Calling a truce in the marriage wars

�By F.H. Buckley & Larry E. Ribstein*

Draft of March 8, 2000

The debate over the changing institution of marriage is marked by the clash of ideas about social institutions. Recent marriage reform proposals highlight these differences. Covenant marriages, in which spouses waive the right to a no-fault divorce, appeal to social conservatives but are regarded with suspicion by social liberals. Authorization of same sex marriages appeals to social liberals but is strongly opposed by social conservatives. Previous articles have focused on same sex marriage and have proclaimed a total victory to one side or the other in the debate. We offer a compromise based on jurisdictional choice and enforcement of contracts. Objections to recognizing covenant or same sex marriages in one�s own state are stronger than those to enforcing at least the contractual elements of marriages solemnized elsewhere. At the same time, a state need not give a non-standard foreign marriage the same effect internally as to government-conferred subsidies and benefits as it has in the state of celebration.� In other words, we would reject the all-or-nothing principle by which a marriage is either wholly valid or wholly invalid in other states for all purposes. This compromise approach is preferable to absolutist, top-down solutions because it offers a chance of building a consensus between the two sides in the debate.

The cultural clash about legal institutions is most prominent in the current debate about marriage law. Social conservatives support covenant marriages, now recognized in Louisiana and Arizona, in which the parties waive their rights to a quick, no-fault divorce.[1] While social conservatives herald the return to tradition, liberals tend to regard covenant marriage as a reactionary step that reintroduces notions of fault and responsibility in divorce proceedings.� Liberals support, and conservatives condemn, recognition of marriage between people of the same sex. Although state court decisions, most prominently in Hawaii and Vermont, have moved toward recognition of same sex unions, a combination of referenda, state statutes and the limited reach of the judicial decisions themselves has so far prevented the full authorization of same sex marriage in any state.[2] The rhetoric about marriage can be hot. A Congressional attempt to prevent the spread of same sex marriage has been described as "an embarrassment to have been offered and an absolute scandal to have been signed into law."[3]

Choice of law is a logical next front in the marriage wars. Once some states recognize the new forms of marriage the others must decide what local effect to give such marriages. Commentators quickly recognized the choice of law issues in same sex marriages,[4] writing more than 70 law review articles on the subject since the first Hawaii decision.[5] Many states have enacted laws that purport to deny local recognition to foreign same sex marriages.[6]

Viewing together the two new forms of marriage, which most commentators curiously fail to do, demonstrates the need for compromise. Choice of law can provide such a compromise rather than serving merely as the next battlefield. Courts should delineate the contractual and non-contractual aspects of marriage and apply principles of contractual choice of law to marriage's contractual aspects. Spouses would then be able to marry under any state law and have that marriage recognized by courts for contractual purposes in every other state. However, states could determine local non-contractual effects of marriage such as subsidies and tax benefits, and thereby make local public policy concerning marriage and family institutions.� At the same time, states could choose to reject the proposed default recognition rule and invalidate all aspects of foreign same sex marriage

Relying on choice of law gives neither side of the marriage debates the total victory it would prefer.� Marriage is endowed with social meaning and therefore has the power to fashion norms and preferences.� Each side of the marriage debates therefore would like to be able to define marriage and its attributes consistently with its objectives. However, a Constitutional or federal statutory solution is inappropriate in the absence of a consensus on or criteria for resolving these issues. Instead, the law should provide a mechanism by which the debate can evolve toward a consensus.

The article proceeds as follows. Part I discusses contractual choice of law generally and its application to marriage. Part II discusses the debate over the most contentious marriage issues of no-fault divorce and same sex marriage. Part III shows how enforcing the parties' choice of marriage jurisdiction can close gaps between the two sides of the marriage debate over many of these issues.� Part IV discusses how most of the remaining issues can be dealt with by "unbundling" the contractual and non-contractual elements of marriage rather than adhering to the general choice-of-law rule of all-or-nothing enforceability. Part V shows how our pragmatic approach is superior to an absolutist solution that declares total victory for one side, and how this analysis helps to make sense out of the current Constitutional law relating to marriage.

I. Contractual choice of law

This Part introduces the contractual choice of law model, which we use as a starting point for analyzing marriage. Subpart A shows that contractual choice of law is well-recognized in Anglo-American jurisprudence.� Subparts B and C discuss costs and benefits of enforcing contractual choice of law as a mechanism for promoting jurisdictional choice in marriage and in other contexts.� Subpart D discusses the enforcement of contractual choice of law in marriage.

A. The enforcement of contractual choice of law

The dominant rule in American choice of law today is to enforce the law contractually chosen by the parties subject to limited exceptions.[7] This is consistent with the classic case on enforcing contractual choice from Commonwealth law. In Vita Food Products Inc. v. Unus Shipping Co.,[8] the court enforced a provision applying English law though there was nothing to connect the contract with England except the choice of law clause.

American law has accepted the Vita Food freedom-of-contract position despite the general ascendancy of "interest" analysis, which emphasizes the interests of governments rather than the rights of private parties.[9] Section 187 of the Second Restatement on Conflicts adopted the choice of law rule as the proper law of the contract, subject to two exceptions.[10] The first exception is where "the chosen state has no substantial relationship to the parties or the transaction and there is no other reasonable basis for the parties� choice. "[11] While in theory this might seem a rejection of the English position, the rule as applied is close to Vita Food.

The second exception is where "application of the chosen state would be contrary to a fundamental policy of a state which has a materially greater interest than the chosen state in the determination of the particular issue and which . . . would be the state of the applicable law in the absence of an effective choice of law by the parties."[12] This exception is consistent with Vita Food itself,[13] as well as with the conflicts rules of every country.[14] However, the scope of the exception is unclear.[15] Among other things, the public policy exception lets states enforce mandatory rules "designed to protect a person against the oppressive use of superior bargaining power."[16] A court might also ignore the contractual choice if the contract is tainted by a defect of capacity under otherwise applicable law but not under the chosen law.[17]�

Beyond these applications of the rule, questions arise as to the extent to which states do and should regard contracts enforceable under the chosen state as contrary to local "fundamental policy."� This requires an analysis of the benefits and costs of enforcing jurisdictional choice.[18]� Section B discusses the considerations that should impel the courts to narrowly interpret the fundamental policy exception, while section C discusses the case where courts should hesitate to enforce contractual choice -- i.e., where potential spillovers are involved.�

B. The case for jurisdictional choice

Enforcing contractual choice provides an efficient mechanism for promoting jurisdictional choice.[19] The parties necessarily can choose their legal regime whether or not courts enforce choice of law contracts. They can always vote with their feet by choosing to live in a jurisdiction whose laws they prefer, and in doing so will sort themselves out by their preferences. That such sorting occurs is indicated by the fact that more than 40% of Americans live in states other than the one in which they were born.[20] Requiring migrants to choose all the laws that govern their affairs through a single choice or where to live, however, constrains their choice. One might like the climate and job opportunities of Virginia, but prefer to contract under New York law, incorporate in Delaware, and marry under Louisiana or Hawaii law. The choice of law is easily made if the parties can specify by contract the law that governs their contracts, corporations, or marriage without having to visit, reside or contract in the chosen state.[21]

The parties' ability to choose the applicable law has several positive effects. First, letting people choose from among various contract forms can serve as a discovery mechanisms.� States have an incentive to adopt various contract forms in order to attract business from other states.� At the same time, society has an opportunity to discover superior rules by observing the contracts individuals select.

Second, contractual choice helps ensure that the chosen law suits the parties� specific needs better than a one-size-fits-all statutory rule. This reduces the error costs of applying an unsuitable rule. The sorting effect is particularly important in accommodating the variety of views that exist regarding marriage and the family. As long as the choice of law rules give each state adequate scope to enforce its regime, people can choose where to live and where to marry based on their differing preferences for family law.[22]

Third, contractual choice of law effectuates efficient competition among the states. Even without enforcement of contractual choice, states compete for the business and tax revenues associated with increased population. The market for migrants disciplines states to reject inefficient or undesirable laws. This account of a race to the top for migrants is most closely associated with Charles Tiebout,[23] but was first discussed by Frederick Jackson Turner at the end of the last century (when states more clearly perceived increased population as a good thing).[24] But when parties may contract cheaply for the applicable law, they may more easily undo a state�s internal illegality rules. Provided third party effects do not intrude,[25] this results in a race to the top that yields the benefits of more efficient laws with fewer error costs.[26] Ensuring party autonomy regarding the applicable law not only produces more variety, but also tends to produce more efficient laws by motivating legislators to be more public-regarding.[27] Legislators can earn political rents, such as campaign contributions,[28] by redistributing wealth among interest groups that have different costs of organizing to promote or defeat laws.[29] Enforcing the parties' choice of law forces legislators to compete with other jurisdictions and so can reduce their rent-seeking opportunities. Indeed, complete party autonomy to choose the applicable law would render wealth-redistributing laws wholly ineffective.[30]�

C. spillovers

Despite the benefits of enforcing contractual choice of law, there are some circumstances where such choices may be value-reducing. A "race to the bottom" may develop when the parties employ choice of law clauses to evade benign internal rules.[31] Whether the race is to the top or bottom depends partly on whether the out-of�state regime inefficiently imposes �spillover� costs on other states.[32] Some choices, particularly marriage, have social consequences that extend far beyond the contracting parties. In such cases, a privately optimal choice might be socially wasteful, and a state might efficiently override a choice of law clause. This explains why a person cannot purchase in Hawaii the right to pollute in Virginia. Because the migration state has inadequate incentives to consider out-of-state harms, its internal rules might be excessively lax.

