marriage between John and Jane will typically be less valuable than a freely chosen marriage between John and Jane. This suggests that the exercise of autonomy in the context of intimate relationships is both intrinsically and instrumentally valuable.Footnote 13
Moreover, we allege that something very similar is true of our relationship to God. Even if God has a deep and immutable desire that we love and enjoy fellowship with him, given the deontological significance of creaturely autonomy, God has no right to a fellowship of love with us, so there is no correlative duty for human beings to seek communion with him. Indeed, human beings are free to respond to God on their own terms; and our response to God's invitation to a union of shared love loses value to the extent that it is coerced or imposed upon us through threats of divine punishment.Footnote 14 Given all this, the intriguing conclusion is that human beings have no general moral obligation to seek a union of love with God.Footnote 15 To be sure, while we do not deny that love of God is both prudentially and morally significant, we think the language of obligation – at least where such terminology is tied to claims of enforceability – does not appropriately capture the dynamics of the divine–human relationship. If there are moral reasons to love and seek communion with God, such reasons belong to the realm of the aretaic, rather than to deontological morality. That is, failing to seek and love God may speak poorly of our character, but it is not obligation-failing. Since relations of intimacy between autonomous agents cannot be forcibly imposed, there can be no divine right to a union of love with human beings, and so, human beings do not categorically (or culpably) wrong God by failing to enter a union of love with him.
What consequences might follow from this conclusion? Consider the doctrine of hell. Some Christian traditions teach that God consigns unrepentant sinners to everlasting conscious torment in hell as a response to earthly transgressions. Given that the Greatest Commandment takes precedence over all other divine commands, it is standard to think that violation of this commandment culpably wrongs God and so makes one vulnerable to divine punishment.Footnote 16 There are numerous problems with this doctrine, the most prominent of which is that eternal damnation hardly seems a proportionate response to even the most egregious of finite wrongdoings. That is, eternal punishment in hell is just far too severe a punishment for sins that have but a finite, temporal impact. But here is another reason to reject traditional accounts of hell: human beings do not culpably wrong God by rejecting a relationship with him. Given that autonomous agents enjoy a general moral prerogative to accept or reject relationships involving intimacy, failure to love God does not result in violation of any divine rights. Yet, only rights and correlative obligations can be backed up by coercive measures such as punishment. Put another way, a person only forfeits her rights against punishment when she culpably violates someone else's rights. But it is not clear that God has any rights in relation to humanity. And if God has no general right that people enter into a communion of shared love with him, it follows that any model of hell which rests upon the assumption that rejection of God makes one liable to divine punishment is in serious error. The Greatest Commandment is simply not the sort of moral principle which can be backed up with threats of fire and brimstone. Failure to love God is, at best, suberogatory.
Objections
Here, we consider objections to the thesis that there is no morally enforceable duty to love God. The response to each objection is broadly the same: though permissible, rejection of God is spiritually and prudentially disastrous and speaks poorly of our characters.
Obligations of gratitude
It might be objected that acts of goodwill, grace, and generosity generate an obligation of gratitude. For instance, one might think that the gift of creation and life, and the grace offered through salvation, establish an obligation of gratitude toward God.Footnote 17 This is not a convincing objection, but by responding, it allows us to emphasize once more that the act of turning away from God, even if permissible, might merit negative evaluation. Suppose Amy and John have a child, Damien, whom they raise to the best of their abilities. They teach and demonstrate love, kindness, and generosity to Damien. Yet, for whatever reason, when Damien turns twenty years of age, he decides to cut himself off from his parents. He says some mean things about how he does not care about or love his parents and refuses to visit or call, shunning them from his life entirely. Damien's parents are emotionally crushed.
We think it is uncontroversial to describe Damien's actions as morally bad and so merit negative appraisal. After all, Damien's behaviour is extremely hurtful. Even so, such behaviour is not the rights-forfeiting type. Even if Damien's actions make everyone involved miserable, his actions do not violate the rights of his parents, and so are best dubbed as suberogatory. To see this, consider how little sense it would make for Damien's parents, in response to his familial evasion, to take him to court and attempt to have him tossed in prison because he has ostracized them from his personal life. So it is with God when those he has gifted with life and grace fail to reciprocate such love. Certainly, God's grace and love are reason-giving, but as we have tried to emphasize throughout, there are different kinds of moral reasons, and not all such reasons generate moral duties. Failure to give to God the respect he deserves may demonstrate insufficient appreciation of God's grace and goodness, revealing selfish and perhaps even morally repugnant attitudes, but while all of this is bad, it is less than obvious that such actions amount to a violation of any rights that God might have with respect to humanity. And so, in turning away from God, a person does not make herself liable to the infliction of any sort of hard treatment.
