The right to a fair exit

Abstract

This paper introduces a novel account of freedom of dissociation, construed as the “right to a fair exit.” It defines freedom of dissociation as the right to end an association without excessive and undue costs. This novel account contrasts with the classic right of exit that some liberal philosophers have theorized as the bedrock of associational freedom. The original right of exit is first and foremost concerned with the protection against excessive exit costs, while the right to a fair exit broadens its scope to include undue costs as well. State policy should aim at monitoring and resourcing exit not only to ensure that leavers have adequate exit options, but also that the costs of exit are fairly divided between parties. This account illuminates a range of cases, such as divorce, dissolution of employment contracts, and contested exits from religious associations.

Introduction

Exit from voluntary associations is a focal point for claims of entitlement and complaints of injustice. Beyond the diversity of our associational commitments, there are strikingly analogous normative issues related to leaving, be it by force or by choice. Consider these two examples of individuals leaving voluntary associations, respectively, employment and marriage.

In 2005, Carly Fiorina, CEO of Hewlett-Packard, was fired. She left with a $40 million golden parachute ($21 million in cash and $19 million in stock and pension benefits according to Time), provoking the ire of stockholders who filed an (unsuccessful) class action suit against her. ¹ In a memoir titled Tough Choices, she described being mortified by how the Board treated her—no open dialogue, no direct explanation, no gratitude, not even a proper good-bye (Fiorina, 2011). “Life isn’t always fair,” sighed Fiorina, reflecting upon the perceived procedural unfairness of her departure. Was Fiorina's dismissal actually unfair? Didn’t she receive colossal benefits? A CEO can be fired for many reasons, ranging from poor performance, reputational damage, to relationship issues with the Board or the team. Fiorina knew very well that she could be fired by the Board without violation of contract. What she denounced, however, is the fact that no justification was given. According to Fiorina's own judgment, there was an issue of due process in the way the dismissal happened, as the Board is bound to provide continuous feedback on CEO's performance. A proper assessment of this case would require to hear all sides of the story, but it suffices to illustrate how claims of unfairness can emerge in spite of tremendous exit benefits.

Meanwhile, the same year, Katie Allison Granju, writer specialized in parenting, was threatened to lose the custody of her three young children in a divorce lawsuit. ² Accused by her ex-husband of being an “unfit mother,” she described how costly the separation was—hearings, mediation, parenting classes and evaluations, legal fees, and the debilitating fear of having her children taken away from her. The costs she bore for divorcing, including the bitter humiliation, were indeed colossal—all this seemed terribly unfair.

Liberal political philosophers insisted that individuals have a right of exit from voluntary associations (Barry, 2001; Billingham, 2019; Borchers and Vitikainen, 2012; Eisenberg and Spinner-Halev, 2004; Green, 1998; Kukathas, 1992; Rouméas, 2020; Spinner-Halev, 2000). Discussions have mostly focused on religious and cultural groups, but have recently expanded to include marriage, the market, and even the state (Taylor, 2017). ³ The primary concern was to ensure that individuals are not coerced into remaining in associations that undermine their integrity. Individuals do not always freely join associations—they can be born into them—but they should always be free to leave. The costs of exit should not be excessive, and the state should sometimes intervene to minimize exit costs.

This paper introduces a novel account of freedom of dissociation, construed as the “right to a fair exit.” I define freedom of dissociation as the right to end an association without excessive and undue costs. Excessive costs make exit prohibitively expensive. As a result, leavers do not have adequate exit options and might find themselves in a state of deprivation after exiting the association. Undue costs are different. They refer to the fair distribution of the burdens of exits among dissociating parties. A divorced spouse may have sufficient means to avoid deprivation, but find herself burdened by an unfair exit deal—as illustrated by the case of Granju threatened to lose the custody of her children.

This definition encompasses both chosen and unchosen exits. This novel account contrasts with the classic right of exit that liberal philosophers have theorized as the bedrock of associational freedom. The original right of exit is first and foremost concerned with the protection against excessive exit costs—ensuring substantive exit opportunities—while the right to a fair exit broadens its scope to include undue costs as well. This account can illuminate a range of cases, such as divorce, dissolution of employment contracts, and contested exits from religious associations. ⁴

The argument proceeds as follows. I first present the right of exit as it is commonly formulated and I critique its sole emphasis on protecting the least advantaged party (I call it “a right for the vulnerable”). Liberal philosophers have been too concerned with minimizing exit costs for non-domination purposes and have overlooked the question of fairness. I then advocate a novel account of the right to exit, conceived of as a right to a fair exit. Specifically, I argue that this right should not be solely concerned with excessive costs, but also with fairness—with the fair allocation of the costs of exit or the avoidance of undue costs—and that this should encompass unchosen exit as well.

