Is It or Isn't It?

[we are republishing an email from Scott Stanley to the SmartMarriages community]

Is premarital cohabitation associated with more risk for divorce (and lower marital quality) or isn't it? Check out these stories in the news and see if you can figure out the answer to that question.

NY Times: Living Together First Doesn¹t Make Marriage Last, Study Finds
http://www.nytimes.com/2010/03/03/us/03marry.html

USA Today: Report: Cohabiting has little effect on couples' success in marriage
http://www.usatoday.com/news/health/2010-03-02-cohabiting02_N.htm

MSNBC: Living together does lead to marriage 2 in 3 who cohabit tie the knot within 5 years, CDC says
http://www.msnbc.msn.com/id/35663243/ns/health-behavior/from/ET

Reuters: Marriages last longer than living together?
http://www.reuters.com/article/idUSTRE62133E20100302

ABC NEWS: Moving In Together Before Wedding Can Decrease Chances of Having a Lasting Marriage, Study Says CDC Survey Says Cohabitation Isn't Always a Good Thing, Couples Say It's All About Love
http://abcnews.go.com/WN/secrets-marriage-cdc-survey-marriage-cohabitation/story?id=9989134

Are you confused yet? Most of these news reports read like a lot of new evidence just came out that changes everything you might have thought you or social scientists knew about cohabiting prior to marriage. It is instructive to read these articles. In fact, it's fascinating to read the headlines and stories and try, from there, to figure out what the new findings really are in the report that came out from the CDC¹s National Center for Health Statistics, on Tuesday (3/2/10). You would do a lot better to read the actual report, and if you want to do that, the link is further on in this piece.

One news article sounds like the report merely confirms the finding that has been well understood for years‹that cohabiting prior to marriage is associated with greater risks for divorce. Another artilce sounds like the risk is no longer there. Gone. Poof. We've grown out of it as a society.

Another is pretty nuanced, suggesting both the "it's still here" and "it's gone for good" messages are true.

My colleagues, Galena Rhoades, Howard Markman, and I found this all pretty interesting. It¹s especially fascinating to compare the NYT piece with the MSNBC piece, if you want to see how far apart the (mis)understandings. Were they reading the same report? Since it mentions our work, we¹re partial to the USA Today piece. That piece also, at least to me, comes closest to telling the more complex story. I believe that USA Today will have an even more detailed piece out on this CDC report, today.
(http://www.usatoday.com/news/health/2010-03-02-cohabiting02_N.htm

So what do we really know, as social scientists? I mean "we" as the field. I (Scott) would sum it up this way. There is a lot of evidence over many decades by many researchers that living together prior to marriage is associated with greater risks for divorce or lower quality marriages. One of the really interesting debates in the field is if this association with risk
will go away as the majority of people live together prior to marriage. Since some scholars have long felt that the stigma of cohabitation contributed in a large way to the risk, they felt that the risk would diminish as the stigma diminished. That's a reasonable theory and there is some scattered evidence to support it. The stigma has certainly diminished. This new CDC report could be taken to suggest that the risk is not what it used to be, and so it fits this way of thinking. It's probably safe to say that the risk associated with cohabiting prior to marriage is not as clear as it used to be. But has it gone away, completely? Well, not so fast. Just like some people¹s relationship status on Facebook, it¹s complicated.

As many of you know, we (our lab) have been doing research in this area for some time, and are presently conducting a rather large study on the question (as well as other important questions). One of the major theories we are testing is what we call the inertia perspective. Here's a quick recap of the idea. Cohabitation may be risky for some people because it makes it harder to break up with someone that they otherwise would have broken up with and never married in the first place. (Or, they never would have stayed together long enough to have a child; a variable that fits all the same logic.) Yes, cohabiters break up all the time. In fact, they break up much more often than people who are married. That is not exactly a news flash. But cohabiters break up much less often than people who are dating. A lot of the time, that's just because cohabiters are more likely to be more deeply involved (entangled). The idea of inertia is that cohabitation makes it harder to break up in comparison to dating without cohabiting. Inertia suggests that some people actually marry someone they would not have married had they not been cohabiting.

This idea of inertia also suggests that, among those who eventually marry, those who were already engaged or mutually planning marriage, when they began to cohabit, will be at lower risk for marital difficulties than those who cohabited before nailing down that big question. It's the latter group that would contain the people adversely affected by inertia. If two people already mutually agree on marriage before cohabiting, it¹s not very likely cohabitation will be a factor in them getting married. Inertia is not their issue. There may be other factors that matter for such a couple, like what their beliefs and values are about cohabitation, but inertia is not part of equation for them. 

Back to the CDC report. What got a lot less attention on Tuesday is the fact that the researchers also found exactly what we have been predicting, testing, and publishing for some years now based on the concept of inertia. They found that those who cohabited prior to engagement were at greater risk for divorce but those cohabiting after engagement were at no greater risk than those who didn¹t cohabit at all before marriage.

