55 Comments
User's avatar
Sean Hazlett's avatar

Marriage before cohabitation seems like a much more deliberate decision. It's very easy to just fall into a marriage from cohabitation, even if signs of incompatibility are present

Mr. Doolittle's avatar

I had some college friends who dated through high school and for years after. They dragged out a long engagement and ended up canceling their wedding like a day before. She met another guy and got married almost immediately, and has been extremely happy. He seems much happier since as well.

They weren't right for each other, but also didn't want to cut things off because there was never a specific reason to do so.

Age of Infovores's avatar

Re: the burnout factor, I think the fact that cohabitors have more time living together than a non-cohab couple that has been married the same number of years actually tilts the odds of divorce horse race in the cohabitors favor. When people live together for a few years before getting married there is time for the most likely to divorce couples to leave the sample and not even be measured

David Friedman's avatar

That is the argument that I started with, the reason one would expect the opposite of the result observed. The puzzle is why it doesn't happen.

I believe the probability of getting divorced in the next year increases over time, which is the burnout effect.

Age of Infovores's avatar

I think I initially talked past you. My attrition/filtering point does basically boil down to a standard trial-period argument you already started with as cohabitation creates additional pre-marriage time at risk during which the most mismatch-prone couples can break up and therefore never enter the married sample at all. I was thinking of it as distinct from couples simply having “more information” in the narrow sense, but I see now how it fits.

Nadav Zohar's avatar

Interesting discussion and possibilities. I write only to add the idea that “marriage” has different meaning and significance to different people, even within just the population who consider cohabitation no big deal.

Also, I wonder how many couples in the evidence sample were married for the n+1th time? If I am a 40 year-old divorcee I am more likely to already have my own place that I find suitable, and won’t necessarily be in a rush to move in with my next serious girlfriend the way I was when I was 20 and eager to leave my ratty apartment and immature roommates.

David Friedman's avatar

Whether previously married was one of the variables they looked at but not how many times.

John Ketchum's avatar

There are many possible explanations for the data.

Isha Yiras Hashem's avatar

Is there any data whether seeial cohabitation with different people is different than monogamous habitation in terms of marriage outcomes?

David Friedman's avatar

I don't know. I would expect a selection effect making the serial cohabitors less likely to have stable marriages but there could be a "try before you buy" effect going the other way.

James Hudson's avatar

I have some reservations about the notion that, “since we evolved in a risky environment; giving up benefits today in order to get larger benefits ten years from now is probably a mistake if you are likely to starve to death in a famine or get eaten by a predator before the benefits arrive.” Evolution has constructed us to act as if we cared about our descendants, assumed to persist indefinitely into the future. We are not designed to act like “rational” (egoistic) calculators.

David Friedman's avatar

That only works for benefits that will be passed on to your descendants after you die.

James Hudson's avatar

Yes, though not just lineal descendants matter: in just the next generation one should act as if he valued benefits to nieces and nephews, first cousins once removed, second cousins once removed, etc. Since we are all (mostly very distant) cousins, that includes everybody.

David Friedman's avatar

But most of them with very little weight.

James Hudson's avatar

Perhaps making that up with volume.

Doctor Hammer's avatar

It is notable that people do not seem to care much about their descendants until they have produced at least one. I have also seen little evidence that people care much about their descendants beyond those they personally know, e.g. great grandkids your grandkids haven't had yet.

Empirically, it seems that our tendency to care about our descendants is pretty limited. Possibly as limited as our tendency to care about our future selves compared to our present self.

Stephen Riddell's avatar

Interesting stuff, David. Personally, my experience in this matter is very limited, so it was nice to read some speculation and evidence for various mating strategy theories.

Michael Beverly's avatar

I read this with a bias I caught at the end.

I think it's common to come to the table with the idea that divorce is bad/wrong/evil and/or and undesirable outcome.

But what if we turn that on it's head?

What if divorce is good?

What if most people who divorce should divorce?

Toxic relationships, abusers, or different paths, etc., could alter the formula here about whether the statistics mean what they seem to mean at first glance (I started reading guilty of this myself).