The right of exit, a right for the vulnerable?

Freedom of association is widely viewed as a fundamental right. It is broadly defined as the right to form intimate unions and collectives to pursue shared purposes, be it the love of God or profit. A rich associational life is considered valuable for liberal democracies: associations can play a political role in counterbalancing the power of government; they can realize essential liberties, such as religious freedom; they can nurture some virtues and dispositions essential to democratic life; or as Nancy Rosenblum argued, they can allow for the “experience of pluralism,” namely the ongoing possibility of shifting association. As she put it, “forming, joining, schism, and disassociation are as much a part of freedom of association as the solidity of identification and belonging” (Rosenblum, 2000). ⁵

Dissociation is thus presented as an integral part of freedom of association, together with the right to exclude and the right to organizational autonomy (Brownlee, 2019). The right to exclude is essential for guaranteeing members “integrity interests” (White, 1997). Without control over membership, the tenets of the association could be significantly altered. The right to organizational autonomy can be justified on similar grounds—some collective autonomy is essential to the pursuit of associational purposes. What is special about the right to exit is that it also protects in-group members against the group to which they belong. They shall not be coerced to remain should they wish to leave. It entails that groups have a duty not to deliberately enforce excessive exit costs on their members.

Which interests does the right of exit protect? Fundamentally, the right of exit protects individual agency by opening the possibility of leaving. It secures a space for choice with the option of opting out. The right of exit guarantees the consensual nature of membership and the ongoing possibility of dissent. It grounds voluntariness. It helps ensure that the associational landscape remains fluid and that individuals are not trapped in unwanted affiliations. In so doing, it is a crucial protection against domination and abuse (Green, 1998; Taylor, 2017). To be sure, leaving is not necessary for the protective function of exit to be effective. The possibility of exit can make voice more effective by increasing the bargaining power of dissenters (Hirschman, 1970). Regardless of whether one leaves or not, having the option of exit can enhance one's voice in an association.

The specific contextual freedoms that the right of exit protects are contingent on the nature of the association. Divorce secures marital freedom and protects spouses from domestic abuse. The right to resign secures market freedom and protects workers from exploitation. The much-discussed right of exit from cultural and religious groups secures the freedom to pursue one's conception of the good and protects members from violations of integrity, for instance having to abide by rules that they find unconscionable. The right of exit protects both the positive and the negative facet of these freedoms, the freedom to and the freedom from.

Whether the right of exit is sufficient to adequately protect these freedoms is up for discussion. There has been a host of criticisms against the emphasis on exit rights (Hill, 2016; Mahoney, 1992; Okin, 2002; Shachar, 2001; Walzer, 1998). Critics have pointed to the gap between the liberal vision of community belonging as consensual and the many reasons why exit can be out of reach, or costly, especially for “minorities within minorities” (Eisenberg and Spinner-Halev, 2004). For many individuals living in communities in which they have been raised, there is no clear exit door. Because of their socialization, they might not even contemplate the possibility of an alternative life. And should they want to leave, they would suffer all sorts of material and immaterial costs, such as the difficulty to find a job and the loss of meaningful social ties. In some cases, it seems that state's effort should be directed at giving individuals the opportunity to remain in their community under fairer conditions (Okin, 2002; Shachar, 2001).

In response, proposals have been made to alleviate some exit costs. Jeff Spinner-Halev argued that communities relying on collective ownership and communal living should develop an “exit fund” that provides financial help to members who choose to leave (Spinner-Halev, 2000). In Spinner-Halev's proposal, the burden is on communities to make exit accessible. The state coerces them to fulfill their duty vis-à-vis dissenters—bearing in mind that religious communities relying on collective ownership are scarce. In contrast, Robert Taylor took a more state-centered approach to examine how exit can be facilitated by legal reforms and policy provisions (Taylor, 2017). Taylor was concerned not only with religious or cultural communities, but also with intimate associations such as marriages, and economic associations, such as business firms. This translated into practical recommendations to help more vulnerable parties, for instance by providing restraining orders, shelters, and job opportunities to victims of abusive marriage. Enabling exit was presented as a tool against domination directed at the most vulnerable members of society.