The race to the bottom is illustrated by state competition in the provision of insolvency laws during the Articles of Confederation period and the early Republic. States offered an excessively lax fresh start in order to promote migration by debtors from other states with more rigorous insolvency policies. The lax state would benefit from new residents with sterling credit, whose creditors had been left behind them. A state might thus enact inefficient fresh start policies because the costs were borne by out-of-state creditors.[33]

The probability of spillover effects and a race to the bottom is increased when the parties can choose the law of a jurisdiction with which they have no connection. But restricting choice of law clauses to states with which the parties have significant contacts often is not an efficient solution to the spillover problem. If there are significant effects on both sides of the border, legislators may be adequately disciplined to resist inefficient legislation by the local presence of affected parties even if the parties to a particular contract have no contact with that jurisdiction. Moreover, some contracts, such as those relating to marriage, have external effects in the other states regardless of the nature of the initial contacts with the selected state. Thus, the benefits of this rule are often outweighed by the costs of imposing a barrier to contractual choice of law that impedes the efficiency-promoting aspects of jurisdictional choice.

D.� Marriage and choice of law

Choice-of-law rules emphasize marriage's contractual nature by giving the parties considerable power to specify which marriage regime applies. The law of the state of celebration generally determines validity in the absence of the strong countervailing public policy of some other state closely connected with the marriage.[34] Although the public policy exception seems to provide a potential escape route, the courts frequently uphold marriages that are valid where made but barred by the internal domicile law,[35] and this authority is reinforced by statutes that specifically provide that a marriage valid where entered into is valid everywhere.[36] Thus, courts generally distinguish internal and external illegality, reflecting a strong policy favoring certainty regarding the validity of a marriage.[37]

It is not clear whether external rules would favor same sex marriages in the same way. There is a difference between formal requirements as to capacity and same sex unions. When the age of consent is 18, a marriage between 17 year-olds might violate norms of internal illegality, but does not offend traditional morality as homosexual unions do. Similarly, the �potential polygamous marriage,� which is generally upheld in the U.S. and England despite the risk of later polygamous marriages,[38] arguably does not violate traditional norms because the marriage is actually monogamous. Some argue that permitting domicile states to invoke the public policy exception as a way to deal with evasion of their marriage laws appropriately resolves differences regarding same sex marriage.[39] But same sex marriage involves an arguably unprecedented clash of fundamental values regarding the essence of marriage that challenges prior law on the public policy exception.[40] Thus, the public policy exception might sharply limit the interstate recognition of same sex marriage, depending on what non-recognition entails. Unless it is declared unconstitutional,[41] the public policy exception would at least leave the issue for case-by-case adjudication.[42] The application of the public policy limitation therefore is not an answer but simply another way to state the question.

Apart from the scope of the public policy exception, it may be significant that a marriage is formally regarded for choice of law purposes not as a contract but rather as a status conferred by the law of the celebration.[43] This is analogous to the treatment of corporations for choice of law purposes.[44] The status approach to marriage has some important implications for choice of law. First, while individuals can determine the applicable marriage regime, they can do so only by traveling to it rather than selecting the state in their contract. Although this imposes higher costs on parties seeking jurisdictional choice than pure contractual choice of law, these costs might not be thought forbidding in so fundamental a contract as marriage.

Second, the status nature of marriage may pose a difficulty for same sex marriages by implying an all-or-nothing approach in which one is either married (with all the incidents of this status) or not (with none of them).[45] Thus, monogamous heterosexual marriages valid under the law of the state of celebration have all consequences of marriage under the law of the spouses' domiciles. The state does not adjust consequences to suit the particular relationship in accordance either with the state's policy or with the parties' contract. Courts that are unwilling to recognize some of the consequences of same sex marriage may recognize none. On the other hand, courts are willing to distinguish between status and the contractual incidents of actual polygamous marriages. For example, the multiple wives of a polygamous marriage may have inheritance rights.[46] Thus, a court might enforce a same sex marriage for some purposes but not others. Indeed, some recent state laws on same sex marriage do not recognize any incidents of such marriage, while others recognize domestic partnerships.[47]

Third, the failure of choice of law rules to treat marriage as a contract is problematic for covenant marriages. Under current law, a divorce valid where entered is entitled to recognition everywhere.[48] Thus, a Nevada divorce might end a Louisiana covenant marriage even if the divorce is contrary to Louisiana law. This suggests a dark future for covenant marriage laws, short of a new federal law or constitutional amendment,[49] and thereby considerably diminishes the value of a covenant marriage as a self-bonding device. If, on the other hand, marriage and divorce were treated more like contracts, the parties could choose the law under which the contract might be terminated through divorce under the Vita Food principle of free choice.[50] Thus, although it is said that marriage is more than a contract, from a choice of law standpoint it might be said that the status aspect of marriage makes it less than a contract.� These rules have arguably led to a race to the bottom in divorce law. Divorce laws were always more lax in western states such as Nevada, which sought migrants from more restrictive eastern states.[51] Because the party who seeks a divorce may have acquired a second family in the new state, cutting off the claims of an out-of state spouse effects a wealth transfer to the new family in the divorce state from the original family in the marriage state. Moreover, divorce havens such as Nevada attract legal and tourist business within the state while imposing some costs on divorced families elsewhere. The competition for migrants may explain why all states liberalized their divorce laws[52] and why lax divorce laws continue to attract migrants.[53]

The application of contractual choice of law to non-standard forms of marriage is, therefore, unsettled. The critical policy issue is whether contractual choice of law would lead to spillovers and a race to the bottom. The issues relating to the types and extent of effects of foreign marriages in the spouses' domicile are discussed below in Part II.� The issues cannot be resolved by simply characterizing marriage as either status or contract for all purposes. As discussed below in Part IV, complete analysis of choice of law issues in marriage requires unbundling its contractual and non-contractual elements. In the meantime, we discuss in the following two parts the substantive policy issues involved in marriage and how choice of law offers an approach to resolving those issues.������������� �����������

II. The marriage debates

This Part discusses the current debates over marriage. This discussion forms the background for the analysis in the remainder of the article of how the contractual choice of law model should be applied to marriage.

A. Waiving No-Fault Divorce Rights

Since 1975, every state has permitted divorce without proof of fault.[54] Liberalization of divorce law appears to have increased the rate of divorce.[55]� Recently enacted laws in Louisiana[56] and Arizona[57] give the parties the option to elect a more durable form of covenant marriage, by permitting them to waive their rights to a subsequent no-fault divorce. Spouses in such a marriage may divorce only upon a showing of adultery, abandonment, crime, one-year separation or a judgment of separation on showing of abuse.[58] Because courts do not enforce promises not to divorce,[59] this law supplements what the parties are able to do for themselves by private contract. Subsection 1 discusses some reasons for covenant marriage. Even if one disagrees with these arguments, the important question is whether there are any problems with making this option available to the parties.[60] The remaining subsections consider some possible objections to covenant marriage laws.

1.� The case for covenant marriage

Covenant marriage can be viewed as a type of standard form contract, in which expanded choice facilitates socially beneficial arrangements. The most important function of covenant marriage is to protect spouses and others (particularly children) from the opportunist spouse who seeks an easy marital dissolution under no-fault divorce law. Because of the possibility of opportunism, and the loss of what the economist calls "appropriable quasi-rents,"[61] the parties will under-invest in the marriage: they will not commit as closely to each other as they would if they could contract around easy divorce rights. The under-investment might take the form of a weaker emotional commitment. Alternatively, the parties might be less ready to embark on projects (such as offering each other material support or buying a house) affected by divorce. These pathologies mutually reinforce one another.[62]

The increased probability of divorce particularly affects a married couple's desire to have children. Children are "marriage specific" assets who increase the costs of a marriage break-up.[63] One or both of the spouses likely will make investments in raising children.[64] Those who invest in children may forego valuable opportunities to develop skills they could sell in employment markets. Spouses who are freed from childcare responsibility can work outside the home and contribute financial assets to the marriage. The childcare and childbirth contributions are normally loaded in the front end of the relationship while the financial contributions are back-end loaded. The asymmetric quality of the spouses� investments creates a risk of opportunism: The �working� spouse can withdraw from the marriage after reaping much of the advantage of specific investment in the children but before making his or her financial contribution (assuming he manages to keep his emotions pliable).[65]

The asymmetry and opportunism are reinforced by gender roles, which to some extent are biologically determined. Mothers are more likely to make the investments than fathers, since mothers have a closer physical connection to their children and place a higher value on time spent with them.[66] This encourages the sort of specialization of activities in the household that gives rise to the opportunism problem. Whether or not they invest during the early years in marriage-specific assets like children, women invest critical years of reproductive capacity and sexual attractiveness.[67]