As Galena Rhoades and I presented in our July plenary at the Smart Marriages conference, the premarital cohabitation effect is really a pre-engagement (or pre-marriage-planning) cohabitation effect. As a side point, it's actually possible that the premarital cohabitation risk as always, mostly, been a pre-engagement cohabitation risk but we just didn¹t know it before now. There are only a handful of data sets in existence where the key questions about the timing of plans for marriage and cohabitation are asked - most of them are the ones we've been working with. The data set used for the CDC report is the only other one we know about that asks the critical question (The National Survey of Family Growth). And, by the way, the critical question is not if a couple has plans for marriage but if they had plans for marriage when they began to cohabit. Did the plans exist before it got harder to break up?

We have published numerous studies using different data sets showing the pre-engagement cohabitation effect, and there is another publication on the way later this year. In fact, in every data set we know of where you can examine how cohabitation prior to engagement relates to marital outcomes, we have found the pre-engagement cohabitation effect. (So I can quit saying it every time, please note that I¹m using the term "pre-engagement cohabitation effect" interchangeably with the more important idea of whether or not there are mutual plans to marry when couples start to cohabit.) We also knew this finding was in the data set used in the CDC because of a one line comment about it in a report some time ago, but no one had actually presented it clearly as a finding before Tuesday.

Here's what the CDC report says about it:
Previous cohabitation experience was significantly associated with marriage survival probabilities for men. In general, men who cohabited prior to their first marriage had lower probabilities of the marriage surviving to the 10th anniversary than those who did not cohabit prior to their first marriage. However, those men who were engaged at the time of cohabitation with their soon-to-be first wife had similar probabilities that their marriage would last 10 years as did those who did not cohabit prior to marriage: 71% for men who were engaged at cohabitation and 69% for men who had never cohabited before their first marriage.

If the couple were engaged when they began cohabiting, the probability that a woman's marriage would survive for 10 years was similar (65%) to the probability for couples who did not cohabit at all (66%). In contrast, the probability that the marriage would survive 10 years or more was lower if the couple were not engaged when they began cohabiting (55%). http://www.cdc.gov/nchs/data/series/sr_23/sr23_028.pdf 

Importantly, we have found that cohabiting prior to engagement is associated with lower quality marriages and divorce, even among those married relatively recently. For example, we found the difference in terms of marital quality in a national sample of those married 10 years or less as of 2007. That's recent. In fact, those marriages are more recent than the marriages studied in the CDC report. We also found the risk for lower marital quality and higher risk for divorce in two other samples of those married in the 1990's, which I believe is as recent as the data used in the CDC report. We have consistently found this result even while controlling for a large number of potential "selection" variables. Selection variables are pre-existing differences between people that explain a risk that appears to be related to one thing (like cohabitation) but it¹s really simply because of the selection variables. For example, those who are more traditionally religious are less likely to cohabit, especially prior to engagement, and they are also less likely to divorce. Those who have lower incomes or who are less educated are more likely to cohabit, and also more likely to cohabit prior to having plans for marriage, and they are also more likely to divorce. Selection can never be completely ruled out, but the more a finding continues to exist while controlling for loads of selection variables, the less reasonable it is to say selection explains it. Selection is at least part of the story of how cohabitation prior to marriage (or at least, engagement) is related to marital outcomes. Yet, we have been unable to explain away all of the risk of divorce and/or lower marital quality associated with pre-engagement cohabitation by controlling for a large number of selection variables. It's important to keep all of this in perspective. There are many risk factors for marital distress and divorce, and living together prior to engagement (or having mutual plans for marriage) is one among many. An even greater risk factor for divorce is having a lower level of education; not finishing high school is a whopping risk factor for divorce. As is poverty. As is marrying young. As is having divorced parents. As is not being able to handle conflict well. As is . . ., well you get the idea. But the most important thing about risk factors is that some are more changeable than others. If it turns out that cohabiting prior to having clear plans for marriage really adds to a person's risk for marriage turning out less well‹and in a way that appears truly causal‹it¹s not rocket science to think a person could learn not to do that, and thereby, lower their risks for marital unhappiness.

Why does this issue of cohabitation always get so much attention? There are at least three reasons.

First, it's something that is connected to values for people, so debates about cohabitation are not just about cohabitation, they are also about religious beliefs and other values. Those are important issues, but my observations here are focused only on the science.

Second, younger people tend to believe that cohabitation lowers their risks in
marriage, so reports that it's not associated with benefits, or even associated with risk, have gotten a lot of attention. People are interested in this.

Third, cohabitation is important societally because it's increasingly the context for childbearing and rearing, and cohabiting relationships are much more fragile than marriages. That makes it an important subject in our national discourse.

With funding from the National Institute of Child Health and Human Development (NICHD), we (our team at DU) are following a large sample of young adults in serious, romantic relationships in order to try to better understand who may be at greater and lesser risk, and why. We are measuring various pathways people follow before marriage and into marriage. The clearest picture of how all of these things relate to marital quality can only come from following people through the various possible transitions before marriage and then into marriage, while measuring a great many variables. That¹s what we continue to do, and we will have more to talk about in the future.

Scott Stanley
Research Professor
University of Denver

[see Scott's blog at http://slidingvsdeciding.blogspot.com/ ]