Statisticians Uncover What Makes For a Stable Marriage

[NobleSavage]

Randy Olson, a Computer Science grad student who works with data visualizations, writes about seven of the biggest factors that predict what makes for a long term stable marriage in America. Olson took the results of a study that polled thousands of recently married and divorced Americans and and asked them dozens of questions about their marriage (PDF): How long they were dating, how long they were engaged, etc. After running this data through a multivariate model, the authors were able to calculate the factors that best predicted whether a marriage would end in divorce. "What struck me about this study is that it basically laid out what makes for a stable marriage in the US," writes Olson


http://www.randalolson.com/2014/10/10/what-makes-for-a-stable-marriage/

 

[NobleSavage]

The most interesting thing I to me was:

Crazy enough, your wedding ceremony has a huge impact on the long-term stability of your marriage. Perhaps the biggest factor is how many people attend your wedding: Couples who elope are 12.5x more likely to end up divorced than couples who get married at a wedding with 200+ people. Clearly, this shows us that having a large group of family and friends who support the marriage is critically important to long-term marital stability.

 

[Loren Pechtel]

A lot of these seem to me to be measures of the casualness/consideration of the marriage. Nobody has a big wedding on the spur of the moment.

 

[bilby]

Interesting to see that casual church attendance appears to be more likely to lead to divorce than either complete non-attendance or regular attendance.

 

[Bronzeage]

No great surprises there, except the one about income, but apparently money can buy happiness. I've been married 3 times, the current relationship is going on 24 years, with 22 years of marriage. For myself, I don't see any correlation with the statistics.

 

[EricK]

None of these really make for good advice. eg a couple who marry after 1 year and divorce 3 years later would very likely have divorced 1 year after getting married if they had waiting another couple of years before getting married. Or inviting extra people to your wedding who you don't particularly want there isn't likely to make your relationship last any longer.

 

[rjh01]

It says nothing about how you should treat your partner. For example how often should you praise or criticize them? Is this sort of thing irrelevant?

 

[exodiasrightleg]

No great surprises there, except the one about income, but apparently money can buy happiness.

Presumably, celebrity marriages weren't counted, or they would seriously skew the results over.

 

[hurtinbuckaroo]

None of these really make for good advice. eg a couple who marry after 1 year and divorce 3 years later would very likely have divorced 1 year after getting married if they had waiting another couple of years before getting married. Or inviting extra people to your wedding who you don't particularly want there isn't likely to make your relationship last any longer.

Actually, it's probably more likely that if they had waited a couple of years, they wouldn't have gotten married in the first place because the issues that would eventually have caused the divorce would have caused a breakup before the wedding.

And you can't create an artificial "support system" just by inviting more people. That's missing the point entirely.

 

[Keith&Co.]

It's probably important to remember this is from a statistician, not an advice columnist or a wedding planner. It's something that's been observed, not necessarily something you can control.
The important thing about number of wedding guests is that you've got a support system, not bums on seats. I suspect the important thing about church attendance is not the church itself, but that both people are either committed to a religion or both are not. If the couple goes only as often as one partner can talk the other into going, that burden's going to get heavy pretty quickly.

 

[Malintent]

They say that the likelihood of divorce increases with cost of wedding ceremony... and then they say that likelihood of divorce decreases with size of wedding... in reality, the larger the party the higher the expense. So, obviously these two 'conclusions' conflict.

I stopped reading the article at that point... I am assuming the overall conslusion will be that we need more pirates.

 

[skepticalbip]

The article seems to suffer from assuming that correlation implies causation or maybe it was just written to have a place to put a lot of unrelated statistics that had been gathered. Apparently the author didn't have any statistics on the astrological signs of the parties or the differences of their hair color or he would have likely included those too.

My unprofessional opinion is that a “good marriage” is one where the two are both well adjusted adults who are compatible.

 

[rousseau]

A lot of these seem to me to be measures of the casualness/consideration of the marriage. Nobody has a big wedding on the spur of the moment.