Covenant marriage protects against this appropriation of the homemaker spouse's marriage-specific investment. A covenant marriage statute that makes divorce available only with the consent of both spouses or on a strong showing of fault is a kind of "property" rule that forces the wage earner to buy his way out of the marriage. The price will depend on the homemaker's evaluation of what she will lose by divorce. By contrast, under no-fault divorce, the homemaker's protection depends on alimony and support payments. This is a liability rule that attempts to fix the value of the homemaker's loss. Yet it is difficult to do this through legal rules and judicial decisions. Although the basis of alimony and support payments is unclear,[68] they probably do not fully compensate the homemaker for her opportunity costs.[69] This is consistent with evidence of women's increased participation in the labor market under no-fault divorce, an indication that women are making fewer marriage-specific investments.[70]

Covenant marriage's ability to protect investments from appropriation not only helps the spouse who specializes in childcare, but also can benefit children, to the extent that it reduces divorce rates, for children are often the principal victims of divorce.[71] Moreover, covenant marriage offers potential self-binding gains for the non-caregiver. Self-binding gains arise when promises are time-inconsistent, that is, where one�s interests at contract formation are different from those at the time of performance. It is often in one�s interest to promise to perform at formation, but to breach when the time comes for performance. In such cases, the promisor may be made better off if, at the time of formation, he can credibly commit to perform, since contractual gains will be lost if the promisee cannot rely on the promise.[72] Covenant marriage may help in this way by making it easier for the non-caregiver to secure a caregiver mate. Thus, barriers to divorce facilitate the traditional homemaker/wage-earner marriage some couples prefer but would be reluctant to enter into without protection from opportunism

The commitment facilitated by covenant marriage also has side benefits for both spouses. Couples who choose the covenant marriage option for its self-binding benefits may enter into higher quality marriages. Because the restriction on exit increases the cost of a bad marriage, the parties will react to this before marriage by conducting a more thorough search for a compatible partner.[73] Making available the covenant marriage option also signals useful information to a prospective spouse. This is particularly important in light of the importance of choosing a mate and the difficulty of learning all of the relevant facts about a prospective mate prior to marriage. An important characteristic many look for in a spouse is the prospective partner's commitment to the marriage. The potential mate's agreement to a covenant marriage sends a strong signal of commitment that is not available when the law provides for only a weaker marriage regime.

The discussion so far in this subsection has focused on tangible effects of covenant marriage.� But covenant marriage has important symbolic value as well.� Through covenant marriage, states endorse the primacy of commitment and responsibility over individual freedom and love.� Perhaps more fundamentally, a state's refusal to break the bonds of marriage represents a return to traditional values from the instability that has characterized modern marriage.

Even with the covenant marriage option, many couples will prefer to retain no-fault exit rights. Covenant marriages are not mandatory. The case for covenant marriage rests on the simple proposition that it offers a useful option.[74] The question is whether couples should be denied this choice.� The remainder of this section discusses some potential arguments against the covenant marriage option.

2.� Protecting covenant spouses.

Lawmakers may conclude that parties who agree to a covenant marriage will thereafter regret their choice, and wish they had elected the easier no-fault option. Barriers to no-fault marriages benefit such parties. There might seem to be special reasons for concern since young adults often will make the choice when they are in love and more willing to forgive faults in their partners than they will be after a few years of marriage. In addition, survey data suggests that newlyweds underestimate the probability of divorce.[75]

It is, however, difficult to evaluate such claims. One would expect the parties to weigh more carefully the risks of divorce when they elect a stronger marriage bond, and the problem might thus be self-correcting. It is not surprising that husbands who enter covenant marriages are reported to be better educated than those who choose non-covenant marriages.[76] Survey data about divorce expectations also are inconclusive, since couples may appreciate the risks of divorce even if they misreport them.

Some commentators make gender-based arguments for restrictions on contract.� They argue that men are better bargainers than women,[77] and that women are more risk averse than men, and therefore more willing to make concessions in negotiations.[78] But these arguments are questionable, even if one accepts their anti-feminist logic. First, even without covenant marriage, the spouses would still be able to bargain over the terms of marriage contract. If men had superior bargaining ability, they might then be expected to reap the same advantages through ante-nuptial agreements over alimony and property rights. At best, barriers to covenant marriage would resemble a fence against foxes erected on one side only of the hen house. Second, if options are to be limited, there is no reason to think that the no-fault option is the right one for women, since there is significant evidence that no-fault has harmed them.[79] If women need paternalism, one might more plausibly withdraw the no-fault option from them.[80] Third, the risk of opportunism in marriage stems from the women's having more to contribute at the front-end of the marriage. The same consideration suggests that women may have a stronger bargaining position at that time and therefore might reap a greater advantage than men from the opportunity to bargain.[81] Fourth, there are less intrusive ways of dealing with biased judgment and bargaining problems. For example, the law might make the election ineffective unless confirmed by the parties six months or a year after marriage, on the model of cooling-off rights in consumer protection law.[82] It also might require the parties to attend counseling sessions before signing on to a covenant marriage, as under the Louisiana statute.[83] In the past, the bargaining problem was addressed by giving the parents a role in the negotiations, and this tradition might perhaps be revived as the children of formerly progressive baby boomers reach marriage age.

3. Protecting children

Covenant marriage might impose costs on parties other than the spouses. Preserving a bad marriage may hurt the children of this marriage.[84] But these effects are ambiguous, for children are the primary beneficiaries if covenant marriages reduce divorce rates. Moreover, spouses arguably will take prospective costs as well as benefits to their children fully into account in deciding whether to have a covenant marriage.

4.� Judicial costs.

Rocky marriages also impose a strain on the courts.[85] Since the courts are supported by tax dollars, covenant spouses would not internalize these costs. However, the problem of potential litigation exists in any type of relationship, and easy divorce has not proven to be a cure for litigation over the consequences of divorce. [86] If covenant marriages reduce divorce levels, they might then reduce litigation levels.

5.� Over-signaling

A potential advantage of covenant marriage -- that it offers couples a way to signal their reliability as marriage partners � might be considered a disadvantage if it leads couples to over-commit to durable marriage. Presented with a choice between hard and soft marriage laws, both parties might prefer the more liberal no-fault regime, since both recognize the possibility that they might subsequently want to leave the marriage. But revealing this preference might be taken to signal that they are less reliable than is the case. Both parties might therefore commit to a covenant marriage when that option is available, although both privately prefer the no-fault divorce option.[87]� This can be termed a problem of "over-signaling."� In other words, by creating the covenant marriage option that spouses can accept or reject, the law forces parties to send a signal that is stronger than the one they wish to make.

But over-signaling concerns offer a problematic basis for policy prescriptions. Private preferences, which the parties are loath to reveal, are virtually impossible to verify. Also, the over-signaling argument makes possibly unrealistic assumptions about behavior.� For example, marriage partners may well under-signal their level of commitment in order to avoid challenging prospective mates. Thus, over-signaling concerns do not support the denial of signaling benefits for those couples who do wish to waive no-fault divorce rights.

In sum, a variety of arguments may be made for and against covenant marriage, and reasonable people might differ on the design of rules of internal illegality. Even if there are potential costs to covenant spouses or others, these costs must be balanced against the benefits of covenant marriages for parties whose marriages are thereby made more durable, and who might have divorced under a no-fault regime.[88] The basic question is whether the costs and benefits of covenant marriage are to be weighed by the individual parties themselves or by the legislators on behalf of all couples. Since a one-size fits-all rule will prove unsuitable for many couples, there is much to be said for the covenant marriage option, even if the option is available only by moving to another state.

B. Same sex marriage

Many commentators have advocated giving same sex couples the right to marry.[89] The debate about same sex marriage has continued for some time[90] but became more pressing when Hawaiian courts held that the refusal to recognize same sex marriages constituted unlawful discrimination.[91] A judge in Alaska also held that individuals have a fundamental right to choose a "life partner."[92] More recently, the Vermont supreme court held in Baker v. State[93] that homosexual couples are entitled to the same rights and benefits as heterosexual couples. Despite all this activity, no state so far has recognized same sex marriage. The Baker decision in Vermont stopped short of compelling the state to authorize same sex marriage and left the issue up to the legislature.[94] Referenda overturned the Hawaii and Alaska court decisions.[95] Meanwhile, a majority of states have passed anti-same-sex-marriage laws[96] and one state court has denied a constitutional right to same sex marriage.[97] On March 7, 2000, California voters passed a referendum, Proposition 22,� that invalidates all same sex marriages wherever celebrated.[98] Thus, the availability of the same sex marriage option still remains an open question.� The following subsections discuss both sides of the debate on this question.

