You are not the marginal case

Jane Galt has a great post about gay marriage. She refuses to pick a side, but makes some analysis that is well worth considering for anyone considering the feedback between legal and cultural institutions:

I would never become an unwed welfare mother, even if benefits were a great deal higher than they are now. It seems crazy to even suggest that one would bear a child out of wedlock for $567 a month. Indeed, to this day, I find the reformist side much more persuasive than the conservative side, except for one thing, which is that the conservatives turned out to be right. In fact, they turned out to be even more right than they suspected; they were predicting upticks in illegitimacy that were much more modest than what actually occurred--they expected marriage rates to suffer, not collapse.

How did people go so badly wrong? Well, to start with, they fell into the basic fallacy that economists are so well acquainted with: they thought about themselves instead of the marginal case. For another, they completely failed to realise that each additional illegitimate birth would, in effect, slightly destigmatise the next one. They assigned men very little agency, failing to predict that women willing to forgo marriage would essentially become unwelcome competition for women who weren't, and that as the numbers changed, that competition might push the marriage market towards unwelcome outcomes. They failed to forsee the confounding effect that the birth control pill would have on sexual mores.

When you push a dynamic system, sometimes it keeps going. Change is not always bad, but analysis that ignores dynamics sure is. Small changes can have big effects - so don't make the changes unless you want the effects.

Share this

Grant - A little decorum

Grant -

A little decorum please! You made the rest of your response very difficult to take seriously, or even to not ignore. The thing is, you may have some very good points in the rest of your comment, but the first sentence has so colored my thinking about your comment that evaluating your argument on its merits is very difficult.

Patri Friedman, I am

Patri Friedman,

I am right.

The "sanctity of marraige" spiel often goes on about the effects on marriage, society, etc. Just note the late Pope's words on the subject: calling homosexuality an "ideology of evil" which seems to wage war on humanity and the family.

Of course also note that Ayn Rand found homosexuality disgusting and that Randroids as a rule have some fairly puritanical views about sex.


Hus - I really think you are

Hus - I really think you are missing some very important analysis. The world is not always how we want it to be. Like Galt, I am very sympathetic to the liberal position. Yet it is crucial not to let that sympathy get in the way of accurate analysis of the effects that changes in these laws have on society.

I'm not sure whether she is right - it would take statistical analysis of empirical evidence to convince me. And even if she was, I might personally value liberty enough not to care about the outcome. But I am sure that she has an important general point, which is that small changes in systems can produce large changes in output, and not always those which we expect.

I am afraid the writeup's

I am afraid the writeup's analysis is just a dressed-up version of the same rhetoric about preserving the "sanctity" of marriage seen from religious adherents who object to gay marriage.

Of course the best solution would be to get the government out of the marriage perference business entirely.

Hus - I fully agree with

Hus - I fully agree with your idea of the best solution. That said, I'd like to see the specifics of why you believe the argument is a dress-up of the 'sanctity' argument. Using empirical evidence of similar situations and deductive reasoning to show why one applies to the other seems to be a pretty rock solid argument to me. Jane's conclusion is not the same as the 'sanctity' argument.

This post is definitely

This post is definitely thought-provoking and important to read, but I think some of the examples - and I would think, many other possible examples of this phenomenon - suffer from a bit of the ol' post hoc fallacy. That is, a set of changes to divorce laws happened, and afterward divorce rates hit 50%. Therefore, the change to divorce laws caused divorce rates to hit 50%. This argument alone is not valid.

While I'm sure it can be demonstrated that many of today's divorces fall under the 'no-fault' category of divorces, that merely proves that the legalization of no-fault divorce is a necessary condition for current divorce rates - not a sufficient condition.

I suspect that many things besides a few pieces of legislation have affected the divorce rate over the past half century. Perhaps one particular piece of divorce legislation was a necessary condition for ALL of them (i.e., it started a chain reaction that caused every other factor to materialize) but that would be very difficult to demonstrate. It is therefore very difficult to distinguish between cases where the phenomenon JG describes actually happened, and cases where a post hoc fallacy can be applied to make it look like it happened.