If we assume that people who live together are less religious (less worried about what others think, etc., and are therefore more freethinking, and maybe even smarter on average) then it doesn't necessarily hold that more divorce in that group is a bad thing --- why should anyone stay married if they don't want to stay married?

It's interesting if we flip this script here.

As a sidebar, I was cohabitating with my girlfriend for about 4 years when she got pregnant, so we got married. Legally, it makes sense to be married in some cases, but I wouldn't use that for a reason to stay married like a religious person would.

David Friedman's avatar

I think the implicit assumption is that marrying someone you will end up divorcing is worse than marrying someone you will never divorce, worse for both you and your kids. But the post and the article I cited are about explaining the observed pattern not evaluating it.

Chartertopia's avatar

I've known three married couples who hated each other with a passion, but dreaded living alone more than living with someone they hated. I'd never thought about the search hassle, but I'm sure that was part of it — they probably figured they would be so desperate that they'd end up worse off, which is probably how they ended up with such miserable spouses in the first place.

Mr. Doolittle's avatar

Whether kids are involved makes a huge difference here. For a kid, parental divorce counts as an ACE (https://www.cdc.gov/aces/about/index.html) because it's correlated with terrible outcomes for kids, similar to child abuse, family member suicide, and a group of terrible childhood experiences. Divorce stands out on the list because it's one of a few we haven't criminalized, yet causes similar levels of trauma.

Even for adults, breaking up with someone causes a pretty high level of pain for one or both people. It's not exactly a great thing. The closer the relationship, the worse the pain. Presumably married people are much closer than lesser forms of bonding, though that's clearly not always the case. That divorce is neutral or good would be very hard to demonstrate, and may only be possible in relation to other kinds of breakups or staying in a relationship that's more damaging. It's not a good within itself, and can't be.

Michael Beverly's avatar

My daughter is married to a violent drug addict.

She's not speaking to me because I objected when her husband physically assaulted me and I called the police.

So, I'm familiar with ACES as it's proof she should divorce. I.e. living with a violent drug addict father who is physically and mentally abusive is worse than divorcing him.

So, yeah, having kids makes a difference, but it doesn't mean divorce isn't often the right choice (a lessor of evils if you'd like).

I've divorced twice from women I've had children with. Both times were better for the kids, ultimately, but yeah, in both cases it was traumatic and hard and difficult and emotionally damaging for all parties.

So, yes, divorce is correlated with terrible outcomes, but so are many of the reasons someone would seek a divorce, so on it's own, it doesn't help us figure out the best solutions (i.e. not many people who have wonderful relationships without abuse and bad conditions and live in a peaceful, wonderful relationship, etc., seek divorces, do they?).

Mr. Doolittle's avatar

Are between 40-50% of marriages involving abuse or similar levels of negative environments? Are couples that cohabitate almost twice as likely to be abusive?

I can agree that there are some marriages that are very bad, and staying in them would be worse than leaving. But that's not most cases of divorce.

Michael Beverly's avatar

I suppose it depends on how you define abuse and negative environments, doesn't it?

Also would matter how those levels of abuse/negative environments damage the kids.

I'm sixty years old, my observation, granted it's anecdotal, is that most marriages after the first few years (if that) of "honeymoon" period aren't happy ones. They're negotiated peace treaties at best and long-term subversive wars at worst.

Maybe the circles (friends/family) I've traveled in are just terrible people...I don't know, I mean, I'm not saying I don't know people with seemingly stable good marriages, but even those are suspect because a lot of people are good at hiding what goes on behind closed doors.

I do think all things being equal, it's way better for a couple to stay together for the children, at least until they're not babies (the worst time for drama, abuse, trauma, and divorce, etc., I think the ACES studies concentrate or expose how deeply damaging these things are at birth to like 5 years, if I recall correctly --- for sure it's true that a 2 year old subject to abuse has a harder time in life than a 12 year old or a 16 year old, it's still bad, but it's more long term damaging when abuse or witnessing it happens in early brain development).

So, yeah, you're asking a good questions but without more definitions and studies and understanding, how can we answer the question of when the tipping point is hit that divorce is better than staying? I don't know...I guess everyone has to decide for themselves based on their best understanding.

Which leads me to my feeling here....most people error on staying together too long, imho.