1.� The case for same sex marriage

Proponents of same sex marriage have emphasized the benefits for homosexual couples. In other words, the asserted benefits of same sex marriage are, to some extent, the same as those for heterosexual marriage. Marriage provides access to benefits that are available to heterosexual married couples, including tax benefits, family benefits from private employers, evidentiary privileges, intestate inheritance rights and rights under immigration law.[99] Also, marriage lets same sex couples opt into the elaborate standard form framework that is now available to heterosexual couples.[100] This includes, among many other things, rules regarding ownership and inheritance of property, and dissolution rules that make break-up costly. These rules are particularly suitable for a long-term domestic relationship. As discussed in the previous Section, couples can signal a stronger commitment to each other by their willingness to incur a loss on a break-up of the relationship. They may then more securely invest in the relationship, and this will reduce the probability of a subsequent break-up.[101] This might benefit the couple, and perhaps society as a whole, if the positive external benefits of stable relationships are not restricted to heterosexual couples.[102]

Perhaps the most important benefit of marriage for same sex couples is that recognizing same sex marriages would place a seal of societal approval on homosexual relationships.[103] This is what distinguishes recognition of same sex marriage from merely giving same sex domestic partners the same rights as heterosexual married couples. This issue is particularly contentious with respect to the Vermont decision in Baker, which as noted above refused to compel the legislature to permit same sex marriage. To the extent that the denial of marriage rights stigmatizes homosexuals, the recognition of same sex marriage might benefit them even if they would not marry themselves. Until recently, same sex relationships remained in the closet, since homosexuality was regarded as socially deviant. The social stigma of homosexuality is now much reduced, and recognizing same sex marriages would accelerate this trend. We discuss the role of the law in shaping norms and preferences in more detail below.

2.� Tradition

Opponents of same sex marriage have argued that these marriages involve several different types of costs. It is important to understand these arguments in order to determine precisely how those arguments play out in the context of choice of law. Advocates of same sex marriage have vigorously challenged these arguments.[104]� It is important to emphasize that, for the purposes of this article, we take no position on who ultimately has the better arguments. Indeed, the inability to make such a decision is an important reason for letting diverse regimes flourish under a choice-of-law approach.� The discussion in this and the following subsections assumes only that arguments both for and against same sex marriage have some rational basis.

Recognizing same sex marriages would represent a sharp change in a major social institution, and this argues for prudence. As Edmund Burke has noted, �it is with infinite caution that any man ought to venture upon pulling down an edifice which has answered in any tolerable degree for ages the common purposes of society.�[105] The existence of a settled tradition, concerning a fundamental institution, should give one pause before welcoming radical changes. The fact that something has worked over a long period says something about its usefulness, and raises questions about the relative merits of what might replace it. �Instead of exploding general prejudices,� theorists might better �employ their sagacity to discover the latent wisdom which prevails in them.�[106] Claims that such institutions have no essential meaning, and are simply defined over time by society,[107] ignore this point. As we learned from Jules and Jim, it often does not make sense to "start from zero and rediscover the rules."[108]

The argument from tradition has limitations.[109] A settled tradition might strengthen the conservative position, but does not provide an independent reason to reject proposals for change. Merely because society defines an institution in a particular way does not necessarily mean that the definition is correct. If the case for reform is in other respects persuasive, the move should be made despite the long acceptance of the existing order. But where the matter is not so clear, and where the tradition is as fundamental as marriage, the additional arguments for conservatism that we consider below must be respectfully examined.

The debate over the role of tradition came to the fore in the recent case of Baker v. State. The Vermont court rejected as illogical and lacking empirical support the argument that giving the same benefits to same sex and heterosexual couples would have unpredictable effects on the institution of marriage.[110] However, in stopping short of requiring recognition of same sex marriage the court noted that arguments about destabilization of the institution of marriage are "not altogether irrelevant. A sudden change in the marriage laws or the statutory benefits traditionally incidental to marriage may have disruptive and unforeseen consequences."[111]

3.� Effect on marriage standard form.

A second argument against same sex marriage is that it might increase contracting costs for participants in heterosexual marriage.� Marriage law economizes on transaction costs by providing married couples with standard form rules they would have had to bargain over in the absence of transaction costs.[112] The extent to which transaction costs are saved depends on how well rules are tailored for marriage relationships. Because a single set of default rules will not suit some participants in heterogeneous relationships,[113] some people may have to incur the costs of bargaining around, or being bound by, ill-fitting rules.[114]

Recognizing homosexual marriages might exacerbate the problem of ill-fitting default rules.[115] The presence of children in heterosexual marriage has implications for default rules in marriage and divorce and inheritance rights. First, the different types of marriage demand different rules regarding the grounds for divorce. Homosexual spouses are less likely to demand sexual fidelity from their partners than heterosexual couples, who are concerned about the paternity of their natural children.[116] This difference between the two types of couples may make it more difficult to reintroduce notions of fault in divorce, which itself may be desirable to reduce the costs of no-fault divorce discussed in Part II. Even under current no-fault divorce rules, homosexual marriage requires different rules regarding the consequences of divorce. As discussed above in Section II(A)(1), where one spouse has specialized in child-rearing in a heterosexual marriage, interspousal payments in a standard form divorce might usefully compensate her for her opportunity costs in foregoing a career, and reflect the bargain the parties would likely have struck at marriage.

Second, the different types of marriage also demand different inheritance rules. A deceased spouse in a heterosexual marriage is likely to want to leave everything to the surviving spouse for support and for the children as well as to recognize joint contributions to maintaining the household. The members of homosexual unions may be more likely to be financially independent and therefore might prefer to leave their wealth to other members of their immediate family.

Some homosexual couples do, of course, raise children, including the biological children of one of the spouses from prior marriage, artificial insemination and surrogacy. More might do so if marriage law changed to make homosexual unions more child-friendly. Conversely, many heterosexual couples are childless, and those who marry late in life know they will never have children. Thus, this argument against same sex marriage depends on whether restricting marriage to heterosexual couples does a reasonably good job of fitting rules to particular cases. If most heterosexual marriages result in children, and very few homosexual couples would raise children, then different standard forms arguably commend themselves.

This is not to say that all heterosexual couples would seek the same set of implied terms in their marriage agreements. Different couples will often have different ideas about their responsibilities, and a single set of default rules will not fit every couple. But this argues for additional standard forms, such as covenant marriage, rather than for introducing greater incoherence by assimilating still more dissimilar unions into the marriage standard form. The law might offer different types of marriage based on the likelihood of children rather than on sexual preferences. On the other hand, the heterosexual couple, even if likely to remain childless, might be more likely than the homosexual couple to seek a standard form that resembles the normal marriage option. Elderly heterosexual couples may desire the same degree of fidelity as younger, fertile couples because their preferences about marriage were fixed when they were younger.

Thus, a concern for the coherence of the stock of default rules might be thought to argue for separate forms of marriage for heterosexuals and homosexuals, rather than a ban on same sex marriages. Yet this intermediate solution is not free from difficulties either. It is difficult to believe that courts would keep the rules of closely related marriage categories in watertight compartments, without allowing slippage from one to the other. Nor would this solution satisfy those who want same sex unions blessed by the full state imprimatur.

4.� Effect on children

As noted in Subsection 3, permitting same sex couples to marry might encourage them to have and raise children.� By marrying, these couples would have access not only to child-friendly inheritance and divorce rules, as just discussed, but also might be more eligible for adoption and custody. This raises questions concerning the effect on children of parents of the same gender.��

A mother and father give the child two gender role models that arguably contribute to its psychological development by encouraging heterosexual behavior.[117] To be sure, whether this is a benefit depends on what one thinks of heterosexual behavior.� However, it is enough for present purposes to assume that encouraging heterosexual behavior is a reasonable, even if debatable, goal. Moreover, if one concludes that heterosexuality contributes to being well-adjusted in a society dominated by heterosexual relationships, the benefits of encouraging heterosexuality would extend beyond the children themselves to building society�s "social capital."[118]

Apart from the effect of same-gender parents on the children they raise, the state also might be concerned in the number of children, and therefore offer subsidies to heterosexual marriages on the ground that they promote births. This argument would justify refusing a similar subsidy to childless same sex couples.[119]

The effect of same sex marriage on children is one of the most hotly debated issues in the same sex marriage debate. That is not surprising, since adverse effects third parties are a strong basis for invalidating otherwise consensual relationships.

5.� Shaping norms and preferences

The above arguments take as given the parties' preferences and seek to design legal rules in light of these preferences. Though this may be enough for rules about commercial contracts, marriage has special social significance. Thus, it is important to take into account the way in which the legal system may shape individuals� preferences.[120]

The law relating to marriage functions in two ways to change preferences. Transfers and tax benefits and burdens change the benefits and burdens of certain relationships and thereby encourage individuals to prefer the relationships that most benefit society.[121] In particular, the law encourages cohabitation and childbearing in a particular type of long-term partnership -- companionate marriage.[122] This has been referred to as the "channeling" function of marriage law.[123]

Less directly, marriage law may encourage some conduct, such as heterosexual sex and companionate marriage, by expressing society's approbation or condemnation.[124] Society clearly has expressed a preference that intimate relationships take place within marriage.[125] Indeed, it is this preference that raises the stakes in the marriage debate. Society's sanction or disapproval sends signals about people's reliability or socialization.[126] Marriage law thereby defines the "social meaning" of a particular type of cohabitation.[127] Getting married is defined as one's passage into adult and responsible society and middle class domesticity, while cohabiting without marriage is defined as immature and irresponsible.[128]� Marriage law thus establishes norms of behavior by setting types of conduct for which people earn the esteem of others in their group.[129]

