The New Marriage Contract and the Limits of Private Ordering(1)†
Gregory S. Alexander(2)
Marriage isn't what it used to be. Traditionally considered a status-based institution, marriage is now increasingly a matter of contract. More and more American couples are executing premarital contracts that expressly define various legal aspects of their marriage, including their obligations to each other during marriage and upon divorce. While in the past it was largely a tool of the rich and famous, the premarital agreement has now become attractive to a much wider spectrum of married couples.(3) The rising divorce rate, combined with expansion in the definition of marital property subject to equitable distribution upon divorce, are among the major factors fueling the growing interest in marriage contracts.
To date, the shift from status to contract is far from complete. As Professors Rasmusen and Stake note in their very interesting and stimulating paper, courts do not always enforce marriage contracts, and the bases on which they review such contracts is not entirely clear. The result is that the legal status of the extant type of marriage contract is somewhat uneasy.(4) Moreover, while more couples are using marriage contracts, most newlyweds, especially first-timers, still rely on the default rules of marriage and divorce that the legal system provides. Still, things are not as they once were. As recently as 1977, Judge Posner could accurately state that "explicit marital contracts are rare."(5) That statement can no longer be made. Marriage contracts may not yet be commonplace, but they certainly are no longer a rarity.(6) Once the province of the rich and beautiful (or at least the rich), marriage contracts have moved into the middle class in recent years.
The thrust of Professors Rasmusen and Stake's paper is to push the inside of the marital contract envelope even further. Currently, prenuptial agreements are primarily devoted to sorting out the property aspects of couples' obligations to each other, especially the division of assets upon dissolution, either by divorce or by death. Rasmusen and Stake propose extending the use of contract to cover a much wider range of issues than the current practice contemplates. They want couples to be able to define by contract not only the terms of divorce but the grounds for divorce as well.(7) Even more striking perhaps, they propose extending the province of contract to include conduct during the marriage.
What motivates the movement to expand the permissible scope of marriage contracts is not knee-jerk contractarian or libertarian ideology. Rather, the advocates of covenant marriage want to use freedom of contract to enhance security of contract in the context of marriage. Stated differently, they want to throw off the traditional limits of private ordering in marriage as a means of returning to traditional marriage. That tradition, they believe, protected not only the stability of marriages but also the more dependent spouse better than the current legal regime does. The current regime, which I will call "NF/ED," is that of no-fault grounds for divorce and equitable distribution of property upon divorce. Under this regime, a spouse who is, relatively speaking, less committed to the marriage and wealthier has strong incentives to opt out of the marriage than he (because the exiting party usually is the man) had under the traditional regime, the regime of fault-based divorce and alimony upon divorce (I will call this traditional regime "FAL"). This is because NF/ED in general imposes fewer financial burdens on the wealthier ex-spouse than did FAL and because the exit costs are lower under NF/ED.(8) Hence, the obvious solution is to raise the costs of exit, or at least permit consenting parties to do so by entering into contracts for covenant marriage.(9)
Without defending the extant NF/ED regime, I want to offer two reasons for being skeptical about this proposal.(11) First, marriage is not the type of relationship in which precommitment is most likely to maximize long-term preferences. Second, the new restrictive marriage contract represents a very significant expansion of the domain of private ordering by including within it an institution--marriage--that historically has been viewed as a matter of public regulation because of its strong implications for the entire society. The defects in the current regime of NF/ED do not warrant such a fundamental change in the relation between society and the institution of marriage.
Professors Rasmusen and Stake model marriage in market terms, treating it as a deal between two rational agents. Their goal is to maximize the aggregate preference satisfaction of the two agents. From that perspective, it is easy to see why marriage law looks anomalous. Marriage law should be like partnership law, a matter of private agreement. That it is not is the mistake that they seek to correct. Their solution is to allow H and W to precommit themselves to each other in a way that NF/ED marriage law currently does not.(12) Put in terms first coined by the economist Albert Hirschman, they ought to be able to contractually eliminate the exit option, or at least greatly restrict its availability, so that only two options remain to each of them--loyalty, that is, stick with the other, and voice, that is, stick with the other but express dissatisfaction to her or him and try to work things out. Where there is dissatisfaction on the part of one or both parties, removing the exit option, as Hirschman's book generally argued,(13) increases the likelihood that the voice option will be taken up effectively. Recognizing this dynamic regarding the relationship among exit, voice, and loyalty, the couple could rationally choose to forego (or severely limit the availability of) the exit option, if only the law would allow them to.