Not to mention wealth correlates with wedding size, and wealth also has a strong correlation with divorce rates.

 

[rousseau]

The article seems to suffer from assuming that correlation implies causation or maybe it was just written to have a place to put a lot of unrelated statistics that had been gathered. Apparently the author didn't have any statistics on the astrological signs of the parties or the differences of their hair color or he would have likely included those too.

My unprofessional opinion is that a “good marriage” is one where the two are both well adjusted adults who are compatible.

Yea a lot of correlation, most implying the cause.

 

[doubtingt]

None of these really make for good advice. eg a couple who marry after 1 year and divorce 3 years later would very likely have divorced 1 year after getting married if they had waiting another couple of years before getting married. Or inviting extra people to your wedding who you don't particularly want there isn't likely to make your relationship last any longer.

Actually, it's probably more likely that if they had waited a couple of years, they wouldn't have gotten married in the first place because the issues that would eventually have caused the divorce would have caused a breakup before the wedding.

Agreed.

And you can't create an artificial "support system" just by inviting more people. That's missing the point entirely.

Yes, but the article itself is what misses the point entirely when it claims .....

Crazy enough, your wedding ceremony has a huge impact on the long-term stability of your marriage. Perhaps the biggest factor is how many people attend your wedding

The article attributes causal impact of the wedding ceremony itself. It is pretty obvious that there is no direct relationship and it is a correlation produced by third variable factors.

A large "support system" can play a role, but just as likely is the impact of being a conformist to tradition and afraid to stand up for yourself and your own happiness. Most large weddings are due to couples caving blindly to conformity and pressures of society and parental expectations that go against what is reasonable and good for themselves. Every massive wedding I have been to have been ones where one or both of the mothers took control, while all the small ones of I have been to have been where the couple themselves did it with little input from the parents (these have also been the funnest weddings with the best food, drink, and music, and least amount of religion). It is often the wife caving to parents and social pressures and the husband caving to the wife.
Now, you might think this would mean more divorce, but that is simplistic. What is means is lifelong unhappy marriages between people without the guts to get divorced. IOW, they don't get divorced for all the same unhealthy reasons that they had a big, debt creating wedding in the first place.

Also, the "wedding expenses" results are very misleading. The linked article shows the results only after controlling for all other variables, which includes # of guests, which is one of the biggest determinants of wedding cost. So it is a bit weird to look at wedding expense controlling for number of guests. It represents the amount you spent on each guest rather than total amount spent. When you look at the original research article, it shows that when you just look at the simple relationship between wedding expense and divorce, the relationship is the exact opposite, with people who spend more being less likely to get divorced. This matches the results for number of guest, meaning that if you have an expensive wedding with lots of guests, you'll be less likely to get divorced, but again I would argue that this is due to "sticking it out" in bad marriages for the sake of appearances and social fear.

IT is a very similar reason why "religious" people are less likely to get divorced.

 

[Bronzeage]

The article seems to suffer from assuming that correlation implies causation or maybe it was just written to have a place to put a lot of unrelated statistics that had been gathered. Apparently the author didn't have any statistics on the astrological signs of the parties or the differences of their hair color or he would have likely included those too.

My unprofessional opinion is that a “good marriage” is one where the two are both well adjusted adults who are compatible.

A good marriage is when you find someone who can put up with your shit.

 

[doubtingt]

They say that the likelihood of divorce increases with cost of wedding ceremony... and then they say that likelihood of divorce decreases with size of wedding... in reality, the larger the party the higher the expense. So, obviously these two 'conclusions' conflict.

I stopped reading the article at that point... I am assuming the overall conclusion will be that we need more pirates.

See my post above about this. The weird result you speak of is the product of looking that the effects of wedding cost after controlling for number of guests.
The OP article is a puff piece that didn't give much thought to the results or to which of the different analyses it chose to pull from the original research article. Its still a whole lot of correlations in which the real cause is almost certainly a third variable factor, but the silly opposite effects are of cost and number of guests are a statistical artifact.