This raises questions concerning what effect marriage law should have on preferences and norms.[130] One set of questions concerns the effect of marriage law on attitudes about homosexuality. Social conservatives use marriage law to defend traditional lifestyles by ensuring that benefits are conferred only on heterosexual relationships.[131] At the same time, activists who push for same sex marriage can be seen as "norm entrepreneurs"[132] who seek to change norms and preferences by legitimizing homosexuality. This is part of what has been termed the social construction of attitudes about homosexuality.[133]

Another set of questions concerns the effect of same sex marriage on attitudes toward marriage. Recognizing same sex marriage arguably weakens the sacramental respect accorded marriage among those who think homosexual sex wrongful. Moreover, once the bounds of traditional marriage are erased, it is not clear what other types of marriage should be recognized, particularly polygamous marriages. Although religious norms condemn polygamous unions, this should not matter in a religiously neutral state, and in any event does not distinguish polygamy from same sex marriage.[134] Any reduction of incentives to marry, as by weakening the sacramental respect accorded marriage, might thus have serious repercussions, particularly when combined with other recent developments that have depressed marriage rates and raised divorce rates.[135]�

On the other hand, allowing same sex marriage might strengthen marriage by encouraging same sex couples to formalize their relationships through marriage. Or if same sex marriage does weaken traditional marriage, this might be a good thing.� It has been argued that, over the long run, recognizing same sex marriage might facilitate changes in existing social roles that ameliorate the loss of individual autonomy caused by the institution of the family.[136]��

Apart from questions about which way the law should go, there is a more fundamental question concerning the appropriateness of deliberately using law to shape norms and preferences. Setting norms arguably is best left to spiritual leaders rather than government.[137] For example, permitting homosexual marriage might inefficiently weaken valuable norms favoring traditional marriage.[138] Conversely, government reinforcement of traditional norms might inefficiently permit the prevailing majority to punish conduct by the minority that is socially productive or useful.[139]

6.� Over-signaling

Same sex marriages might create a more serious over-signaling problem than that for covenant marriage.[140] Many same sex couples will not want to opt into the marriage standard form because it does not suit their relationship. But if marriage is available and homosexual couples do not take advantage of it, they may be sending a negative signal about their social worth that they would not be sending in the absence of the marriage option.[141]

III. The marriage debates and choice of law

This subpart discusses how the policy debate over same sex and covenant marriage plays out in the context of choice of law. It shows that most of the arguments against non-standard forms of marriage can be resolved by letting states enforce foreign marriages that they would not enforce if contracted within their borders. This discussion assumes that states would retain some power to decide the precise local attributes of a foreign marriage.� Part IV discusses more precisely which elements of foreign marriages would be enforced.

A. Covenant marriage

Part II examined the case for restrictions on covenant marriages, as a matter of internal illegality. We now examine the same arguments from a choice of law perspective.

1. Protecting spouses from improvident marriages.

Paternalism arguments against covenant marriage are not particularly compelling.[142] But even if they were, the argument from paternalism does not supply a reason to impeach a covenant marriage entered into under the laws of a state to which the parties have traveled to marry.

If the argument from paternalism justified the ban on no-fault waivers, it would also be sensible to impeach end-runs around the ban through choice of law clauses. A party who cannot bargain wisely over divorce rights is also unlikely to bargain wisely over the attempt to modify the divorce regime through another term in the contract. This explains why a party�s capacity to contract is governed by the system of law with which the contract is most closely connected or by the law of his domicile or residence, notwithstanding the choice of another law to govern the contract.[143] However, while the spouses may not be able to enter into a binding choice of law clause, they can accomplish the same thing indirectly, by travelling to another state to marry. For example, minors might avoid internal barriers by travelling to a state with a lower age of majority, and this will be recognized by his home state.[144]

It makes sense to let the parties skirt internal barriers in this way. It would be very troubling to introduce an element of doubt about the status of a marriage. For example, a second marriage might be bigamous in one state and not in another. Moreover, the laws as to capacity to marry are fairly uniform across the United States, so that recognizing the marriage will rarely shock the conscience.[145] And if the problem is a concern that decisions should be made deliberately, the parties will often have provided striking evidence of this by travelling to the foreign state.

These arguments are particularly compelling in the case of covenant marriage. First, certainty is vitally important here, since an agreement not to divorce is worthless unless the parties can be sure it will be enforced. Second, the difference between covenant and no-fault marriage options is not such as to shock the conscience. Liberalization of the divorce regime appears to have weakened marriages, and left women worse off.[146] This suggests that the covenant marriage option is a reasonable choice for many couples. Third, and perhaps most important, the decision to travel to another state to take advantage of a rule that ensures the stability of the marriage provides even stronger evidence of deliberation in adults than it does of maturity in minors. The decision to have a marriage performed far from the homes of family and friends is likely to be a significant inconvenience. These costs will be most strongly felt by couples that care enough about their marriage to seek out a foreign law in order to make it more durable. Thus, even if a state has concerns about covenant marriages performed locally, these concerns are not enough to justify refusing to enforce a covenant marriage performed elsewhere.

2. Protecting children and other third parties.

The domicile state may have a strong reason for protecting local children, courts and welfare agencies from the negative effects of binding a couple to an unhappy marriage.[147] As discussed in Part II, there is a strong debate over whether no-fault divorce or binding marriage offers the best set of solutions. Most importantly for present purposes, the effects of either option are felt in the state where the children are domiciled. Accordingly, there is a potential spillover problem if State A's covenant marriage law is enforceable in State B where the effects of unhappy covenant marriage occur. Thus, it is in this respect that requiring local enforcement of covenant marriages is most problematic. As discussed in the next Part, however, separating out the purely contractual elements of marriage offers a possible solution to this problem.

3. Protecting prospective or rational spouses: over-signaling.

A third objection to covenant marriage is that a couple might seek to over-commit when presented with a choice between hard and soft marriage laws in order to avoid sending a negative signal about their unwillingness to commit.[148] Thus, the availability of covenant marriage may impose costs even on couples that reject the covenant marriage option, or that enter into a covenant marriage with full appreciation of the consequences.

Even granting this argument, it is considerably weaker when the parties must travel to another state to waive no-fault divorce rights. Because the likelihood that the parties will over-signal depends on the availability of the option, the concern is greater for Louisiana couples than it is, for example, for Virginia couples who travel to Louisiana to marry. As noted above, couples normally marry at home, where their family and friends are. Most non-Louisiana couples will find it a great sacrifice to give this up, unless they strongly prefer a covenant marriage. This significantly dilutes the signal inherent in foregoing a covenant marriage. In other words, a potential spouse might credibly refuse an out-of-state covenant marriage because of cost and inconvenience without appearing to reject durability.

Even if making an interstate option available makes over-signaling possible, such costs are likely to be less than the significant signaling benefits of offering the covenant marriage option. Permitting spouses to choose from among the marriage standard forms available across the states would allow them to signal more precisely the level of commitment they are prepared to make.[149]�

B. Same sex marriage

A state that rejects same sex marriages under its internal laws might nevertheless, through a narrower doctrine of external illegality, recognize a same sex marriage contracted in another state with more liberal internal laws. Part II reviewed the case for banning same sex marriages as a matter of internal law. This Part shows that such arguments are weaker when it comes to the design of external illegality rules.� Note, however, that this may depend on the domicile state's ability to control some of the consequences of marriage.� Unbundling of marriage consequences is discussed further below.[150]

1. Effect on marriage law.

Recognizing same sex marriage would weaken the coherence of marriage rules by forcing them to deal with two very different types of marriages.[151] Suppose, however, that a state such as Hawaii chooses to make standard marriage available for same sex couples. Another state should not refuse to enforce the marriage on coherence grounds because a concern for coherence justifies internal and not external barriers. Applying a foreign law does not affect the coherence of the enforcing state's internal marriage regime. It might be otherwise if the enforcing state applied the foreign law as to the status of marriage, but its own law as the consequences of marriage (such as inheritance rights). At best, therefore, coherence concerns might argue for full-fledged enforcement of foreign same sex marriages, with the foreign law governing the consequences of marriage.

2. Effect on children.

If children of same sex spouses are worse off than those of heterosexual couples,[152] it should not make a difference whether the couple married under the law of another state.� However, as long as the domicile state would control the consequences of such a marriage, including such things as custody and adoption of children, the mere fact of the foreign marriage would not have such effects outside of the state of celebration.

3. Norm and preference shaping.

The most contentious argument regarding recognizing same sex marriages is that it places a seal of approval on homosexual relationships.[153] This is deeply desired by many homosexuals but just as strongly opposed by many religious conservatives. Side-stepping the debate by declaring one side the unambiguous winner and anathematizing the other side does not serve to resolve it. Instead, we suggest a choice of law solution to the dispute.

Even if one accepts, as all sides in the debate about covenant and same sex marriage do, that the state has a reasonable role in play in shaping norms through its domestic laws, it does not follow that such policies should also inform the state�s private international laws. A state that objects to same sex marriages might signal its views with sufficient clarity through its rules for in-state marriages. There is far less need to go further and seek to legislate on the expressive function of foreign laws, when out-of-state marriages comes before in-state courts.