One might push the objection further though, claiming that there is an important disanalogy between our relation to God and the relation between parents and their children. Parents give life to their children, but God creates and sustains all creatures. And so one might think that our deep metaphysical dependence on God generates a special sort of obligation of gratitude that might not be present, or wholly present, between parents and their children. While thankless children are merely nasty, thankless creatures violate deep moral obligations.
In response, while we admit that the analogy is imperfect, in that creatures depend on God in a much more robustly metaphysical sense than do children upon their parents, we do not think this objection changes much. It is not clear what it is about the creation relation that could change the nature and value of relations of intimacy among autonomous beings. Whether I owe gratitude to my parents for giving me life or the stranger who simply holds the door open for me, I still have no obligation to form an intimate relationship with either of them. So it is with creation. While I ought to show God greater gratitude for my being than I show my parents for my life, it is true nonetheless that I have violated no divine rights by failing to seek out God. It might be true that my character is even more deplorable if I reject a union of love with God than if I reject a union of love with my parents, but here we are still operating in the realm of the aretaic.Footnote 18
The value of autonomy
We have argued that the value of autonomy makes it such that intimate relationships may not be coercively forced upon persons. One might find this generally plausible but think it disanalogous when applied to the divine–human relationship. Perhaps when considering relationships among people, we ought to treat autonomy as sacrosanct (at least in the realm of intimate attachments), but, when dealing with our relationship to God, autonomy matters much less. Suppose you think that the God–human relationship is modelled best by the parent–child relationship. For, when compared to God, we are small, helpless, and needy, like young children. But if we are tantamount to small children in the eyes of God, then it is not clear that respecting autonomy is morally important. Parents do not need to take up considerations of autonomy when dealing with young children, as the latter are not capable of making rational decisions but rather are completely dependent for their survival and well-being on their caretakers. In the same way, one might think human autonomy is inapplicable when analysing the divine–human relationship; God can coerce humans into an intimate relationship with him just as parents can force a diaper change upon a baby.
We find this to be a serious objection, but simply have different intuitions about what sort of analogy best captures the moral relations between God and humanity. In some contexts, it may make sense to speak of human beings as small children in comparison to God – for instance, when we consider how fallible and short-sighted we are in our decision-making, and then contrast our limited epistemic condition with that of an all-knowing and perfectly wise God. In that case, the parent-child analogy might seem apt. But when considering the moral relations between God and humans, the most obvious of empirical facts is that God values creaturely autonomy, for in fashioning humanity in his own image, God did not create beings which operate mechanistically via automatic reflex or through some simplistic stimulus-response function; rather, God gifted Man with the ability to exercise reason and to be the author of his own actions and choices, including the ability to respond positively or negatively to God's invitation to share in a communion of love with him. Thus, in keeping with the line of reasoning pursued above – that freely chosen relations of intimacy are more valuable than those which are coerced – we allege that, to the extent that humanity's freedom and autonomy is de-emphasized in the spiritual domain, the divine–human relationship is devalued. A mature love of God is not the result of some puppeteered process or spiritual browbeating, but is borne out of the enduring decision of an autonomous agent who, after reflecting upon the various ways by which she might pay heed to her own humanity and spirituality, chooses to turn towards God in recognition of the divine grace, love, and generosity which precedes all. But if this is right, that God has conceded having complete and total control over us and our affairs, and autonomy is a sufficiently important moral consideration even in the context of divine–human relations, then the rights-based account of morality sketched above seems an appropriate lens through which to view God's relationship to us.Footnote 19
Nature fulfilment
One might object that human beings have an obligation to love and worship God because communion with God is the final end of humanity. To act against God's calling is to act against one's human nature, to deny one's telos. Put another way: to fail to love God is to be an ill-functioning human. This might be correct, but we would allege that such teleological considerations provide well-being-related reasons to love God, not moral reasons. Compare: eudaimonistic theories of welfare claim that a person flourishes to the extent that they develop and realize certain natural capacities tied to the core part of their humanity. Nature-fulfilment theories focus especially upon the exercise of our rational capacities, since rationality is allegedly a uniquely human characteristic. Supposing this were all true, we would argue that a person who lacks discipline and endurance, squandering away whatever innate talents she might possess, makes a prudential error – not a moral one. In a similar vein, we allege that rejection of God, while morally permissible, might have prudentially disastrous results, but it is doubtful whether such welfarist or teleological reasoning has any bearing upon deontological morality. Given that the core machinery of our argument is thoroughly deontological, we do not find teleological considerations especially pressing, at least with regard to talk of rights and obligations.