It is not always the case that the unavailability or unattractiveness of the exit option will lead either to greater loyalty or more effective use of the voice option. There are many circumstances in organizations in which exit is, practically speaking, unavailable (e.g., high exit costs), but the result is neither to mitigate dissatisfaction nor to produce greater efforts to remedy the reasons for the dissatisfaction. There is such a thing as suffering in silence, as Hirschman himself recognized. Voice is not costless, and in some organizational settings the costs can be quite high. I have no data on this, but intuitively and anecdotally it seems likely that this was the case in more than a trivial number of marriages under the old regime of FAL.
Moreover, recognizing that the relationship with which we are dealing is a long-term relationship, one is likely to ask, what happens if, down the road, one or the other spouse does become dissatisfied with the partner, tries the voice option, and it does not work? The partner is unwilling or unable to change enough so that remaining in the relationship continues to satisfy both parties. As one would expect, Rasmusen and Stake anticipate this concern, and answer it with the following statement: "Requiring the traditional grounds for divorce would result in more unhappy couples staying together. The unhappiness of the couple, however, is not enough reason to prevent them from binding themselves, unless a case for paternalism can be made."(14) Not surprisingly, they can find no credible case for paternalism in this context. The basic idea is that people are the best judges of what is in their self-interest, and if they believe that binding themselves into the future will maximize their individual preferences, they ought to be free to do so.
Now, I have to confess, I am at a bit of a loss to understand this statement. Rasmusen and Stake appear to be treating the marriage contract as a precommitment device designed to overcome problems of rationality.(15) But the problem with which they are dealing is significantly different from the usual case in which precommitment is rational. The classic instance is that of akrasia, or weakness of will. If I am weak-willed about getting out of bed in the morning and I am aware of this, acting rationally, I will put my alarm clock on the other side of the room from my bed. This is, of course, the familiar "Ulysses" strategy: the individual, recognizing that she might later change her mind, binds herself to the course that she initially prefers but would later not prefer and so achieves consistency. She precommits herself to the initial preference because it maximizes her preferences over time. The later, inconsistent preference --sleeping in--derives from weakness of will, and if acted upon, is one that she knows she would later regret.
In the marriage setting, the difference in preferences--at Time 1, both wanted to be married to each for life; at Time 2, one or both want out--is not the result of weakness of will or of discounting the future. Rather, it results either from endogenous preference changes which imply some change of character, or from exogenous factors such as changes in the mate's character, or other external factors. Regret should not always be a basis for avoiding contractual obligations, but sometimes it should. That is especially so where there were substantial and unforeseeable errors in information or other related cognitive defects in the agreement. Indeed, interfering with consumption choices under these circumstances is not a straightforward rejection of private preference at all. If a contracting party lacks relevant information and could not reasonably have obtained that information, a legal decision to override the choice is not interference with liberty, at least not obviously so.
Public-choice considerations aside, one way of explaining why the no-fault revolution occurred in the first place is that society recognized that the risk of high information costs and cognitive errors is especially high in the marriage context. In no other relationship is information about the other party's character more important and more difficult to obtain. Married individuals not only seldom fully know their mate's character at the time they marry, but they cannot possibly know what their mate's character will be like in the future. A significant percentage of divorces occur during the first seven years of marriage.(16) Since youth at the time of marriage is positively correlated with divorce, it is unsurprising that the parties to most of these divorces are young.(17) The plausible inference is that most of these divorces were the result of inadequate information going into the marriage. Under these circumstances, where the risk of error is substantial, it is not at all irrational to make the exit option readily available. Nor is it incompatible with a general concern for individual preference satisfaction to say that precommitment devices in this context should not be binding.
