Bryan Caplan  

The Coase Theorem and Divorce; Or, the Familial Economy of Hate

Defending Market-Value Account... Is "Contagion" the Result of S...

In The Logic of Life, Tim Harford repeats an argument about divorce that I've often heard economists make. One of the challenges that Gary Becker tackled, he explains, is the sharp increase in divorce. Tim goes on:

Some commentators have blamed changes in divorce laws for the trend: Ronald Reagan, then governor of California, signed a bill introducing 'no-fault' divorce in 1969, meaning that either partner could simply walk away from the marriage by demanding a divorce... But Becker knew that couldn't be the answer: If the husband wanted a divorce to run off with his mistress, no-fault divorce didn't make it easier for him to do that, just cheaper. Before no-fault divorce, he had to get his wife's agreement, which might mean higher alimony payments. This reasoning suggests that no-fault divorce rules wouldn't change divorce rates at all. The only thing that would change was who paid whom to get the divorce. (emphasis mine)
This is of course the application of the Coase Theorem to divorce: If a couple's total surplus is higher apart than together, then with zero transactions costs, the assignment of property rights only changes distribution, not outcomes.

Tim argues that the empirical evidence supports this view. Perhaps it does. But most economists who make this argument have no idea what the data says; instead, as Tim says of Becker, they "know" that changing divorce laws "couldn't be the answer."

Unfortunately, the Coasean argument overlooks a pretty obvious fact: Couples contemplating a divorce often hate, loathe, and despise each other. We've all heard of stories of divorcing couples deliberately destroying objects of sentimental value to each other. Indeed, many couples in this situation wallow in petty spite; they can't stop bad-mouthing each other to anyone who will listen.

With these facts firmly in mind, how confident are you that Coase's zero transactions costs assumption is remotely true? At risk of sounding Austrian, transactions costs are subjective: Bargaining with your mortal enemy hurts. And while you might think that people would be eager to part from their mortal enemy, many decaying marriages contain at least one person eager to torture the other for as long as possible. (Yes, it's ugly, but don't blame me - I'm only a messenger!)

Maybe no-fault divorce didn't matter. Maybe divorce law never matters. But I'm skeptical. How about you?

Comments and Sharing

COMMENTS (16 to date)
Daniel writes:

How about hiring a divorce lawyer? In marriages with a lot of assets, I'm sure quite a lot of money is spent on legal fees. That's about as straightforward and literal a case of transaction costs as I can imagine.

Also, getting a divorce may mean moving out of a house which you own into someplace that you have to rent. If one party keeps the house after the divorce, but the other party is paying rent on a new place, between the two of them they're spending more on housing. Even if a couple doesn't own their home, splitting up will usually mean that between the two of them they have to pay 2 rents, rather than just 1.

It seems to me that there are lots of ways that divorce can involve transaction costs that are far from negligible; this seems like a poor candidate for an application of the Coase Theorem.

Daniel writes:

Wait, my post was beside the point. Sure there are transaction costs associated with divorce; the question at hand was whether no fault divorce decreases those transaction costs. The stuff I said above doesn't bear on that.

Nevertheless, I do think you're right; having to bargain to obtain the permission of somebody you're in a rotten marriage with seems sufficiently unpleasant that it constitutes a non-negligible transaction cost that makes the Coase Theorem inapplicable here.

undergroundman writes:

Although most people don't include perfect information in the conditions of the Coase requirement, it seems like a requirement to me. The husband doesn't really know what the wife will demand for a divorce, and the wife doesn't really know what the husband is willing to pay. In cases of pollution this knowledge requirement can be a huge barrier: I have no idea how much damage smog does to my lungs, and I can't even begin to estimate the long-term costs of global warming on my well-beinng. These information barriers can lead to a deadlock.

Overall, though, you're certainly right in this case.

FXKLM writes:

I think the bigger problem with applying the Coase Theorem is that the wife in that case has no way of paying the husband to stay if he's willing to walk away from his share of the couple's assets in a no-fault divorce.

Suppose that in a no-fault divorce, either party can choose to divorce and keep 50% of the couple's assets. The wife is indifferent between staying married or leaving the marriage with 80% of the assets. The husband is indifferent between staying married or leaving the marriage with 40% of the assets. In a no-fault divorce state, the husband can choose to leave the marriage and keep 50% of the assets.