Both critics and proponents of this understanding of freedom of association have been concerned with the “most vulnerable”—and to a certain extent, rightly so. The liberal view of community belonging “gives a great deal of authority to cultural communities” (Kukathas, 1992). To keep this authority in check, individuals should have substantive exit opportunities. The most vulnerable are more likely to suffer from the existence of high exit costs, whereas powerful members are more likely to be able to afford these costs. Therefore, Taylor's insistence on low exit costs is directed at the protection of abused wives, exploited workers, and oppressed citizens. ⁶ Taylor argued that a partial alleviation of exit costs may occur at the detriment of the least advantaged, as only the most privileged will be able to enjoy the opportunity to leave. However, if the costs continue to diminish, the most vulnerable will eventually benefit from it. To take the example of marriage, the formal right of divorce can initially benefit powerful husbands who want to leave their wives, but if substantive opportunities are added to the table, they will benefit the least well-off who can either leave or at least credibly threaten to do so, and enhance their bargaining power within the relationship.

Although the most vulnerable might ultimately benefit from low exit costs, there is something misleading in framing the right of exit as a mere instrument to protect the most vulnerable. Recall that the raison d’être of the right of exit is not merely the protection against undue interference or domination, but also, more positively, the fulfillment of positive freedoms realized by the association. Yet, the most vulnerable have been the focal point because laissez-faire policies leave them deprived of substantive exit rights. Their position of vulnerability gets in the way of their ability to leave, and as some authors suggested, policy provisions could help make exit less costly (Taylor, 2017). All members of an association, qua members, hold a formal right to exit, but it is only the most vulnerable who would require intervention.

It is perfectly reasonable to argue that the right of exit should provide adequate protection to vulnerable members of groups who cannot afford to pay the price of exit. ⁷ Yet, this view is too narrow. Labor laws surrounding dismissal and resignation should not only target cases of slavery or exploitation, but ought to protect any employee. Divorce laws should not only be helpful to victims of domestic abuse, but to any spouse. Freedom of dissociation should not be construed merely as protecting freedom from abuse, exploitation, and so on. The right of exit secures essential associational interests and positive freedoms. It is not merely “a right for the vulnerable.” In what follows, I propose a more robust account of freedom of dissociation, which concerns both the most vulnerable and the most powerful—from the desperate divorced mother to the wealthy fired CEO.

Exit ought to be fair

Freedom of dissociation matters, and framing it too narrowly as a right to substantive exit opportunities is doing it a disservice. I revise the standard view along two dimensions. First, I contend that exit should not be merely about excessive costs, but also about undue costs. Broadly speaking, an exit should not only ensure substantive exit opportunities, but also allocate the burdens of exit fairly among the parties. Second, the right to a fair exit should not be narrowly defined around “voluntary exit”—involuntary exit also ought to be fair. I reframe the right of exit into a “right to a fair exit,” whether chosen or not.

Let us begin with the first move: exit ought to be fair. Freedom of dissociation should not be merely about avoiding excessive costs, which are likely to make exit prohibitively expensive for leavers, but also about the fairness of the allocation of the burdens of exit. Excessive costs can include (a) material costs that are disproportional to the financial situation of the leaver (e.g., a single mother paying unreasonable legal fees; a spouse having to leave their house behind without adequate compensation); (b) immaterial costs that significantly burden exit (e.g., the prospect of potentially losing custody of one's child; an overly humiliating dismissal). Undue costs can include (a) material costs that are disproportional to the contribution that the party has made to the association during their involvement (e.g., a stay-at-home parent not receiving any alimony); (b) immaterial costs that are inappropriate (e.g., losing custody of one's children based on a wrong accusation of being an unfit mother; suffering reputational damage while being fired without proper justification). In practice, excessive and undue costs may overlap, but the distinction lies in whether it is about expensiveness or unfairness.

It goes without saying that an exit cannot be cost-free; dismissals are likely to be psychologically heavy and economically costly in a context of mass unemployment. Yet, the significance of these costs ought to be assessed in terms of how they unduly burden the dissociating parties. Focusing only on low exit costs misses an important point about dissociation. Many associations involve collective assets, shared responsibilities or the exchange of goods. Dissociation is not only about individuals being able to physically leave an association without excessive costs (although this is one important element of it), but about ending an association with a fair exit deal, that is, a deal whose costs are allocated in a way that does not unduly burden one of the parties.