Divine opprobrium
We have argued that failing to seek a union of love with God is bad only in the sense that it reflects poorly on our characters. When Susie rejects God, she acts deplorably, but she does not violate any obligations to God, for God has no right to an intimate relationship with creatures. One consequence of our view is that Susie is not liable to divine punishment for rejecting God, since punishment is warranted only when rights are violated. But it seems plausible that punishment, or something very much like it, is warranted in certain cases where no rights have been violated. If it is discovered that Bob the carpenter holds racist views, many will find a suitable response to be to no longer seek Bob's services. Bob might be socially shunned, cast out. This response to Bob's holding racist beliefs may strike many as reasonable, even though Bob has violated no one's rights by holding racist views. So, it seems, even though Bob has failed to violate anyone's rights, punishment, or something like it, via an opprobrium of sorts, is an appropriate and justified response to him. Might something like this be true of the divine–creature relationship? Can God shun or otherwise stigmatize us for evincing vice and failing to seek him out?
A few points in response. First, to reiterate, we understand punishment as essentially involving hard, stigmatizing treatment – the type of treatment that would typically violate the rights of the punished.Footnote 20 So, if God were to reproach, shun, or otherwise ostracize creatures for failing to seek a union with him, he would not be punishing us, strictly speaking, for we have no right to the fruits of a relationship with God (just as he has no right to our seeking him out). Just as Bob violates no one's rights by merely holding racist views, we have no obligation to patronize his business, as he has no right to our money. Second, and most importantly, our view is wholly compatible with a ‘mild’ or ‘Natural Consequences’ view of hell. According to such views, hell is not imposed upon sinners in response to their wrongdoing, but rather, damnation is the natural state of being which follows from our free and wilful rejection of God – a state of spiritual disattunement, deprivation, and emptiness.Footnote 21 Importantly, however, such disattunement and alienation is not punitively inflicted hellfire; it is simply the experienced absence of a union with God.
Conclusion
Up to this point, we have restricted our terminology to the use of concepts and language employed by contemporary ethicists. In conclusion, it is nonetheless worth asking how the above argument might relate to the Christian doctrine of sin. Unfortunately, we take the notion of sin to be one of the more complex and nebulous doctrines taught by the Church. Noting this, we believe one might incorporate the above arguments in one of two different ways. First, if one defines sin in relational terms, as a disruption in the relationship between God and humanity, the thesis defended here merely entails that sinning against God (by failing to adhere to the Greatest Commandment) is best construed in aretaic rather than deontic terms. Our aim has not been to discard or undermine the significance of the Greatest Commandment, but rather, to argue that the command does not establish a rights-correlating obligation. However, if one assumes a version of Divine Command Theory and a more legalistic conception of sin, as a transgression against the law of God, the above arguments read a bit more antagonistically to Christian ethical theory, and raise the following questions. Why think God can make enforceable claims in a domain of action where the ability to freely choose seems to be of utmost importance? Is personal autonomy negligible in the context of the divine–human relationship? Does the threat of punishment not diminish the value of a person's response to God? While we doubt that our argument here will convince defenders of Divine Command Theory, our hope is that it at least pushes them for clarification on these important questions.
Acknowledgments
Special thanks to Kit Wellman, Zachary Manis, Jonathan Kvanvig, Billy Dunaway, Gina Cordovi, and three anonymous referees for their helpful feedback on this paper.