It can be said that cultural institutions are "organic", that is they develop without being directed by any particular individual or group, but are influenced by their environment - which includes laws. However in a democratic society, the causation flows the other way as well - shifts in culture cause legislative changes. Suppose political efforts in opposition to no-fault divorces had been more successful, and prevented such bills from being passed. Would that have stopped the overall push for less restrictive divorce laws, or merely slowed it down? I suspect the latter.

Similarly, if that amendment to limit the income tax to 10% had been passed, how long would it have lasted? I seriously doubt it would have survived the Great Depression or World War II.

Society is a complicated, intertwined thing, and any attempt to pin down single causes for such broad attributes of society as "the divorce rate" or "the rate of income taxes the population is willing to tolerate without revolting" is futile. The phenomenon Jane Galt describes probably does happen, and it's a good snapshot of society's overall complexity. But I seriously doubt we'll ever be able to see the phenomenon coming; that is, we won't be able to know which laws will cause disastrous cultural chain reactions, which will cause beneficial cultural chain reactions, and which won't have much effect at all.

Micha wrote: "I don’t

Micha wrote: "I don’t even know how to begin to respond to that."

Of course you don't, you're a weak-willed, wishy-washy person, Micha. :-P

I'm not sure which parts of

I'm not sure which parts of Grant's post are tongue-in-cheek and which parts are to be taken seriously. If the claim is that there is no such thing as a difficult decision, with good reasons on both sides, and that all who fail to immediately see the correct answer are "wishy-washy" and "weak-willed," well, I don't even know how to begin to respond to that.

Trust an economist to trust

Trust an economist to trust the margins. "The margins knew," says the economist, "what we did not: That [change x] would cause [consequence y]. Had we but listened to the margins!"

Fuck off.

There are two problems with this argument, one concrete, one general. The concrete one is that there are lots of margins. Consider Ms. Galt's charaterization of easing divorce restrictions. She considers the margin at which an epsilon-short-of-divorcing couple suddenly can divorce. And that margin tells us that we will get more divorces. But let me consider a different margin.

In New York at least and probably many other places prior to divorce liberalization there was in industry of professional divorce procurers. These were people who would counsel you on how to invent plausible-yet-compellingly-awful stories of abuse, who would drug your partner and take pictures of him or her in compromising positions, who would bribe judges to get them to accept divorce petitions. You know what happened at this margin? Divorce reform kicked its ass. The notion that paying people to arrange fraud and perjury was a necessary social arrangement for an orderly and happy society got a serious kick in the teeth.

At one margin, divorce reform worsened things. At another margin, it improved them. Frankly, I'm happy with the outcome -- the rule of law won in a big way. So point 1 is, there are plural margins, and when an economist speaks of "the marginal case" as if there were only one margin, he or she is either stupid or, more likely, attempting to defraud you.

Point number two is more general: The margin always advocates tyrrany in retrospect.

There is this odd little phenomenon about margins, particularly in the case of social rather than economic changes: They consist of wishy-washy, weak-willed, undecided people. Why can't these folks just make up their minds, eh? Divorce, to pick Ms. Galt's example, is a big step. What right-thinking person would ever be balanced on that margin?

What the marginal case needs, in each an every case, is for the firm hand of the state to set him right. We after all "know" what the right choice is, and there's no reason not to just impose it. At the margin, nobody will make good use of their rights -- if they used their rights well, they wouldn't be marginal. Free speech's margin is a bunch of ill-informed, unmotivated whiners. Gun control's margin is people who didn't really need that gun anyway. Trial by jury's margin is scumbags who would be in prison if not for twelve marginally-ignorent peers. The marginal case for your social reform always cries out for tyrrany.

Of course, a real economist spots the flaw in this argument instantly: Tyrrany has different margins than social reform. But this is a silly argument, hinging on tyrrany as a separate object from rights. It is not. Resolve every hard case by the margin, and you will arrive at tyrrany without ever having considered the margins of increasing tyrrany. I seem to recall that Mr. Hazlitt has a nice little motto for this sort of difficulty in planning.

Fond though economists may be of the marginal case, it is in matters of rights and customs nothing but a consequentialist roadmap to statist tyrrany. The margin is seldom improved by having any rights, and the margins that are are seldom the ones visible in any particular debate. Look not to the margins in complex questions: They will lead you astray.