An exit can be reasonably priced, but unfair. Consider the case of a parent who is sufficiently well-off to leave the household, but yet is declined any financial contribution from the other parent for child raising duties. The additional costs of exit are not excessive, but there is something potentially unfair in the exit deal, which makes it unduly costly. Now, recall Granju, the divorcing mother deemed “unfit.” She might have been well-off and able to afford the colossal legal fees and the parenting classes; the question of the fairness of these costs remains open. Similarly, consider the case of a monk who has worked diligently for twenty years in a monastery, but does not receive any compensation upon leaving, not even a penny. While participation to the communal work was fully voluntary and based on a spirit of unconditional giving, the former monk argues that the significance of his contribution to the collective life of the monastery over the years should give him the right to some form of compensation. The question of whether the monk is able to make a living elsewhere is distinct from the question of whether the community owes something to the monk in light of his long-term services. The exit might be easy, and not too costly, thanks to the help of generous family members, but its fairness is up for discussion.

Exit can arguably be unfair also from the standpoint of the stronger party, when there is one. Consider the case of an employee resigning from a firm as soon as the training period is over. This sudden departure is costly for the institution. The firm has invested a lot in training them before the employee was able to do any productive work. The firm is therefore losing a lot from their departure, while the former trainee benefited from a paid training. Yet, it is generally not expected from former employees that they compensate their employer for the cost of their training period. This may be because the employee's interest in an affordable exit outweighs the unfair burden put on the business. The institution has a duty to let the individual go, even if it entails paying an unfair burden for it. Associations, and stronger parties, very much have an interest in a fair dissociation too, but this interest is outweighed by the individual's right to an affordable exit. If it was requested from flaky employees that they reimburse their training costs, their exit opportunities may be undermined. This example highlights again the relevance of the distinction between affordability and fairness when it comes to exit, and potentially trade-offs between the two: a fair exit for one of the parties may in some case undermine the affordability of exit for the poorer party.

The protective function of the right of exit seems to imply that fairness implies affordable costs for the weaker party. An unreasonably expensive exit cannot possibly be fair. ⁸ Affordability is a necessary, but insufficient condition of fairness. My view thus supplements rather than contradicts previous views focused on lowering exit costs. I reiterate that substantive exit opportunities are paramount, but I add that fairness also deserves attention.

Excessive and undue costs can interact in different ways. On the one hand, a fair exit deal can make leaving more costly for one of the parties. Striving for a fair distribution of assets and arrangement about child custody through a lengthy divorce burdens exit in the short-run in order to ensure a fairer long-term arrangement. Perhaps the ordeal that Granju went through during her divorce has ultimately paved the way for an equitable deal. On the other hand, the possibility of a fair exit can make exit a more attractive option, and consequently increase its likelihood. Some members may indeed renounce ending an association on the expectation that they will have to pay undue costs upon leaving.

To recap, I contend that substantive freedom of dissociation requires exit opportunities to be substantive, but also sensitive to considerations of fairness, which means that leavers should not pay excessive and undue costs. My account is therefore more demanding that existing accounts, as I am not merely concerned with maximizing exit opportunities for the vulnerable party, but also with rightly allocating the burdens of exit.

The fairness of exit: two desiderata

The question that naturally ensues is which conception of fairness applies to exit. My account is compatible with a range of conceptions. To this end, a fair exit is defined in a very minimal way, as not imposing excessive and undue costs to the leaver. By focusing on eliminating costs that are inappropriate or disproportionate, it concentrates on what Shklar calls “damage control” (Shklar, 1989: 27) and on securing the “conditions that are necessary for the exercise of personal freedom” (Shklar, 1989: 21). It also takes as “basic units of political life” “the weak and the powerful” (Shklar, 1989: 27), although it argues that the right to exit should not merely be a protection for leavers otherwise likely to suffer absolute deprivation. Should this account be specified further with a more demanding conception of fairness, it could become more ambitious and involve a maximalist version of liberalism. For now, I primarily want to make the general claim that an account of dissociation should be revised to become fairness-sensitive. More context-specific accounts should be developed to work out the right procedures and criteria of the fair allocation of exit costs. ⁹

In what follows, I formulate two important desiderata on what such accounts should include. The first is that assessments of fairness should not be made solely on the basis of standards which are internal to the associations. The second desideratum is that assessments of fairness should cover both some material and immaterial costs.