Especially when the abuse is coming from a domineering male over a submissive female. I watched this in my mom's 2nd marriage (I lived in this abusive house and it really hurt my chances at a good life) and I'm currently watching in my daughter's marriage where for sure she'd be better off without staying married, heck, if I could get away with offing the abuser without making it worse, I would in a heartbeat. He's a monster. She's choosing to stay with him for reasons that will ACES the crap out of my grandson, who is facing worse chances at everything in life due to the ACES check list, i.e drug abuser father, who is violent, I mean, he physically assaulted me in front of his son, that was the day of last contact with my grandson, who is now isolated from extended family. His mom, my daughter is obviously mentally ill as she defended her husband beating me up as, "oh, he's from a different culture and it was just two grown men," and "oh, he's not really a drug addict," and "I know how to handle him to keep him calm," and all the stuff abused women say when they're co-dependent and don't want to leave.

For sure the kid is WAY better off if his dad got hit by a bus tomorrow.

So a divorce would the next best thing.

How many women stay in this kind of relationship? And men? FAR too many.

And these kinds of stats are hidden. My son-in-law wasn't arrested when I called the police, in fact, my daughter lied to the cops to protect him. He's never been arrested for drugs. He comes across as a high functioning businessman who is a "nice" guy.

My step-father was the same. The guy was a drug dealer, actually, but nobody knew it, and of course, we in the family knew to keep our mouths shut. In public he was a super nice man. Kind, generous, fun, funny, well liked. Which is why it took decades of abuse for my mom to finally leave him.

Anyway, I know this isn't a therapy session, but thank you. Send me an invoice.

Doctor Hammer's avatar

I think the trouble with divorce is what it says about the state of the marriage that ends that way. For every divorce there are going to be other couples trapped in a miserable marriage, etc. As a result, trying to minimize the number of divorces itself is not desirable per se, but if divorce is a sign of a bad marriage then things that lead to more divorces must be leading to worse marriages.

An analogy would be "When repairing a small engine, taking photos of where every part came from before you remove it leads to fewer cases where the motor has to be taken apart again afterwards." Of course if you put the thing back together wrong you definitely want to pull the motor back apart and do it right, it is the right thing to do, but if taking careful photos leads to fewer cases where that rebuild is required it is better to do so.

A divorce is evidence that a mistake was made initially in the marriage (possibly through no fault of one's own!) and that starting over was better. Avoiding early mistakes to avoid later divorce is good even if divorce is good as in your example.

David Friedman's avatar

"then things that lead to more divorces must be leading to worse marriages."

I didn't discuss it but the article had evidence of correlation with features of marriage quality other than divorce.

"Among married individuals, premarital cohabitation is related to lower marital satisfaction, less time spent together in shared activities, higher levels of marital disagreement, less supportive behavior, less positive problem solving, more reports of marital problems, and a greater perceived likelihood of marital dissolution."

Doctor Hammer's avatar

I'd believe that. I suspect there is a lot to the notion of relationship inertia, where after a while it seems harder to dissolve the relationship than just keep going, and so cohabiting becomes a bit of a trap where you are "trying out" living with someone but your lives become enmeshed enough that even if you find out you don't really like the person it is easier to just stay together and hope things get better.

I had a girlfriend like that in undergrad, where we started dating right at the start of our freshman year and did everything together. We realized we didn't like each other much after a while, but every time we broke up, after a day or two we would call the other up to hang out because we didn't really know what else to do with ourselves and all of our shared friends just expected us to be together. The relationship drug on and off for a few years like that until we had pretty much excised the shared friends and found new, non-overlapping groups. If we were living together and had to deal with finding a new apartment etc. we probably would have gotten married during one of the relationship upswings eventually. To no good end.

David Friedman's avatar

I have been long and long away from sorrow,

Close havened as it seemed against the storm,

But now it seems that on some soon tomorrow

My walls are mist and I must cloak me warm.

(The first verse of a poem I wrote at the end of my first marriage.)