State recognition of the validity of same sex marriages that are validly celebrated under the laws of a foreign jurisdiction, while according different consequences to the two types of relationships, offers a Solomonic solution to the problem of preference-shaping effects of marriage law. Although the couple might earn some social respect as a �married� couple, as discussed in more detail below in Part IV, states might deny them many of the tangible benefits received by heterosexual couples. The absence of such benefits would not only directly affect conduct at the margins, but would also have the indirect effect of sending a message that homosexual relationships are not the norm. This, in turn, would attach additional emotional costs to the relationship.��

4.� Over-signaling.�

Similar considerations apply to over-signaling arguments in the same sex context as regarding covenant marriage[154]: the concern is greater for couples in states whose law provides for same sex marriage than it is for those who must travel to another state to marry.

IV. a contract-based approach to recognizing foreign marriage

The distinction between external and internal enforcement of marriage does not wholly eliminate the kinds of difficulties with covenant and same sex marriage that discussed in Part II. The remaining question is precisely what effect states must give to foreign marriages. This Part proposes a judicial rule enforcing the contractual elements of marriage (such as property division and support), while stopping short of recognizing the public elements (such as welfare subsidies and tax breaks). Thus, we would unbundle marriage law, by discarding the status-based approach of recognizing the marriage either for some purposes or for none.[155]

Subparts A and B set forth the analogy between marriage and contract.� This analogy argues for a conflicts rule of interstate recognition of contractual elements of marriage. Subpart C discusses the distinction between contractual and non-contractual aspects of marriage for choice of law purposes.

A. Covenant and Durable marriages

Spouses could replicate many features of covenant marriage by contract, through private agreements that attach monetary penalties to divorce, or that condition divorce rights on a showing of fault, as under a covenant marriage. Through such contractual restrictions, the parties might privately craft a �durable marriage.� To the extent that the parties can mimic a covenant by a durable marriage, there is less reason to resist our conflicts solution to the problem.

Courts might decline specifically to enforce a durable marriage because of the judicial burden of supervising performance, by analogy to personal service contracts or partnerships,[156] or because of paternalism concerns about holding the parties to onerous obligations.[157] However, both concerns might be alleviated under a contract that merely imposes fault conditions or time requirements rather than forbidding divorce.[158] Although courts might now be reluctant to enforce these conditions because they fetter divorce rights that are generally assumed to be mandatory,[159] the expansion of free contracting rights suggests that courts might relax such barriers over time.

Even without specific enforcement, the parties might be able to contract for something close to covenant marriage by permitting divorce but providing for liquidated damages unless the contractual conditions of dissolution are satisfied.[160] In the debate over no-fault divorce, commentators sometimes overlook the extent to which property divisions and support duties penalize fault. Many states take marital misbehavior into account in assessing financial obligations, sometimes through explicit statutory injunctions, and sometimes through implicit judicial standards.[161] Could the parties take this one step further and bargain over fault penalties in a durable marriage? Such terms would be similar to those already included in some ante-nuptial agreements. The main legal issue would be whether the liquidated damages constituted a penalty.[162] From a policy standpoint, there might also be an over-signaling concern with offering the parties the ability to bind themselves by contract. On the other hand, the over-signaling problem is weaker than with covenant marriage if the state merely enforces contracts rather than offering a clear choice of a more binding commitment than standard-form marriage.

An important element missing from a durable marriage contract as compared with covenant marriage is a binding constraint on remarriage. A state might enforce the contract but permit divorce and thereby facilitate remarriage that is effective everywhere, including in the covenant marriage state. The contract might deter remarriage through a damage remedy, but this would not significantly constrain spouses who have little wealth. On the other hand, barriers to remarriage similarly may under-deter. While bigamy is illegal everywhere, legal restrictions on adultery are ineffective.[163]

Even if courts hesitate fully to enforce durable marriages, legislatures might fill the gap by promulgating statutory standard form contracts that spouses could enter into when they marry requiring the party seeking a divorce to pay damages in some circumstances.[164] Such statutes would have several potential advantages over simple enforcement of customized contracts. First, they would encourage the parties to make agreements by reducing contracting costs. Such costs may be particularly high because negotiating over divorce rules at the time of marriage introduces contentious issues at a sensitive time. Thus, without statutory standard forms, spouses might be reluctant to make contracts for durable marriage even if such contracts would be enforced. Second, statutory standard forms provide clear directions concerning complicated issues that courts cannot decide with any certainty in case by case adjudication. For example, legislatures might sort out the issue of appropriate remedies for breach, as well as impose conditions such as counseling that deal with some of the paternalism concerns.

Although statutory standard forms offer many of the advantages of covenant marriage, they would be less radical, and therefore more politically acceptable, than inventing a whole new type of marriage in which divorce rights are fettered. Standard forms are offered in a variety of other contexts, from business associations to living wills. Offering standard form contracts also might not be viewed as a sharp break with current law once legislatures recognize the extent to which private contracts could replicate covenant marriage.

In addition to offering many of the same advantages as covenant marriage, durable marriages have some legal advantages over covenant marriage. First, to the extent that durable marriage contracts do not prevent remarriage, they may be less subject than covenant marriage to paternalism concerns. Second, and more importantly, durable marriage contracts may be a more binding constraint against out-of-state divorces than covenant marriages. As discussed in Part III, a covenant spouse can get a unilateral, no-fault divorce in a non-covenant jurisdiction. However, in order to adjudicate the spouses' contract, the court would need jurisdiction over both spouses. Accordingly, the contractual aspects of an ex parte divorce would not be recognized elsewhere.

The discussion so far has focused on the parties' ability to contract effectively under the law of their own state. But the contractual approach also provides an alternative way for the parties to take advantage of more liberal laws in other states. Rather than traveling to those states to marry as required for effective choice of marriage regime, the parties might simply insert a choice of law clause in their agreement. Courts might enforce such a clause even if they would not enforce the contract as a matter of internal legality. Adopting such a form might provide many of the same advantages regarding certainty and legislative judgment that would be offered by a standard form in the parties' home state. Even if the home state does not generally enforce no-divorce contracts, the court might be willing to accept the legislative judgment inherent in a statutory standard form as an answer to paternalism arguments. In other words, courts may be more willing to enforce commitments fashioned by cold-hearted legislators in other states than by the starry-eyed couples alone.

B. Contracts and Same sex marriage

As with covenant marriage, contractual approaches are also available for same sex marriage. Again, these include both enforcement of statutory standard forms and enforcement of contracts. Contracts that replicate marriage for same sex couples would include terms on property ownership, income-splitting and support.[165] They would be analogous to cohabitation (or "Marvin") contracts, now enforceable in most states.[166] Indeed, courts give even polygamous marriage this kind of recognition.[167] The contracts would not be marriages, and therefore would not be subject to proscriptions on same sex marriage.

An important problem with the contractual alternative is that it may be more costly for couples than simply opting into the marriage standard form.[168] Since marriage is a complex long-term relationship, it may be costly to contract for many incidents of marriage, particularly divorce.[169] To be sure, couples already do this in ante-nuptial agreements. But unlike those agreements, a contractual substitute for marriage would have to replicate all of the terms of standard marriage, which are extensive and embedded in the common law. Ambiguities in these terms would not have the advantage of clarifying court decisions. Moreover, without recognition of status it may be uncertain which terms states will enforce. For example, courts may not enforce contracts for short-term relationships because they too closely resemble contracts for sex, or terms that force long-term relationships, such as covenant marriage, because of potential problems of over-optimism at the time of marriage.[170] States that would refuse to recognize contractual incidents of same sex marriage also may refuse to enforce same sex cohabitation contracts as end runs around these prohibitions.

States can reduce these problems in the same way that they can for durable marriage -- by promulgating statutory standard forms.[171] As with covenant marriage, such statutes might reduce transaction costs without going significantly beyond what the parties legally could do without the statute. Some jurisdictions offer "domestic partnership" ordinances or statutes.[172] For example, the Danish registered partnership authorizes a standard form for same sex couples that provides for the contractual aspects of heterosexual marriage, including property in marriage, inheritance, support and alimony, but not rules relating to children such as adoption and custody rights.[173] Such statutes offer many of the advantages of marriage, including secure contract and property rights as well as state-conferred benefits, without compromising the preference-shaping function of giving marriage special status.[174] Indeed, the Vermont supreme court suggested in Baker that the legislature could satisfy its constitutional obligation to offer the same benefits to same sex as to heterosexual couples by authorizing domestic or registered partnerships.[175] Marriage substitutes also can facilitate recognition of same sex relationships by private parties without extending the full panoply of government subsidies. For example, the California domestic partnership law provides only for hospital visitation rights and health insurance coverage for the government employees' dependents.[176]

The above discussion focuses on legal alternatives within a given state. As with covenant marriage, even if state that does not enforce contracts that offer close equivalents to marriage, it might enforce a contract that provides for application of the law of an enforcing state. It also might enforce the local adoption of another state's standard form, subject to the other state's law.[177] This would be a less drastic approach than recognizing the same sex relationship to be a "marriage."� The main legal impediment to this alternative is that a state might apply the public policy exception[178] and refuse to enforce the contractual elements of a same sex relationship just as they might refuse to enforce the marriage itself.� However, the main point of our analysis is that, the public policy exception should have less force as long as states are not called upon to recognize the marriage itself, with all of its symbolic trappings.