Determining the exit deal is primarily a matter of the internal business of the dissociating parties—divorcing parents striving to compromise over child custody; an employment contract establishing some exit terms; a religious organization having procedures in place for when a member breaks their vows. Questions arise, such as: what was initially agreed, if anything? What and how much do dissociating parties owe to each other now? How to divide collective assets, how to compensate for past contributions? Is the exit procedurally fair, and is the exit deal equitable to the parties involved? Exit is costly. It can be costly for both the leaver and the association being left and it is arguably first and foremost the responsibility of the dissociating parties to allocate the burdens of exit in the fairest possible way.

Letting parties negotiate their own dissociation seems in line with associative freedom. There are significant differences between the purpose of an intimate association like marriage, collective associations, such as religious groups or business corporations. Intimate associations have been described as “purposeless”—existing “for their own sake” and “comprehensive”—“marked by emotional and material commitment” and central to one's life (Brownlee, 2015). In contrast, many business corporations have a clear purpose—profit maximization—and are only meant to be a transactional activity. Religious associations can oscillate between loose networks and tight communities, and members be irregular attendees or fully committed participants (Laborde, 2017).

The self-understanding of an association very much determines how exit is construed and how the exit deal is negotiated. In more transactional cases, exit terms can be negotiated in advance and may reflect some kind of risk assessment. For example, a CEO risks being fired at any moment, without warning, depending on results. Unlike the average employee, their firing is very public, and this is what partly justifies a high compensation. In the marriage case, the commitment of the spouse comes with a bundle of rights and duties, to which a contract can be added (Chambers, 2017). Religious groups are a harder case, because they do not typically include a “contract-like” moment and encompass a wide diversity of situations. It may seem odd that the spiritually-motived participation of a monk in a monastery could be scrutinized with the same lens as the financially-motivated participation of CEO employed in a multinational firm. If unconditional giving is part and parcel of the doctrine of the monastery, why require fairness in exit? Isn’t it undermining associational freedom to require some standards of justice to be applied in the exit of various voluntary associations?

Associative freedom is subject to many constraints, among which the consent of members (Brownlee, 2015). If we take the right of exit to be paramount for the protection of individuals against excessive costs, we ought to also protect them against undue costs. And merely applying the internal standards of the association will not do, as many associations do not have an interest in leaving exit unobstructed, even less in making it fair. The right to a fair exit is likely to collide with internal standards of the association and this is unavoidable—as it protects individuals from the group to which they belong. The implication is rather straightforward: state policy should not merely aim at resourcing exit, but also at monitoring its fairness.

This does not entail that the association, or the state should always bear the lion's share of exit costs, that individuals should be exempted from all costs or that all exits should be state-monitored. It seems reasonable that someone who entered voluntarily an association with full knowledge of the costs they will incur upon leaving should pay at least some of these costs. In that case, the fair share might be close to the affordable share—the costs should not render exit inaccessible, but the association or the state ought not to subsidize all the exit costs. Voluntariness of entry gives a pro tanto reason for the individual to bear at least part of the burdens of exit.

An agent may sometimes waive their right to be protected from certain exit costs, like in prenuptial agreements. These are written contracts that establish some financial or property-related arrangements in the case of divorce or death of one of the parties. Given the high rate of divorce, some couples want to mitigate the costs of a potential separation by formalizing its financial modalities ahead of time. In many cases, it entails that one of the parties—typically the poorer party—renounce some exit benefits. For example, a business owner might want to exclude their property from the principle of equitable distribution in case of a divorce with their spouse. A wealthy homeowner might want to privilege their children from a previous marriage rather than their spouse, should the marriage end.

Why would a poorer party relinquish exit benefits? A plausible explanation goes as follows: while those with more to lose do have a legitimate interest in avoiding predation, a poorer party does have an interest in signaling that they are sincere and are willing to accept a form of sacrifice to prove it. One might think that poorer parties might simply have less access to richer mates if they could not signal that they were interested in the person rather than the wealth.

There have been some warnings that prenuptial agreements largely benefit the wealthier party. Sam Margulies argued that, most often, prenuptial agreements are sought by the economically stronger party as a way to avoid a loss of assets and income in case the marriage fail, putting the weaker party in a difficult decision-making position (Margulies, 2003). Some states in the Unites States have even put limitations on prenuptial agreements to ensure that the poorer party does not become in need of public assistance. ¹⁰ Although there might be good individual reasons to waive one's rights of exit with a prenuptial agreement, there are equally good reasons to be concerned with how it potentially undermines both the affordability and fairness of exit.