We should also look out for

We should also look out for unforseen large positive consequences, a point often lost in the "conserve the status quo for its own sake" type arguments.

Hus - "I am right" is not

Hus - "I am right" is not much of an argument. In fact, its not an argument at all, its practically a meaningless utterance.

Nor is the fact that people make bad arguments that changing marriage has bad consequences a proof that there are not good arguments to that effect. If you wish to reason logically, you need to reply to the actual argument being made, not a totally different argument trying to prove the same thing.

I am bisexual, tattooed, pierced, and drive a car with the license plate "FRRREAK" - so its not like I have any respect for religious anti-gay bigotry. But Galt's argument is entirely different, and I think worthy of consideration. As Jon points out, she may well be wrong - which is why I would need more careful statistical analysis to believe it. But I think its an argument well worth considering, not one to be rejected out of hand like "Homosexuality is a sin, Marriage is sacred to god"-type bullshit.

Patri Friedman, Yes, I am

Patri Friedman,

Yes, I am quite sure that any change which enhances liberty can be viewed this way. There are indeed consequences to greater liberty; periods of transistion, etc. where some costs may be high. But I hardly this as a reason to stop its advance.


BTW, why am I getting so

BTW, why am I getting so many run time errors from your blog? :wall:

I don’t think that laws

I don’t think that laws should be made based on moral consequences, but on rights. I believe that marriage and intact families have social benefits, but even if it could be proven that homosexual marriage had a negative effect on these, I don’t think that it would be sufficient reason to outlaw it.

Just like with Galt’s other point, birth control. I think that the introduction of birth control has had several negative ramifications on society, and am morally opposed to it. This means that I would never use birth control, not that I would ever support a law against it or to regulate its use by consenting adults.

I think that an argument of this nature could only convince someone not to engage or to encourage others engage in homosexual marriage, not convince someone that force should or could legitimately be used against someone who would like to engage in homosexual marriage.

Thea, There is also the fact


There is also the fact that government discrimination against an activity (be that discrimination based on criminal sanctions or something else) tends to drive behavior underground. There is a cost to this as well as we see readily in the drug war.


Hus- In addition to the


In addition to the costs, the activity often becomes more dangerous; again look at the drug war.

However, I'm not sure that this alone is ever a reason to not outlaw something that is a violation of rights. For example, saying that hit men are more expensive when murder is illegal is not a compelling reason to legalize it.

The reason that drugs, contraception, and homosexual marriage should not be illegal is because it wrong to use force against someone for doing something you disagree with for whatever reason unless it's violating your rights.

I don’t think that laws

I don’t think that laws should be made based on moral consequences, but on rights.

I agree, but what does this have to do with homosexual marriage, or with no-fault divorce? How is it a violation of anyone's rights for the state to decline to sanction certain relationships or to refuse to dissolve a contract entered into voluntarily?

Gay Marriage and "The

Gay Marriage and "The Margin"
Jane Galt has a confused, rambling non-sequiturfest at Asymmetrical Information in which she essentially calls on same-sex marriage advocates to "chill out" --
My only re...

I agree, but what does this

I agree, but what does this have to do with homosexual marriage, or with no-fault divorce? How is it a violation of anyone’s rights for the state to decline to sanction certain relationships or to refuse to dissolve a contract entered into voluntarily?

I'm sorry if my previous statement was ambiguous. I'll try to clarify, if we are assuming that the state is going to be enforcing laws, I think that those laws should only be regarding rights violations, i.e. the government would only have the power to step in if someone's rights are being violated. Therefore, since homosexual marriage nad divorce do not a violate anyone's rights, I think that it should not be illegal.

Therefore, since homosexual

Therefore, since homosexual marriage nad divorce do not a violate anyone’s rights, I think that it should not be illegal.

But it's not a question of whether or not homosexual marriage should be legal. There is no serious support anywhere for a Straightness Enforcement Agency to conduct midnight raids on the homes of gay couples.

Rather, it's a question of whether or not the state should grant certain privileges to homosexual couples who jump through a certain set of hoops, just as they do for heterosexual couples now. Whether or not the state should do so is a legitimate question, but they aren't initiating force, violating rights, or "stepping in" by not doing so. In fact, it's the advocates of legally sanctioned homosexual marriage, not its opponents, who are asking the state to step in.