C. Unbundling marriage

So far this Part has shown that private contracting offers a method through which the parties might obtain many of the benefits of covenant or same sex marriage. To the extent that this contracting is possible even in states that do not permit same sex or covenant marriage, it supports recognition in those states of at least the contractual elements of those marriages. This subpart distinguishes such contractual elements from other elements, including government-created consequences of marital status, as to which local law should control.� Thus, it seeks to unbundle marriage, and thereby discard the rule that incidents of marriage necessarily follow marital status.[179]

1.� Same sex marriage

Since the same sex marriage issue emerged courts and commentators have considered the various incidents of marriage that recognition of same sex marriage might entail.[180] As the Vermont supreme court said in Baker, "the marriage laws transform a private agreement into a source of significant public benefits and protections."[181] However, marriage is more contractible than might first appear.

The list of marriage benefits in the Hawaii Supreme Court's 1993 Baehr opinion is typical and at least representative of the possibilities:

(1) a variety of state income tax advantages, including deductions, credits, rates, exemptions, and estimates, . . .; (2) public assistance from and exemptions relating to the Department of Human Services . . .; (3) control, division, acquisition, and disposition of community property . . .; (4) rights relating to dower, curtesy, and inheritance . . .; (5) rights to notice, protection, benefits, and inheritance under the Uniform Probate Code, . . .; (6) award of child custody and support payments in divorce proceedings . . .; (7) the right to spousal support . . .; (8) the right to enter into premarital agreements . . . ; (9) the right to change of name . . .; (10) the right to file a nonsupport action . . .; (11) post-divorce rights relating to support and property division . . .; (12) the benefit of the spousal privilege and confidential marital communications . . .; (13) the benefit of the exemption of real property from attachment or execution . . .; and (14) the right to bring a wrongful death action . . .[182]

All of the above categories except numbers 1, 2, 12, 13 and 14 are generally comparable to contractual rights, even if not strictly controlled by contract under current law. As discussed above regarding covenant marriage, the spouses can broadly contract for divorce rights, including spousal support, property division and liquidated damages, [183] and might even be able to affect custody by contract.[184] They can, of course, sue to enforce these rights. They can also contract for many inheritance rights or obviate these rights by inter vivos gifts. Although they may not get an automatic name change from marriage, they can petition for change of name.� Finally, in addition to these rights are others explicitly available by contract, including employee benefits such as group insurance.[185]

States should enforce foreign marriages at least as to these basically contractible elements. In other words, spouses who are legally married under some state's law would be entitled in the state of their domicile to all of the elements of the contractual standard form of marriage under the law of the state of celebration.� That should be so even if the domicile state's law would not make these elements available to locally celebrated marriages.�

The remaining incidents of marriage listed by Baehr to which same sex married couples would not be entitled include tax and other benefits that married people cannot obtain solely by contract.[186] The Baker decision in Vermont lists similar rights, including those to consortium, workers' compensation survivor benefits, spousal benefits for public employees, statutory rights of inclusion in employee group insurance policies, homestead rights and protections, presumption of joint rights of ownership and survivorship, and rights incident to medical treatment of a family member.[187] This allocation of benefits is a critical aspect of marriage law's function of channeling people into relationships favored by society.[188] Some of these benefits may not be very important to most couples. The average couple will never have to be concerned with evidentiary privileges or wrongful death actions. The supposed tax benefits of marriage matter most to one-earner households, since most homosexual couples form two-earner households.[189]

In short, while states would enforce out-of-state contracts, they would retain the power to shape the meaning of marriage by denying some of the government-created incidents of marriage to same sex couples.

2.� Covenant marriage

The main questions regarding covenant marriage under our approach are whether a covenant marriage should be dissolved contrary to the law of the celebration state and whether any such divorce should be entitled to recognition outside the dissolution state. Although, as we have discussed, contracts for durable marriages may be generally enforceable, they are also likely to have unenforceable elements.� Moreover, the parties clearly may not contract to make unlawful sexual relations outside the marriage notwithstanding a purported divorce. Thus, it is important to consider what the rules should be concerning entry and recognition of any such divorce.

It would arguably follow from our approach to same sex marriage that domicile state courts should recognize the contractual incidents of durability. This would include continued support obligations.� The question is whether the domicile state should have to give full-fledged recognition to the covenant marriage, and thereby is precluded from divorcing the couple on petition of the spouse seeking divorce and permitting that spouse to remarry.[190] Such effects might include subsidies and tax breaks that apply to new spouse rather than to the "covenant" spouse and non-application to the divorcing spouse of polygamy and adultery laws. The domicile state might prefer to provide this extra encouragement only to marriages involving currently viable relationships, rather than stressing the need for family stability even in an unhappy marriage. Since the effects of such remarriage may be felt primarily in the divorcing spouse's domicile (as long as the original family is protected under the durability contract), permitting that state to regulate local incidents would minimize spillovers.

V. Evolving Toward a Consensus

The approach outlined in Part IV provides for interstate enforcement of contracts while giving states the option to determine the local non-contractual effects of non-standard marriages. Like all compromises, this approach does not fully satisfy either side of the debate. Social conservatives might want complete recognition of covenant marriage in non-covenant states, including a federal ban on divorce and remarriage in violation of the covenant. They also might favor a federal law banning same sex marriage. At the same time, proponents of same sex marriage might object that our proposal denies homosexual relationships the complete recognition they deserve, with important consequences for same sex couples and for the acceptance of homosexual relationships.� They would therefore urge that recognition of same sex marriage be constitutionally mandated.� But declaring victory for one side would no more settle the marriage debates than Roe v. Wade ended the abortion controversy or Dred Scott the fight over slavery. Thus, we would oppose a Constitutional rule that mandates state acceptance of same sex marriage.[191]

In contrast to an absolutist approach that would prolong the controversy, enforcing choice of law promises to build a consensus and thereby to reconcile the two sides of the marriage debate. First, it is an approach that Americans with divergent views on marriage can endorse because they gain at least as much from the freedom to choose as they lose by granting the freedom to others. This should be apparent from the juxtaposition in this article of same sex and covenant marriage. As repugnant as each side of the marriage debate might find the other's preferred form of marriage, both sides may choose to live by a principle that accommodates both forms of marriage. Same sex couples want a place to get married and that tolerates their lifestyle while some opponents of same sex marriage may want to live where marriage retains its original meaning. Allowing states to withhold non-contractual aspects of marriage from same sex couples or to confer non-contractual aspects on a couple despite their subsisting covenant marriage accommodates both types of preferences. On the other hand, if states had fully to recognize marriages valid in their states of celebration, social liberals would have won acceptance everywhere as to same sex marriage, while social conservatives would win everywhere on covenant marriage while finding no states that preserve the privileged status of heterosexual marriage.

Second, jurisdictional choice promises not only to accommodate competing views but also to widen the acceptance of the various rules within each jurisdiction. Giving lawmaking responsibility to smaller governmental units brings lawmakers closer to the governed, and thereby may lend greater legitimacy to the resulting laws than would be the case for federal laws. Devolution draws on loyalties to local communities that can supplant more polarizing loyalties to larger groups.[192] And local laws can build on local preferences.[193]

Third, permitting states to offer a variety of marriage rules can promote evolution toward consensus through state experimentation. State adoptions or rejections of various approaches can indicate the level of social acceptability of increased durability or same sex marriage.[194] Based on the information provided by other jurisdictions, legislators might decide to expand enforcement by extending greater levels of durability for covenant marriage or by recognizing more incidents for same sex unions. And opposition to full internal legalization may be reduced as judges incrementally narrow the differences between legal and illegal relationships by enforcing the contractual aspects of relationships recognized under foreign law.[195] This evolution through experimentation reflects Hayek's observation that �if left free, men will often achieve more than individual human reason could design or foresee.�[196]

The choice-of-law approach contrasts with the method of achieving compromise prominently espoused by Cass Sunstein.[197] Sunstein relies on "incompletely theorized agreements" through which courts eschew high theory in favor of achieving consensus on narrow principles and particular outcomes.[198]� Courts thereby can let democratic processes operate rather than imposing judicial solutions.� But Sunstein has little to say about how to produce consensus in the face of such fundamental disagreements.[199] Although he relies heavily on the power of analogical reasoning,[200] finding the right analogy may be the key problem. For example, whether society should embrace homosexuality as it has racial equality is at the heart of the debate over marriage.� If the court should select a narrow principle on which people can agree, why should this be that same sex couples can marry despite statutes everywhere to the contrary? And what should courts do when legislatures refuse to subscribe to their principles?[201] A test may come soon in Vermont, where the supreme court explicitly adopted Sunstein's reasoning in deferring to the legislature as to the precise nature of the same sex relationships that should be authorized[202] in the face of a clear indication of legislative hostility to its approach.[203]

The Sunstein approach is, in the end, just a kinder and gentler way of declaring a victor rather than a way of reconciling the two sides of the debate. By contrast, rather than adopting a single, possible arbitrary, compromise, enforcing jurisdictional choice accommodates many different approaches and permits a consensus to develop that reflects peoples' actual preferences.� Democratic processes remain viable at the state level as more than just a way to effectuate the court's pre-ordained result. This approach is consistent with a Burkian emphasis on local and particular institutions of the �little platoons" over top-down rules.[204]