A policy reinforcing the fairness of exits is liable to have the counterproductive effect of dissuading certain relationships. A fair exit is likely to be more costly for stronger parties, such as employers. This increased cost may have undesirable consequences in terms of relationship formation. For example, employers may be dissuaded to hire new employees if labor laws are too protective. Employers may be incentivized to multiply precarious employment to avoid the costs of ending so-called “permanent” contracts. This, in turn, is not in the interest of more vulnerable parties. Weaker parties are dependent upon the creation of positions by stronger parties. Why demand conditions that make one's position worse overall to secure a better exit?

A fair exit policy cannot overlook the potential counterproductive effects of monitoring and enforcing fairness. The concern for allowing relationships to freely flourish and proliferate conflicts with the concern for reducing excessive and undue costs. In the same way that a fair exit should not make exit unaffordable, it also should not make forming relationships too risky or unrealistic. If the anticipation of paying high exit costs would act as a deterrent, society as a whole would suffer from the fair exit policy, as it would lead to an increased scarcity of associative opportunities.

A possible answer to this conundrum lies in the distribution of exit costs. The allocation of the burdens of exit should not rely solely on principled concerns—what should be an equitable distribution among parties—but also on principled and strategic considerations—will exit be feasible for the more vulnerable party? Will new relationships be realistic and sufficiently incentivized? This entails that the decision-making process does not only take into account the interests of direct stakeholders, but also the collective interests of future potential associates. To achieve this, it might be necessary to involve third-parties—such as taxpayers—to mitigate the increased costs paid by the strong party and correct the potential counterproductive effects of a fair exit policy. ¹¹

There cannot be one formula that determines the right distribution of exit costs in all relevant cases of voluntary associations—from marriage to employment. The allocation of exit costs should be context-sensitive. The procedure will vary from the verdict of a judge to a multiparty negotiation involving trade unions. The allocation of exit costs might involve third-parties. The negotiation of the exit deal should ultimately aim to avoid both excessive and undue costs, while taking into consideration the potential undesirable side effects of a fair exit policy. How this is implemented is a matter of good governance.

In addition, the state ought not to monitor all cases of exit from voluntary associations. I endorse Chiara Cordelli's approach which consists in identifying some properties of civil associations that trigger demands of egalitarian justice (Cordelli, 2016, 2017). Within relevant voluntary associations, some types of association might just not trigger the same demands than others. Someone who is simply attending church services without substantial material relationship with the association, with the exception of an occasional small donation, will not have a claim for a state-enforced fair exit. Not all relationships within a paradigmatic voluntary association, such as a religious association, trigger demands of justice.

Similarly, not all voluntary associations are eligible for a state-monitored exit. Consider the example of an intimate association like friendship. Ending a friendship can be costly emotionally, but also financially, if you have generously lent a sum of money to your friend in need. Upon falling out, you might want to recuperate this money. Your friend might have a moral duty to give you back the money, but there is little you can do to make this duty legally enforceable. In practice, this has to do with the fact that the state does not organize friendship arrangements the same way that it intervenes in the marriage or the labor market. The level of intervention in the religious market is highly context-dependent and up for debate.

My theory of dissociation as fair exit is congruent with existing legal practices. For instance, legal arrangements surrounding labor do not aim solely at preventing excessive exit costs—by providing a welfare minimum—but also include elements of fairness—for instance, by enforcing a mandatory insurance scheme that tracks the employee's wage. If the focus was only on excessive costs, then a minimal cover of living expenses would be enough, not an income proportional to one's contribution. When the state actively regulates a dissociation, it ensures that it follows a due process, procedurally fair for the affected party. This is the case for divorce policy, which determines the distribution of assets, arbitrates disputes over child custody, and if applicable, determines an alimony for one of the spouses. Enforcing the fairness of exit requires some level of intrusion in associations and a potential clash with their internal standards.

