Statistics Canada - Government of Canada
Accessibility: General informationSkip all menus and go to content.Home - Statistics Canada logo Skip main menu and go to secondary menu. Français 1 of 5 Contact Us 2 of 5 Help 3 of 5 Search the website 4 of 5 Canada Site 5 of 5
Skip secondary menu and go to the module menu. The Daily 1 of 7
Census 2 of 7
Canadian Statistics 3 of 7 Community Profiles 4 of 7 Our Products and Services 5 of 7 Home 6 of 7
Other Links 7 of 7

Warning View the most recent version.

Archived Content

Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.

Skip module menu and go to content. Online catalogue record for Canadian Social Trends Canadian Social Trends main page Articles by subject Previous releases Coming articles Social indicators Attention educators! News Other research More information about Canadian Social Trends Order the print version

Till death do us part? The risk of first and second marriage dissolution

by Warren Clark and Susan Crompton

Marriage has been on just about everyone's mind for the last few years. While the discussion was sparked by the debate over same-sex marriage, many thoughtful Canadians were led to consider just what marriage means in today's society.

Marriage as we have understood it over the last 50 or 60 years seems to be losing its appeal. Marriage is being "de-institutionalized", in the words of American social researcher Andrew Cherlin, as old social norms crumble and couples must negotiate new, mutually acceptable standards of behaviour.1

Certainly, there is now less marriage, partly because young adults are delaying marriage and partly because common-law union is increasingly replacing marriage among Canadians of all ages.2 Also, there is more divorce; well over one-third of Canadian marriages will end in divorce before the couple celebrates their 30th anniversary.3 Finally, marriage is no longer a prerequisite to childbearing, as more and more children are being born to single mothers or unmarried couples.4

Nevertheless, the great majority of people do marry. This article uses the General Social Survey on family history to briefly examine the basic characteristics of Canadians who have legally married once, twice or more than twice. It then uses a proportional hazard model to identify some of the factors that are associated with ending a first and a second marriage by divorce or separation.

What you should know about this study
The first marriage
Age at marriage and living common-law are key factors in first marriage failure
Religion and mother tongue are linked with staying married
The second marriage
The first failure may help to set the stage for the next one
Being a member of a minority population is associated with subsequent marriage failure
The third marriage
Believing in Marriage produces a stronger marriage

Appendix table 1. Selected characteristics of ever-married Canadians aged 25 and over
Appendix table 2. Selected characteristics of twice-married Canadians aged 25 and over
Appendix table 3. Selected characteristics of serially-married Canadians aged 25 and over
Appendix table 4. Attitudes and religiosity of ever-married adults aged 25 and over, by number of times married

Chart 1. About nine in ten adult Canadians who have ever been legally married have tied the knot only once

The first marriage

According to the 2001 General Social Survey (GSS), just slightly more than 16.6 million Canadian adults — 80% of the population aged 25 and over — have married at least once.

On average, Canadian adults entered their first marriage when they were about 25 years old (for 89%, their first marriage is their current marriage). The grooms had been about two and a half years older than the brides, at 26.2 and 23.6 years old, respectively. (See Appendix table 1.)

Most people married another single person, but a few of them (6%) exchanged their first matrimonial vows with someone who had been married before. And although living common-law was not widely acceptable before 1980 (when most of them were courting), about 15% had lived with their spouse before the wedding.

About 9 in 10 ever-married Canadians (88%) have raised at least one child and at the time of the survey, 60% of them still had children living at home. Having children tends to bring people back into the places of worship they may have neglected in their youth,5 and indeed the majority (86%) of ever-marrieds reported that they belonged to a religious faith. Of these, 42% had attended religious services at least once a month in the year preceding the survey. (The corresponding rates for adults who have never married are 77% and 22%, respectively.)

At the time of the GSS, over two-thirds of ever-married people (69%) were still with their first spouse and they had been married for an average of 23.5 years. But for 23%, their first marriage had ended in dissolution following about 11 years of matrimony. (For the remaining 9%, their first marriage had ended in their spouse's death after 34 years together.)