This is a good lesson is why multiple regression analyses and entering "control" variables needs to be theoretically driven and care must be taken to realize that you can change what a predictor variable is actually measuring when you enter other control variables, such as this case where "total cost of wedding" changes statistically into "cost per guest" once you control for number of guests. A closely related lesson is that economists are full of shit and almost never do real science, and their "models" are usually theoretically void (and thus meaningless) piles of uninterpretable covariances that they are yet still happy to draw wrong inferences from.

 

[Bronzeage]

They say that the likelihood of divorce increases with cost of wedding ceremony... and then they say that likelihood of divorce decreases with size of wedding... in reality, the larger the party the higher the expense. So, obviously these two 'conclusions' conflict.

I stopped reading the article at that point... I am assuming the overall conclusion will be that we need more pirates.

See my post above about this. The weird result you speak of is the product of looking that the effects of wedding cost after controlling for number of guests.
The OP article is a puff piece that didn't give much thought to the results or to which of the different analyses it chose to pull from the original research article. Its still a whole lot of correlations in which the real cause is almost certainly a third variable factor, but the silly opposite effects are of cost and number of guests are a statistical artifact.


This is a good lesson is why multiple regression analyses and entering "control" variables needs to be theoretically driven and care must be taken to realize that you can change what a predictor variable is actually measuring when you enter other control variables, such as this case where "total cost of wedding" changes statistically into "cost per guest" once you control for number of guests. A closely related lesson is that economists are full of shit and almost never do real science, and their "models" are usually theoretically void (and thus meaningless) piles of uninterpretable covariances that they are yet still happy to draw wrong inferences from.

My oldest daughter has a half dozen cousins on her mother's side and she was in the wedding party for all of them. The weddings were all staged and managed by the committee of the mother and aunts of the bride. She swore she would not suffer the same fate. She purchased a wedding package from the Mandarin Orient, a nice hotel, halfway between Coco Head and Diamond Head, Oahu, Hawaii. The package included the minister, musicians(harp and keyboard), photography, video, wedding cake, and a night in the honeymoon suite. All told, it would be a modest package, if every guest had not spent about $1200 on the plane ticket just to get there. For what it's worth, they are still married.

 

[fromderinside]

Daughter in law's family sprang for a large room for all three of their daughters' weddings in west Los Angeles. Bands, full meals, appetizers, other entertainment, many changes, all are still married. Daughter looked in parents eyes and opted for a cozy wedding at Gaetano in  Calabasas, Ca near Woodland hills and Beverly Hills then marriage ceremony at our home conducted by one of son-in law's aunts. All are still married, all are happy, all have exactly one child three of whom entered college this fall.

For what it's worth wife and I got married on less than $700. Had a high mass, three hundred guests, buffet, flowers in church and at apartment for buffet and we paid the priests $20 each. Uh, we're still married. Don't see the connection between money, social status or anything else.

Statistics. Asians: 4, American Catholic: 2, American Jewish and Catholic: 1, Unitarian, Methodist, Baptist, Lutheran, latter Day Saint, Atheist: 1. Incomes top 2 percent: all, Career: all, Within 25 miles of US home: 6, More than 500 miles from US home 1, more than 800 miles from US home 3. Married outside ethnic group or race: all, Born more than 2500 miles from current residence 5.

 

[Emily Lake]

this is due to "sticking it out" in bad marriages for the sake of appearances and social fear.

This same thought struck me about halfway through: People who regularly go to church, had a lot of people at their weddings, were together for a long time before getting engaged, etc. While it's possible that there's some causal relationship there... the actual reason hasn't been addressed. And the assumption of these being happy or successful marriages is just that: assumption. A long marriage isn't guaranteed to be a good marriage. It could very well be a marriage wherein one or both parties feel socially pressured to stay together due to having children, family pressure, or religious expectations.

On the other hand, I dated my spouse for a year before we eloped at age 21 and spent all of $150 on our marriage license and the Justice of the Peace. We had no honeymoon at all, and we've been happily married for twenty years now.