The second desideratum is that assessments of fairness should not consider only material costs, but extend to some immaterial ones as well. Recall Fiorina, the fired CEO, who judged her situation to be “unfair.” Clearly, she was not referring to the material side of the story, given the tremendously high compensation she received. She referred instead to the way she was treated—in a humiliating, potentially disrespectful way. In common language, we routinely use the vocabulary of unfairness to capture instances of humiliation and disrespect (“It's so unfair!”). In doing so, we are being analytically sloppy. Jonathan Wolff fruitfully distinguished between fairness—“the demand that no one should be advantaged or disadvantaged by arbitrary factors” (Wolff, 1998)—and the respect we owe to one another. ¹² Wolff further argued that the two can come into conflict, and especially that the pursuit of fairness can lead to instances of disrespect—for instance when welfare claimants are required to reveal private information about themselves. Despite the good analytical reasons to distinguish fairness and respect, I want to side here with common sense wisdom and argue that the immaterial costs of exit (such as the humiliation of being fired) can also count as undue costs and warrant the vocabulary of unfairness. Instances of disrespect and subsequent humiliation can weigh heavily on the burdens of exit, adding to the material costs. This creates situations in which individuals are unduly burdened with inappropriate and potentially disproportionate costs. Immaterial costs, such as the consequences of being disrespected, do not necessarily weigh on the possibility of exit, but can create unjustified costs.

It is wrong to argue that immaterial costs are largely beyond state intervention. Brian Barry distinguished between different types of costs associated with exiting, ranging from costs that are inevitable and beyond the scope of state action (“intrinsic costs”) to undue costs that must be alleviated or compensated by state action (“external costs”) (Barry, 2001). Examples of the former include the excruciating pain of losing a community of belonging. Examples of the latter include losing one's job after having been excommunicated. In practice, intrinsic and external costs can be hard to disentangle. If the distinction lies between psychological and material costs, then psychological costs are not disconnected from material conditions. If the distinction lies between what the state cannot control without undue interference, or does not have the means to control (e.g., preventing the break of social ties) versus what the state can control, then it overlooks the possibility of receiving compensation for psychological harm.

To be clear, the psychological harm of leaving an association does not warrant compensation in itself. It is plausible to argue that we should be held emotionally responsible for our associational attachments. When someone leaves their church after a crisis of faith, they should not be compensated for the psychological pain that ensues. They should take responsibility for their emotional hardship, as they aware that entering an association carries with it the risk of ending it.

What might warrant compensation, however, is the moral damage resulting from an unfair exit process. Suppose that after leaving their religious community, a dissident is publicly presented as sinful, unreliable, hysterical, in a way that significantly damages their reputation and constitutes defamation. This person might want to claim that they should be compensated for psychological harm. The psychological harm does not stem for the dissociation itself, but from a wrongdoing committed during the process of dissociation. To elaborate on the divorce example, Granju should not be compensated for the emotional cost of her divorce in itself, but she might deserve compensation for having been wrongly accused of being “an unfit mother” during the process (providing this claim is false and constitutes defamation).

If it is demonstrated that the material and immaterial burdens of exit were significant enough to cause severe psychological distress, then the fairness of exit should be called into question. An exit ought to be affordable and fair, meaning it should not involve undue material and immaterial costs.

A fair exit, whether chosen or not

Once I have reformulated the right of exit into a right to a fair exit, a second move is in order: exit ought to be fair, whether it is chosen or not. The narrow focus on “voluntary exit” echoes the liberal concern to make exit a safeguard against some pathologies of associational life. For theorists of exit, exit is normatively relevant when it is voluntary, as it expresses and protects dissidence (Green, 1998). Exit is the weapon of the weak, the ultimate way to escape oppression. In this view, state policy aiming at resourcing exit can be a critical tool against domination (Taylor, 2017).

A standard objection against my view is that it undermines personal responsibility and risk-taking. Considering the case of religious association, a distinction could be drawn between people who are raised within a religious community, and those who join as adults. While a right of exit would be essential for those who have not initially elected to be where they are, it does not apply in the same way to those who joined as adults. Adults are in a position to decide whether they want to commit to a faith, and self-bind in order to resist the temptation to leave in the future. They are free to enter on terms that make future exit costly. As long as the initial decision was made autonomously, and standard prohibitions of coercion are taken, there is little more the state should do to facilitate exit.

My response to the standard objection is twofold. First, I argue that my emphasis on fairness is in fact less liable to the personal responsibility objection. A fair exit allocates the burdens of exit among relevant stakeholders, it does not merely alleviate the costs for the leaver. In the example above of unemployment benefits, an important part of the mandatory contribution is typically paid by the employee; what the state does is making the contribution mandatory. Making exit fairer does not make it less costly, as exemplified by lengthy divorces.