The Coase Theorem would suggest that the wife could pay the husband 10% of the couple's assets to entice him to stay. But where do those assets come from if not from the assets of the couple? It doesn't really improve his situation at all.

It's possible that we're talking about a community property state. In that case, the couple would have community property that would be split on a divorce plus each would have their own separate property. In that case, it's possible that the wife could transfer those assets from her separate property to her husband's separate property. But after that payment, what's to stop the husband from getting a no-fault divorce and still taking half of the community property? Can they sign a post-nuptial agreement making divorce financially punitive for the husband? You need to assume a lot of other legal rules for the Coase Theorem to apply to no-fault divorce.

The only way the Coase Theorem would prevent no-fault divorce from increasing the number of divorces is if couples are somehow able to make side payments to keep each other in the marriage if (i) one party is willing to leave for less than 50% of the couple's assets and (ii) the combined reserve price for both parties is greater than 100% of the combined assets. Even aside from the legal issues, that's just not going to happen emotionally. The Coase Theorem is completely inapplicable here even if you ignore the spite factor of divorce.

FXKLM writes:

That last comment was fairly long. Here's the shorter version: The Coase Theorem states that, wherever the law initially assigns assets and liabilities, parties will trade among themselves to most efficiently distribute those assets and liabilities. Assuming the total utility from divorce is higher than the total utility from marriage, a couple will use side payments or work out some sort of agreement where both parties are willing to divorce rather than stay married. But if the total utility from marriage is higher than the total utility from divorce and one party can get a better deal from no-fault divorce than from staying in the marriage, there is no practical way to structure side payments to keep the marriage together. And since the total utility from those marriages is positive, the extra divorces that result from no-fault divorce are value destroying.

Tobbic writes:

I agree that if the total surplus of the couple is greater apart than together both forms of divorce lead to divorce.

Another situation: the total surplus of the couple is lower apart than together but the one party (say the husband) gets a lot of more surplus when apart. If I understand correctly in the no-fault divorce husband could just decide to walk away unilaterally to get the higher surplus. The wife could of course trade some of the in-marriage surplus to retain the marriage. But how realistic is it that the wife pays to the husband so that they will stay together? If this is not realistic, the husband will unilaterally divorce. However, in the 'old divorce' case there is no quantity of money that the husband is willing to pay that will make both of them better off and thus there's no divorce.

Stuart Buck writes:

Assuming zero transaction costs isn't really "applying" the Coase Theorem. In other words, I think Coase's real point was not that people should analyze real-world problems by pretending that transaction costs don't exist.* Instead, I think his point was to show that 1) economists had been ignoring transaction costs; but 2) if there were really no transaction costs, the law would have no purpose, which doesn't seem to be the case in most instances; therefore, 3) you should look for transaction costs when analyzing a problem.

* This means that if you do find transaction costs, that doesn't mean that the "Coase Theorem doesn't apply here." It means that the Coase Theorem does apply -- Coase said that where you find transaction costs, then the initial assignment of rights does matter.

Scott Wentland writes:

Shouldn't the surplus be higher for higher transaction cost cases? If they are your "mortal enemy", then you'd have a much higher surplus being with someone else than if you were indifferent. The exception may be that your torturing the other person actually gives you a higher surplus (i.e. you'd rather bring down your enemy than love someone else). Even if this is the case, there might be an opportunity cost for these kind of people. There might be someone else they'd enjoy torturing even more. Moreover, if your torturing gives you a higher surplus, it only does so because it brings down the other person. This should be a wash in terms of surplus too (from torturing, if their utility goes down 1, yours goes up 1). The numbers may not be one for one, but it'll likely be pretty close.

So, the higher transactions cost should (for most cases) be a wash because the surplus should increase too. Transaction cost need not be zero here for Coase to be applicable, considering the relationship between surplus and transaction cost.

P.S. Empirically, Wolfers is pretty convincing too.

Mason writes:

In a no-fault divorce, do both parties have to agree to the divorce? Not according to Wiki.

“If a couple's total surplus is higher apart than together, then with zero transactions costs, the assignment of property rights only changes distribution, not outcomes.”

But no-fault divorce didn't reduce transactions costs, it changed property rights.



Total surplus is higher together, but one party has higher surplus apart, so there will be a divorce; the outcome has changed.