Philippe DARREAU's avatar

Brinig's paradox says that marriages preceded by cohabitation have a higher divorce rate, whereas one would expect that cohabitation would allow a more thoughtful and better choice that would lead to a decrease in the divorce rate. Both of your explanations are that cohabitation prevents the search for the best partner. Because we are subject to emotion or sex and not to rationality. Your explanation assumes that the alternative, marriage shortly after the meeting, was preceded by a period of rational and effective research to make a wise choice. But this period of research can also exist before the cohabitation.

What can be the basis for the wise choice? Why would a quick marriage strengthen it?

David Friedman's avatar

My alternative isn't marriage shortly after meeting. It was two or three years after I met Betty that I proposed to her for the first time, two or three more before she accepted.

That was more than forty years ago.

The alternative is a substantial courtship without cohabitation. Depending on circumstances, it might be a substantial non-romantic relationship followed by a brief courtship.

Mr. Doolittle's avatar

Cohabitation is a lesser commitment. That means it's easier to get into, and there are fewer reasons for people to stop the relationship from progressing. Someone only interested in a long term marriage will refuse to join an intermediate relationship that doesn't look like a good long term fit. Someone open to cohabitation is signaling an acceptance of lesser criteria and therefore a worse long term matchup. David's theory of relational inertia would then kick in.

Philippe DARREAU's avatar

I believe that discovering a person takes time and that an extended trial period reinforces a wise choice. So cohabitation implies a wiser choice and a reduction in the divorce rate. But this does not imply that divorces can be explained by the absence of cohabitation. Divorces have multiple explanatory variables.

Mr. Doolittle's avatar

Please reread David's post. He thought the same thing, but the actual evidence is that people who cohabitate divorce at almost twice the rate.

Dr. Nicole Mirkin's avatar

I wonder whether what’s labeled an “experience effect” is less about cohabitation per se and more about decision-making under sunk costs and diminished search incentives.

Zoe Baum's avatar

Cohabited before marriage, going on 38 years of marriage. We married for insurance and tax reasons, not for ‘commitment’ reasons. Do the studies look at those elements of a relationship? Does my anecdotal experience cancel out yours😁?

David Friedman's avatar

I only read one study and it doesn't.

When you started living together had you already decided it was going to be permanent?

Laura Creighton's avatar

I know several cohabiting people who took one look at the tax consequences of being married and headed to the courthouse. They didn't see their marriage as particularly much of a commitment.

Mikhail's avatar

I think the reason is that moral standards are less strict, which is why people break up more often among those who lived together before marriage.

David Friedman's avatar

You are saying that the moral standards of those who cohabited are less strict than of those who did not, probably a selection effect. Are you using "moral" to refer specifically to sexual morality? Traditional sexual morality? When I cohabited with the woman who became my first wife I didn't think of it as immoral and doubt she did.

Perhaps "have different moral standards" would be more accurate, standards that put less negative weight on divorce, perhaps also on infidelity.

Mikhail's avatar

I don't put any negative weight to divorce or infidelity or anything else here, nothing about that at all. I am about: if a person sticks to traditional morals or rather more liberal. The more a person liberal, the higher is probability of divorce, I think so. Equally, the more a person liberal, the higher is probability of sexual relations before marriage. Wrong guess?

Governology's avatar

An explanation I don't see here is that the type of marriages that aren't preceded by cohabitation seem likely to be correlated with more religious people, or people who have stronger adherence to traditional norms. Premarital cohabitation and divorce are both much more strongly taboo in those kinds of cultures. If this is the case, one would expect couples who don't have strong cultural beliefs about such things to have little difference between cohabiting before marriage or not.

Wasserschweinchen's avatar

That's what I read into the what the post calls a selection effect.

David Friedman's avatar

Except that I don't think they had the data to test it, except by testing things that might correlate with it.

Gian's avatar

Cohabitation is largely a phenomenon of modern West. So I wonder if any general conclusions regarding human biology or nature could be derived from so peculiar a (non)-institution.

Vladimir Vilimaitis's avatar

It might be that people who don't cohabitate before marriage tend to be more traditional and therefore take divorce much more seriously.

David Friedman's avatar

That is a plausible version of the selection effect. Presumably the authors did not have data on church attendance and the like and so couldn't control for it, which makes their failure to eliminate the effect by controlling for confounders less convincing as evidence for an experience effect.