The most appealing argument in favor of precommitment marriage contracts is that they would produce more egalitarian marriages by evening the bargaining positions between husbands and wives. On average, men and women share unequally the benefits and burdens of marriage. Men tend to have greater bargaining power and use that power advantage to satisfy their preferences at the expense of their wives.(18) The most plausible theory explaining this phenomenon is that the threat of exit from the marriage is greater for men than it is for women on average since men tend to have more appealing options outside of marriage than do women.(19) Precommitment marriage contracts, the argument goes, tend to equalize husbands' and wives' bargaining power by reducing the credibility of the husband's exit threat. One problem with this argument is that while the contract may reduce the husband's exit threat, exit is not the only source of bargaining inequality in marriages. Husbands can and sometimes do rely on noncooperative forms of conduct during the marriage as an alternative means of getting their way. Theoretically, a marriage contract could neutralize this source of bargaining inequality as well by specifying unacceptable forms of conduct by each party. Indeed, this is part of why Professors Rasmusen and Stake argue in favor of permitting marriage contracts that regulate intramarital conduct. But marriage is too unpredictable and too complex a relationship to allow parties to draft a long-term contract that would regulate all of the relevant points of conduct and specify meaningful standards of performance. A conduct-regulating marriage contract would inevitably be so open-ended that it would fail to preclude opportunistic conduct by either party.
A second reason why such contracts are unlikely to succeed in creating more egalitarian relationships between husbands and wives is that men who in fact have greater bargaining power are the very men least likely to agree to such a contract. Faced with the demand to agree to a contract restricting exit or regulating intramarital conduct, such men are most likely to look for a different, more "pliable" partner. Only if such contracts became a near-universal practice--a very unlikely development--would this option be unavailable.
Finally, my guess is that, given the notorious assumption of newlyweds that their marriage will be among the fifty percent of marriages that work, most people will not use the new marriage contract. This will be especially true if, as seems likely, the man is the more opportunistic and therefore disinclined to enter into such a contract, and the woman lacks either information about their availability or wants to avoid derailing the marriage by insisting that her prospective partner enter into a contract severely limiting both his exit and intramarital options. Intended as a tool to benefit women with weaker bargaining positions, these contracts would, as the law of unintended consequences predicts, produce perverse results.
Thus far, my discussion has engaged Professors Rasmusen and Stake on their own grounds. That is, I have remained within the framework of rational-choice theory in responding to arguments in favor of their new marriage contract. I now want to approach their paper from a different perspective, one that argues that maximizing the satisfaction of a couple's aggregate preferences should not be the sole criterion by which we evaluate the desirability of the new form of marriage contract. Marriage is not strictly a private matter, and recent proposals for new types of marriage contracts expand the boundaries of private ordering in ways that fail to respect the strong and legitimate interest that society as a whole has in the regulation of marriage.
I begin with an extreme example. Suppose, as implausible as it may seem, that a couple has with genuinely mutual consent entered into a marriage contract allowing the husband to beat his wife if she serves dinner late. I would hope that we could all agree that this contract should never be enforceable, no matter how genuinely consensual the agreement and no matter what compensation or consideration is provided to the wife. Such a contractual term should be per se unenforceable for the obvious reason that it blatantly and grossly offends the widely shared social values of human dignity, respect, and equality. The point of this example is to emphasize what I had thought was common ground: that there are many spheres of social activity in which the domain of private ordering is properly narrowed by shared social values; and that marriage is one of these areas. Professors Rasmusen and Stake's article, together with recent legislation permitting precommitment marriage contracts,(20) indicate that this shared commitment regarding the limits of private ordering is no longer common ground. Before we release the restrictions on the domains of private ordering, though, we ought to acknowledge that we are giving up. Marriage is not like widgets. It is not a strictly private affair, and neither are marriage contracts.
