Premarital agreements are rare. This observation used to dismay the late Nobel laureate George Stigler: He maintained that the grand institution of matrimony is demeaned by those who can't be bothered to negotiate its details.
A marriage is a contract. You can write that contract yourself (in which case it's called a "premarital agreement"), or you can accept the default contract written by your state legislators. Now comes the state of Louisiana, determined to expand its citizens' options. Henceforth, Louisianians will be able to choose between two prefabricated contracts, each with very different provisions for divorce. The first option is similar to the no-fault contract that is standard in other states. The second--the so-called "covenant marriage"--makes divorce far more difficult.
Even if you never divorce, your choice among contracts can affect the entire course of your marriage. That's because the possibility of divorce alters your incentives to keep your spouse happy (and vice versa). Of course, you might want to keep your spouse happy for other reasons, the most notable of which is love. Sometimes, love is all you need. But because we're talking about divorce law, I want to focus on cases where love is not enough--and in those cases, to ask which contract provides the best incentive for good marital behavior. The answer may not be what you think.
While we're at it, let's compare three kinds of marriage: a no-fault contract (where either party can obtain a divorce on demand), a mutual-consent contract (where both parties must agree to a divorce), and a covenant marriage (where even mutual consent is not enough). You might think that no-fault marriages are always the most likely to end in divorce. That isn't true, and here's one reason why: A lot of marital issues are negotiable--like who should do the dishes, who gets to operate the remote control, which one wears the anti-snore device and which one wears the earplugs, and so on. Here the negotiating process itself provides all the right incentives to respect your spouse's needs. What you won't do for love, you'll still do for a bribe. And those things you won't do even for a bribe are, presumably, sufficiently distasteful that you shouldn't do them.
Bribery works equally well under no-fault and mutual consent (though the choice of contract alters the balance of power and therefore might alter the size of the bribes). Under either system, the marriage survives as long as it's possible to keep both partners happier together than they would be apart. Therefore, the two systems produce the same number of divorces. (If you're not convinced by that argument--which is a special case of a general principle that economists call the Coase theorem, click
On the other hand, if you're in a covenant marriage--where you can't get a divorce even by mutual consent--divorce might be impossible even when the marriage turns bad for both of you. If we assume that all marital conflicts are negotiable, the covenant marriage has no offsetting advantages: It keeps couples together only in those cases where they'd both be happier apart.
The analysis changes if there are important decisions that can't be negotiated, like the decision whether to bring home a surprise bouquet of flowers. Chronic thoughtlessness on such matters can cause a marriage to deteriorate. The knowledge that divorce is impossible might make you strive harder to avoid such deterioration--and it might do the same for your spouse. In that sense, a covenant marriage is like the old nuclear-war Doomsday Machine: You are each on notice that you'd better work hard to preserve a good marriage, or you'll both be forced to live your lives in a bad one. Doomsday Machines can be very effective. But sometimes they blow up. So the covenant marriage is a mixed blessing.
It's the issues you can't negotiate that make the covenant marriage worth considering. But that same inability can make no-fault marriages the strongest of all. In a no-fault marriage, a happy spouse will treat you well to prevent your leaving. That gives you an incentive to keep your spouse happy. And this process feeds on itself: Your spouse works to make you happy, which makes you want to preserve the marriage, which makes you work to make your spouse happy, which makes your spouse want to preserve the marriage, and so on, in a great virtuous circle.
By contrast, if divorce required mutual consent, your spouse could accept your efforts to make him or her happy without feeling a strong need to reciprocate. This prospect discourages you from bearing gifts in the first place. But when either partner has the power to end the marriage, kindness tends to be repaid with kindness, and therefore kindness thrives. Notice, once again, that this analysis applies only to surprise efforts. Efforts that are negotiated in advance can be negotiated equally well under any contract.
So here is the bottom line: When marital issues are negotiable, we are in the domain of the Coase theorem, where no-fault and mutual consent do equally well and where covenant marriage is always a mistake. But when important issues can't be negotiated, both the covenant marriage and the no-fault contract become more attractive, for different reasons.
This analysis is far from exhaustive, and I know from much recent experience that Slate readers will forcefully call my attention to scenarios I've failed to consider. Let me pre-empt them and go a step further by pointing out a basic question I've ignored: How does a change in the marriage contract affect a couple's decision about whether to get married in the first place? There's a lot of interesting economics in that question, and if I manage to sort it out, I'll let you know.