Second, judgments on the voluntariness of entry and of exit can be a bit swift. A religious community may give plenty of opportunities to its members to express and renew their consent, while an adult member may be “indoctrinated” into a cult. Similarly, exit can occur in many different ways. Members can of course be expelled, they can be made redundant from their spiritual mission or excommunicated. A group can be partially dissolved. Some individuals can leave reluctantly, because they made decisions at odds with the norms of the group. A member can resign herself to leave because she wants to get married with an outsider. While she wishes to remain, her failure to conform to one of the group's fundamental norms places her as a de facto outsider. Further, deliberate acts of dissent can lead to exclusion. Consider the case of Robert Bear who was a Mennonite Dissident in the 1970s and who had publicly denounced his church doctrine and hierarchy. ¹³ Bear was subsequently excommunicated by his church and shunned by his wife and six children. Here, voice was a preliminary to exclusion, or a first step towards exit. Bear was most likely aware that the expression of his criticisms would place him at the margins of his community. That was at least the view of one of the Bishops, who declared, with or without a touch of irony: “We believe he excommunicated himself.” ¹⁴ The frontier between voluntary exit and expulsion can be blurred.

A theory of dissociation that narrowly focuses on chosen exit leaves aside cases that would be fruitfully examined with the same lens, given the difficulty to determine in many cases whether someone left deliberately—as the expression “excommunicating oneself” vividly illustrates.

Whether an exit is chosen or not might be morally relevant in some cases to determine the fair exit deal—for example, when a dismissal generates psychological harm. In some cases, it might be justified and necessary that the circumstance of exit be properly investigated to assess its fairness. In other cases, not wanting to differentiate between chosen and unchosen exit amounts to a question of respect, and relates to an argument formulated by Wolff about “shameful revelation” (Wolff, 1998). Gathering information to determine whether an exit is chosen or not is likely to have the unfortunate consequence of forcing individual leavers to reveal too much of themselves—this is especially true for marriage and religious associations. If we adopt a conception of fairness which implies to determine how much of exit was the result of unchosen circumstances, we run the risk to make the process of exit even more disrespectful—and in my own conception, even more unfair.

What I resist is a general account of exit that from the outset narrowly focuses on chosen exit. Individuals have a right to a fair exit, whether chosen or not. Yet, the (in)voluntariness of exit may raise specific moral considerations. This is to be further explored in light of applied conceptions of procedural and substantive fairness.

Conclusion

My new account of dissociation thus revises the right of exit in two distinct ways. First, exit becomes accountable to considerations of fairness. Second, chosen and unchosen exits are scrutinized with the same normative lens. To conclude, I wish to come back to the two cases highlighted in introduction and to exemplify how this change of frame sheds some new light on real-life cases.

Recall Fiorina, who was fired with a $40 million golden parachute, and yet humiliated by the way she was treated by the Board. Let us, for the sake of the argument, take her at her words and assume that her distress was not only real, but justified by something akin to a lack of respect on the part of the board. The way her case was handled showed not only a lack of recognition for the work she had diligently accomplished for the company, but also for who she was as a person worthy of respect. Fiorina was not treated as someone who is owed a justification for a decision that will deeply affect her. There might be good reasons for her dismissal, but those have not been properly formulated to her. This lack of respect is part of her exit deal; an incredibly high financial compensation which does not suffice to soothe the bitterness of the humiliation she suffered.

If freedom of dissociation is approached only through the lens of the traditional right of exit, and the worry around excessive costs, this case is unproblematic. A case could even be made for excessive benefits for the leaver. Fiorina has more than enough to survive and thrive a lifetime on the compensation she received. There are no excessive costs of exit, and therefore her right of exit is adequately ensured by private means.

Considering this case in light of “the right to a fair exit” sheds light on the fact that Fiorina *might* have a claim for a procedurally unfair dismissal on grounds of disrespect, in spite of her having received disproportionate compensation. Whether she should receive additional compensation is up for debate, but the point has been made that her material compensation did not suffice to make the exit fair—it only made exit largely affordable.

As far as Granju is concerned, the unfortunate mother deemed “unfit”, the account shows that both her material and immaterial burdens (legal fees, psychological distress) should weigh on the assessment of the fairness of her exit deal. The story does not say whether she left or was left by her husband; she ought to have a fair exit either way.