Historically, very little, if anything within marriage, was left to private ordering because marriage was considered a sacred rather than a secular institution. Marriage was a sacrament according to Christian doctrine, for example. That perception has changed, of course, but vestiges of it remain. Though largely secularized now, marriage remains a quasi-public institution, a relationship created through official state sanction rather than through private agreement alone. An institution is a legitimate public institution when it directly implicates public interests or shared public values. For all of its changes, that remains true of marriage not only because of the obvious public interest with respect to children but also because marriage, for better or worse, remains the primary foundation of the family, which is the foundational unit of our society's structure.
The public aspect of marriage does not rule out all forms of private control over the terms of marriage. Some forms of marriage contracts are permitted today, and properly so. By and large, however, these are limited to issues concerning the parties' property obligations to each other upon dissolution. Those obligations intrude far less on public interests and values than would the kind of marriage contract that Rasmusen and Stake's proposal contemplates, and even then, courts often regulate marital property arrangements to ensure that they do not violate the legitimate public interests in the welfare of the two parties, especially the one having a weaker bargaining position. The courts' wariness about these contracts does not bode well for the new marriage contract, which takes private ordering to a different plane.
My point is not to defend the extant NF/ED regime. It is flawed and needs correction. But those defects should be corrected directly and universally rather than through an unprecedented extension of the reach of private ordering.
1. † © 1998 by Gregory S. Alexander.
2. Professor of Law, Cornell University; Visiting Professor of Law, Harvard Law School, 1997-98. This Response was presented at a Symposium on Law and the New American Family at Indiana University School of Law-Bloomington, on April 4, 1997. I am grateful to Professors Rasmusen and Stake and to participants in the conference for comments and suggestions and to the conference organizers, especially Professor Jeffrey Stake, for inviting me.
3. See Andrée Brooks, Wedding Bells, Dotted Lines, N.Y. Times, Dec. 19, 1992, at 35.
4. There are three different uniform laws regulating the validity of premarital agreements, the Unif. Probate Code (amended 1993) ("UPC"), the Unif. Premarital Agreement Act (1984) ("UPAA"), and the Unif. Marital Property Act (1983) ("UMPA"). Section 6-201 of the UPC adopts the writing requirement of section 2 of the UPAA but relaxes the formalities of execution needed. The UPAA, enacted in about 16 states so far, provides that premarital contracts are enforceable unless the party seeking to avoid enforcement proves either (1) that the agreement was involuntary, or (2) that it was unconscionable when made, and (a) that prior to enforcement, s/he was not provided a fair and reasonable disclosure of the other party's property or financial obligations, (b) did not voluntarily and expressly waive any right of disclosure, and (c) did not have adequate knowledge of the other party's property or financial obligations. See Unif. Premarital Agreement Act § 6(a). The UMPA distinguishes between pre- and postmarital agreements. As to the former, the standard of validity is basically the same as that of the UPAA, see Unif. Marital Property Act § 10(g); as to postmarital contracts, it adopts a higher standard of unconscionability, see Unif. Marital Property Act § 10(f). Apart from statute, some courts have adopted yet more stringent standards of review. At least two courts have ruled, for example, that premarital contracts are voidable if they are substantively unfair, with unfairness measured at the time of enforcement rather than the time of the agreement. Under this approach, the contract is not enforceable if it leaves the nonenforcing party without adequate means of support according to her or his accustomed standard of living. See Newman v. Newman, 653 P.2d 728 (Colo. 1982); McHugh v. McHugh, 436 A.2d 8 (Conn. 1980).