Same deal with no-fault divorce. As far as I know, there were never any laws against married couples dividing their property or living apart. Again, it was the advocates of no-fault divorce, not its opponents, who were asking the state to step in and assist with these arrangements. Legal marriage is essentially a three-way contract between the bride, the groom, and the state. A couple entering this contract voluntarily can't complain that their rights are being violated if they don't like the terms.

You can argue that the state shouldn't step in to sanction marriage in the first place, and I agree. But once we accept that the state is going to do so, discussion of the details must necessarily be consequentialist, because rights go out the window when you enter into a contract.

Brandon Berg, Quite

Brandon Berg,

Quite obviously you've never read the platform of the Texas Republican Party. It specifically calls for the re-introduction of anti-sodomy laws, for laws which prohibit unsupervised contact between homosexuals and children, etc.

Which plank is that, specifically?

Brandon, The linked page is


The linked page is simply the preamble. On Page 11 of the PDF there are planks regarding sodomy & homosexuality, which includes an explicit call for denial of custody to homosexuals (as part of the "including but not limited to the following"). It also calls for revocation of broadcast licenses for FCC violations, prohibition in toto of any/all forms of pornography (as if none of them ever looked at Playboy- and if they didn't, how do they call themselves Texans?).

Though a prohibition is not exactly the same thing as saying "we want a dept. of straightness" to actively police the population and generate this outcome. I don't think they want that; they want it on the books so that when they *want* to legally smack around a homosexual, they can. Witness how often or with what fervor they went about enforcing the anti-sodomy laws beforehand (instances are few and far between). Certainly they didn't lock up hetero couples for engaging in sodomy, nor have they censured their buddies for looking at porn or tit mags. Its just to put a club in the law's hands to use against people they don't like.

Interesting, Brian.

Interesting, Brian.

Despicable as well.

Maybe I'm too dumb to grasp

Maybe I'm too dumb to grasp the profundity so many of the commenters on Galt's blog seem to see in this post. Once someone starts talking about "societal institutions", they usually seem to be describing something that needs to exist, for its own sake, that is more important than the individuals who make it up. The people who are getting married exist for the sake of the institution, not the other way around. Doesn't Galt's argument against changing societal institutions pretty much boil down to "Your choices should be restricted, because I fear the unforseen consequences of whatever you might choose to do"? That reasoning could apply to just about anything. After all, one poster complained that those libertarians are always talking about how we should legalize drugs without thinking about how you don't really know what the consequences of doing so will be. You might not become a total junkie because drugs were legalized, but think about the marginal case! This argument seems to end, essentially, with the conclusion that people's choices should be as restricted as possible because you can never tell what the consequences will be.

The linked page is simply

The linked page is simply the preamble. On Page 11 of the PDF there are planks regarding sodomy & homosexuality, which includes an explicit call for denial of custody to homosexuals (as part of the “including but not limited to the following").

Well, there's Texas for you. I stand corrected.

Still, the "gay marriage" question on the national agenda is whether the state should sanction homosexual marriage, not whether or not sodomy should be legal.

Brandon Berg, My point is

Brandon Berg,

My point is that the threat is hardly as chimerical as you paint it.

Doesn’t Galt’s argument

Doesn’t Galt’s argument against changing societal institutions pretty much boil down to “Your choices should be restricted, because I fear the unforseen consequences of whatever you might choose to do"? That reasoning could apply to just about anything.

It is not about unforsoon consequences, at least as I read it. It is about trying to honestly foresee consequences which can be unpredictable, and much larger than you might expect. Of course we could hand-wavingly conjure up potential bad consequences for any change, and to let that deter us would be the definition of extreme conservatism. But sometimes changes that have obvious small positive benefits do have large negative consequences, and we should watch out for them.

Three Arguments Against a

Three Arguments Against a Bad Argument Against Same-Sex Marriage
Three Arguments Against a Bad Argument Against Same-Sex Marriage Through Patri Friedman's Catallarchy post on the topic, I found Jane Galt's treatise on how changes at the margins might augur against allowing same-sex marriage. There are several flaw...