"If the husband wanted a divorce to run off with his mistress, no-fault divorce didn't make it easier for him to do that, just cheaper. ... This reasoning suggests that no-fault divorce rules wouldn't change divorce rates at all."

That would be like saying, “while the new Wal-Mart is cheaper than Giant, I still have to get in the car, go to the store, and walk through the isles, so because of that I won’t buy anything differently than I would have at Giant, even though buy shopping at Wal-Mart I can afford that tasty tart that I couldn’t afford at Giant”

Making divorce cheaper will increase the quantity consumed unless it’s a Giffin Good, but I don’t see why it would be.

John Fast writes:

I must be overlooking something here, probably because I'm a political theorist rather than an economist. As far as I can see, no-fault divorce did lower the transaction costs of a divorce in several ways: time and money spent for the legal transaction, and also social opprobrium.

This topic is particularly fascinating because it involves /s/o/c/i/o/b/i/o/l/o/g/y/ evolutionary psychology -- read Gilder's _Sexual Suicide_, for example, about the social and psychological effects of easier divorce -- and also involves the dark side of game theory, where players hate each other so much that they value the suffering of their opponents more than they value their own success in the game (by objective measurements).

BTW the Dark Side of Game Theory also comes from evolutionary psychology; it's the part of the psyche which makes us feel vengeful when we feel our boundaries (preferably set by Schelling points) have been violated.

What kind of work has been done on those sorts of games?

8 writes:

If the husband wanted a divorce to run off with his mistress, no-fault divorce didn't make it easier for him to do that, just cheaper.

It's been a few years since I studied Coase, but since when does something being cheaper result in no change in behavior?

If a couple's total surplus is higher apart than together, then with zero transactions costs, the assignment of property rights only changes distribution, not outcomes.

If the wife has a boyfriend or really loves money, then it is possible to have the same outcome.

ajb writes:

Mason's example only works if there is no proper bargaining (which is another way of saying tc are too high). Otherwise the wife would bribe the husband to stay in the marriage. In the Victorian period, this meant (for example) her letting the husband have a mistress. If we assume this sort of bargaining away, then certainly the Coase theorem doesn't hold and no-fault divorce changes outcomes.

The interesting issues others point out are the cases when both parties purposely diminish the joint surplus out of spite.

James B. Shearer writes:

In my opinion the argument is insane and the sort of thing (absurd real world conclusions confidently obtained from inappropriate theoretical models) that gives economists a bad name.

James A. Donald writes:

In those few divorces that I know what happened, it seems clear the couple's total surplus is not greater apart. At least one person was harmed far more than the other person benefited. The ending of the marriage was always a result of a destructive act of betrayal. The divorces resulted from sexual infidelity combined with massive financial fraud. While it is common for both parties to behave extraordinarily badly, it is usual that at least one party behaved extraordinarily badly.

Eric H writes:

I would like to amplify something said by Stuart and Scott. The popularized version of the Coase Theorem -- as distinguished from Coase's actual article -- may require zero transaction costs, but this is definitely not Coase's view. About 1/3 of the way through The Problem of Social Cost, he said,

The argument has proceeded up to this point on the assumption that there were no costs involved in carrying out market transactions. This is, of course, a very unrealistic assumption.


Once the costs of carrying out market transactions are taken into account it is clear that such a rearrangement of rights will only be undertaken when the increase in the value of production consequent upon the rearrangement is greater than the costs which would be involved in bringing it about. When it is less, the granting of an injunction (or the knowledge that it would be granted) or the liability to pay damages may result in an activity being discontinued (or may prevent its being started) which would be undertaken if market transactions were costless. In these conditions the initial delimitation of legal rights does have an effect on the efficiency with which the economic system operates.

The remainder of the essay emphasizes the effect of transaction costs, as I recall. It is unfair to saddle Coase with the world of zero transaction costs.

SheetWise writes:

Marriages involve three parties, and the third party (the State) is frequently irrational. Getting a divorce can be a lot like playing Russian Roulette, which makes it pretty hard to determine the transaction costs.

Consider gay "marriages" which have been taking place for years, but are unofficial. The government hasn't inserted itself into the contract, so the transaction costs of a breakup are nearly zero. They're far less stable than marriages involving the government.

Hardly conclusive -- but a legal marriage ending in divorce is always negative sum when considering property only.

Comments for this entry have been closed
Return to top