5. Richard A. Posner, Economic Analysis of Law 104 (2d ed. 1977).
6. See Claudia H. Deutsch, More Couples Are Taking No Chances on Love, N.Y. Times, Aug. 13, 1995, at C10.
7. In this respect their proposal substantially tracks the "covenant marriage" idea adopted recently in Louisiana. A party to a covenant marriage contract can obtain a divorce only after counseling and if the other has committed adultery, has committed a felony and been sentenced to death or imprisonment at hard labor, has abandoned the matrimonial domicile for a period of one year and constantly refuses to return, has physically or sexually abused the spouse seeking the divorce or a child of one of the spouses, or if the spouses have been living separately without reconciliation for a period of two years. If a judgment of separation from bed and board was previously obtained, the period of separation is reduced to one year or, if there is a minor child of the marriage, to one-and-a-half years. See Act of July 15, 1997, No. 1380, 1997 La. Sess. Law Serv. 2358 (West) (to be codified at La. Rev. Stat. Ann. §§ 9:224-:225, :234, :245, :272-:275, :307-:309, and at La. Civ. Code Ann. arts. 102-103).
8. See Elizabeth S. Scott, Rational Decisionmaking About Marriage and Divorce, 76 Va. L. Rev. 9 (1990).
9. Assuming for the moment that this standard diagnosis for what ails the modern marriage is correct, why is not the cure reform of the legal default rules to return to FAL? After all, as Professors Rasmusen and Stake acknowledge, initiatives are under way in several states to do just that. See Eric Rasmusen & Jeffrey Evans Stake, Lifting the Veil of Ignorance: Personalizing the Marriage Contract, 73 Ind. L.J. 453, 495-96 (1998). These initiatives seem to indicate a growing public awareness that the shift from FAL to NF/ED was a well-intentioned mistake and now needs to be undone or at least amended in significant ways.
The answer appears to be that Rasmusen and Stake are not confident that the tide in fact has
shifted. Marriage law is in a state of disarray, and its future is ambiguous. Moreover, legal
reform takes time. The no-fault revolution occurred relatively quickly, but it did not occur
overnight. With public opinion so mixed about a return to the traditional FAL regime, any
major shift away from NF/ED will very likely occur slowly. Under these conditions of
uncertainty, private contracting, they argue, provides the quicker, more secure means for
couples who wish to precommit themselves to the traditional obligations of marriage to achieve
their goal. See id. Consequently, they urge "[c]ourts . . . to . . . enforce private agreements
regarding grounds for divorce and terms of an ongoing marriage." (10)
10. Rasmussen and Stake, at 8.Id. 11. For a thorough and penetrating analysis of proposals for new forms of marriage
contracts, see Ann Laquer Estin, Economics and the Problem of Divorce, 2 U. Chi. L. Sch.
Roundtable 517 (1995).
12. For an argument in favor of restrictive marriage contracts explicitly analyzed as a
precommitment device, see Elizabeth S. Scott, Marriage as a Precommitment
(University of Va. Sch. of Law Working Papers Series No. 97-8, Spring 1997).
13. See Albert O. Hirschman, Exit, Voice, and Loyalty 36-40 (1970).
14. Rasmusen & Stake, supra note 7, at 471.
15. For a similar analysis of marriage contracts, see Scott, supra note 9.
16. See Gay C. Kitson & Helen J. Raschke, Divorce Research: What We Know; What We
Need to Know, J. Divorce, Spring 1991, at 1, 12-13.
17. See id.
18. For some evidence supporting these propositions, see Dair L. Gillespie, Who Has the
Power? The Marital Struggle, 33 J. Marriage & Fam. 445 (1971); Gerald W. McDonald,
Family Power: The Assessment of a Decade of Theory and Research, 1970-1979, 42 J.
Marriage & Fam. 841 (1980); Janice M. Steil & Karen Weltman, Marital Inequality: The
Importance of Resources, Personal Attributes, and Social Norms on Career Valuing and the
Allocation of Domestic Responsibilities, 24 Sex Roles 161 (1991).
19. See, e.g., Shelly Lundberg & Robert A. Pollak, Bargaining and Distribution in
Marriage, J. Econ. Persp., Fall 1996, at 139; Shelly Lundberg & Robert A. Pollak,
Noncooperative Bargaining Models of Marriage, 84 Am. Econ. Rev. Papers & Proc. 132
(1994).
20. See Kevin Sack, Louisiana Approves Measure to Tighten Marriage Bonds, N.Y. Times,
June 24, 1997, at A1.