VI. The Constitutionality of Consensus

Several commentators advocate mandating validity of same sex marriage and prohibiting state courts and legislators from banning local recognition of such marriages.[205] This article joins with others that have rejected this absolutist position.[206] In other words, while we propose a choice of law rule that provides for interstate recognition of non-standard marriages and their contractual elements, we would not compel states to adopt this approach.� Thus, states could decide not to recognize any aspect of a foreign same sex marriage, as provided in California's Proposition 22.[207] This Part shows that this approach is consistent with current Constitutional law in this area, and helps point the way toward future Supreme Court decisions.�

Federal constitutional law generally gives states significant power to decide choice of law issues. As the Court has said, "long-established and still subsisting choice-of-law practices that come to be thought, by modern scholars, unwise, do not thereby become unconstitutional."[208] The Court has propounded very broad constitutional guidelines for interpreting both due process and full faith and credit that rely on such elastic and easily met standards as "fairness."[209]

The full faith and credit clause of the Constitution,[210] which requires the states to recognize "public Acts, Records, and judicial Proceedings of every other State," is arguably the most directly relevant provision. This clause gives wide latitude to the states in deciding choice of law issues,[211] and therefore probably does not require a state to recognize the validity of a same sex marriage under another state's law.[212] Marriages almost certainly are not entitled to the higher level of protection the Clause accords public acts or judicial decrees.[213] The full faith and credit clause not only would not protect a covenant marriage, but seemingly would require interstate recognition of a divorce decree even if entered in violation of the covenant.�

The main Constitutional challenge to the choice of law approach concerns the treatment of same sex marriage under provisions intended to protect individual rights.[214] Several commentators have argued that same sex marriage is protected under the equal protection clause of the U.S. constitution.[215] This argument has been accepted by the Hawaii supreme court in Baehr and in dictum by a lower court in Alaska.[216] It is also indirectly supported by the Supreme Court's holdings in Loving v. Virginia[217] that anti-miscegenation laws violate equal protection,[218] and in Romer v. Evans[219] that Colorado could not prohibit anti-homosexual discrimination by the state or subordinate jurisdictions.[220] A series of recent decisions by the Supreme Court of Canada struck down laws that distinguish between same sex and heterosexual couples as contrary to the equality clause of the Canadian Charter of Rights.[221]� However, in Shahar v. Bowers[222] a divided U.S. appeals court indicated that there was no first amendment free-association right to a religiously recognized intimate association between members of the same sex.

This article's analysis bears on these constitutional issues. Equal protection arguments depend on the state's reasons for distinguishing same sex and heterosexual marriage. The law condemned in Loving was intended solely to preserve white supremacy.[223] By contrast, the state may have legitimate interests in preserving the exclusivity of heterosexual marriage that are not based on simple anti-homosexual bias.[224]� Recognizing preference/norm-formation as a compelling state interest provides a basis for refusing to recognize such marriages, even for courts that regard sexual-orientation discrimination as actionable.[225]

With respect to Romer, although the Court implicitly held that some discrimination against homosexuals impermissibly denies equal protection, this applies only to laws singling out homosexuals for a total disability.[226] Mere denial of the right to marry probably is not so fundamental as to approach the total disability involved in Romer.� By unbundling marriage into its contractual and non-contractual components, this article's analysis further clarifies the answer under Romer.� A state that recognizes a foreign same sex marriage and permits the enforcement of the contractual elements of that marriage would clearly not be imposing a total disability even if it denied the spouses the same subsidies and tax breaks accorded heterosexual couples.

Less directly, Romer invites use of a compromise choice-of-law approach. Sunstein characterizes Romer as embodying a "minimalist" approach that seeks to build a national consensus on homosexuality rather than prematurely making the kind of sweeping statement that might trigger a counterproductive reaction.[227] As noted above,[228] Sunstein does not offer a precise mechanism for resolving producing consensus in the face of deep disagreement.� For the reasons discussed in Part VI, private contracts and heterogeneous state regulatory approaches provide such a mechanism. To be sure, the Court retains an important role in protecting fundamental individual rights. However, if both sides of a contentious issue hold rational positions, as this article assumes for same sex marriage, the Court should be sensitive to the potential advantages of state law in achieving consensus.� Moreover, such an approach would be consistent with the Court's current general approach of taking a structural, rather than substantive, role in political decisions. As Judge Wilkinson recently noted, "[a]s a matter of oxen, the gored are determined by infringements upon our federal system, not by judicial disdain for enacted policies."[229]

Even if the Court ultimately decides to Constitutionalize marriage law, choice of law can provide a valuable tool for minimizing disruption. The Court might decide that there is a Constitutional right to same sex marriage, or that full faith and credit does not mandate complete recognition of a decree dissolving a covenant marriage, but only after a state law consensus has developed through jurisdictional choice. Indeed, this may already have occurred with respect to other controversial types of marriage.[230] The first example is inter-racial marriage. In 1952, thirty states, mostly in the West and South, banned multiracial marriages.[231] These states not only refused to enforce multiracial marriages but also imposed criminal penalties. Some, like the Virginia statute involved in Loving, penalized those who married elsewhere and returned to Virginia. Thus, the rule was one of both external and internal illegality. Nevertheless, by 1967, when Loving struck down such laws under the Equal Protection clause, about half of these laws already had been repealed, leaving the bans in place mainly in the South.[232] The shift seems partly due to the growing belief, beginning with a 1948 California decision,[233] that these laws violated the Equal Protection clause. But varying state laws might also have had the effect of a referendum on interracial marriage. As racial attitudes changed, so did the interstate enforcement of the statutes. Ultimately the statutes themselves disappeared. The outcome of the referendum, which was becoming apparent in 1967, made the time ripe for the Supreme Court to reject miscegenation laws.[234] A shift in opinion on same sex marriage indicated by state legal developments might similarly affect constitutional recognition of the right to same sex marriage.[235]

The second example is polygamy. Unlike interracial marriage, no state ever has enforced polygamous marriages. That is not to say, however, that polygamy is inherently irrational.[236] Moreover, jurisdictional choice was briefly available in this context, as Mormons seeking the right to polygamous marriage migrated to the territory of Utah. However, the federal government prohibited polygamy and made Utah's statehood contingent on rejecting polygamy.[237] The constitutional context is at least theoretically analogous to inter-racial marriage, since Mormons had a constitutional argument based on the free exercise clause of the First Amendment for being allowed a form of marriage that was central to their religious beliefs.[238] But unlike inter-racial marriage, no state was willing to "vote" for polygamy. Although federal law ultimately precluded effective competition, Congress, like the Court in Loving, might have acted otherwise had states not rejected polygamy so forcefully.

The jury is still out on same sex marriage, which still is not recognized anywhere. Notably, there is a kind of competition with respect to non-recognition of same sex marriage. Thus, conservative interest groups can counteract at the legislative level what homosexual rights groups win at the judicial level. Although most states have passed statutes providing for non-recognition,[239] there are significant variations in the statutes[240] and a significant minority of the states has not passed the statutes. It is no accident that the latter are the more politically liberal jurisdictions.[241] Thus, even at this early stage, the states already appear to be sorting on the issue of external legality.

Finally, it is important to keep in mind that the choice of law approach is fully consistent with state constitutional and statutory responses to the marriage debates, including recognition of same sex marriage.[242] Denial of the right to same sex marriage was held to violate the right to privacy under the Alaska constitution.[243] The Baker decision in Vermont was based exclusively on a state constitutional provision promising "common benefits" to all Vermonters.[244] A state constitutional decision preserves the viability of jurisdictional choice.�

VII.� Concluding Remarks

The debate over covenant and same sex marriage is marked by the clash of irreconcilable ideas about social institutions. This Article suggests a compromise based on interaction among different jurisdictions. State courts that do not recognize covenant and same sex marriages under local law might reasonably enforce the contractual aspects of such marriages when they have been validly entered into under the laws of another state. Concerns for paternalism, over-signaling, the coherence of internal rules, and the law's expressive or symbolic effect need not be accorded as much weight when it comes to recognizing the contractual elements of foreign marriages as when states make internal laws.� At the same time, states should be able to control the local subsidies and tax and regulatory effects of foreign marriages. Individuals could thereby choose both the marriage rules that apply to them and their preferred cultural environments.�

This approach does not satisfy everybody. Because symbols matter, those opposed to covenant or same sex marriage would object to any local enforcement, while those in favor would object to any limitations. But this article offers a compromise that attempts to achieve consensus rather than awarding complete victory to either side through a Constitutional rule.� If the Court must ultimately take a stand, consensus-building based on jurisdictional choice is preferable to support-maximizing judicial politics Cass Sunstein advocates.

This approach has potential application to other legal issues that invoke clashes of beliefs similar to those relating to marriage. Numerous public debates, including those over abortion, parental rights[245] and the right to die, would gain from preserving jurisdictional choice rather than intervention by an absolutist, winner-take-all, Supreme Court decision.� We should recognize the virtues of deferring to a multitude of state laws in debates where neither side is likely to be happy with a single rule.