Age at marriage and living common-law are key factors in first marriage failure

The success or failure of a marriage is ultimately decided by the deeply personal dynamics of the couple and their unique situation. However, a hazard model can be used to calculate the relative likelihood that a person's marriage will end in separation or divorce, given that the individual has certain socio­demographic characteristics. (See "What you should know about this study.")

One of the key factors associated with a first marriage breaking down is a newlywed's age. Someone marrying in their teens faces a risk of marriage dissolution almost two times higher than a person who marries between the ages of 25 and 29. In contrast, people who wait until their mid­30s or later to marry run a risk 43% lower. (The hazard ratio – or risk – is estimated for each variable when all other factors in the model are controlled for. See "What you should know about this study" for the list of variables included.) Age difference between spouses is not a significant risk factor if the husband is more than 5 years older than his wife, but it is 29% higher if he is more than 5 years younger.

People with less than high school education at the time of their first marriage face a 38% greater risk of marital dissolution than those with secondary completion; those with a university degree are at 16% less risk, when all other factors in the model are controlled for. This finding may seem contradictory – presumably people with lower socioeconomic status are least able to afford to leave their marriage – but it supports evidence which suggests that people with higher social status (especially women) are happier and less likely to divorce.6

Living common-law is also strongly associated with a first marital breakdown. In fact, the risk is 50% higher among people who lived with their partner before the wedding than among those who did not. This finding is supported by recent Canadian research which clearly shows that marriages preceded by a common-law union are distinctly less stable than those that began at the altar,7 possibly because the tradition of marriage is less important to people who have participated in non-traditional conjugal relationships.8

Table 1. Different factors are associated with risk of marital dissolution in first and subsequent marriages, but being young and more recently married is common to both. A new window will open.
Table 1. Different factors are associated with risk of marital dissolution in first and subsequent marriages, but being young and more recently married is common to both

The longer a couple has been married, the greater their chances of staying together. For example, someone who married in the 1960s is at 13% lower predicted risk of first marriage dissolution than someone married in the 1970s; however, the risk is a notable 67% higher for someone married in the 1990s, even when all other factors are accounted for. This difference across the decades probably reflects people's changing expectations of marriage, particularly the shift in emphasis from family-oriented child-rearing to individually-based personal fulfillment.

Having children significantly reduces the predicted risk of first marriage failure: it is 73% lower than that for married partners without children, after controlling for all other variables in the model. This finding bolsters the fact that, although children can put a strain on the adult relationship, marriage dissolution is actually less likely to occur among couples with than without children, an observation which is true across most societies and cultures.9

Religion and mother tongue are linked with staying married

Religious belief can also have a protective effect on first marriage. Although religious affiliation does not seem significant, religious observance is associated with marital durability. People who attend religious services during the year, even if only several times, have between a 10% and 31% lower predicted risk of marital dissolution than those who do not attend at all. (This excludes attending services on special occasions like weddings, christenings and funerals.)

The GSS does not provide information about respondents' cultural heritage. Nevertheless, given that language is a key transmitter of values and norms within a social group, mother tongue can be used as an indirect indicator of the attitudes to which a person was exposed while growing up.

People living outside Quebec, and whose mother tongue is neither English nor French, have a significantly lower risk of first marriage dissolution than the reference group (Anglophones outside Quebec), at almost 26% lower. The large majority of these allophones report that at least one of their parents was born in Asia or Europe, cultures which tend to have traditions that place strong emphasis on the importance of marriage and family.

On the other hand, Francophones in Quebec have even less risk of first marital failure, at 29% lower than Anglophones outside the province. This result is quite puzzling since Quebec posts a divorce rate higher than elsewhere,10 common-law unions are much more acceptable, and Quebec generally has a more socially liberal attitude than the rest of the country.11 In fact, being a francophone Quebecer is no longer a significant factor in lowering the risk of first marital dissolution if the attitudinal variables are removed from the hazard model (that is, importance of being in a couple, being married, and having children. Results of model not shown.)

The second marriage

The great 18th century English lexicographer Dr. Samuel Johnson famously remarked that remarrying was "the triumph of hope over experience."12 But about 43% of Canadian adults whose first marriage had ended in divorce had married again by the time of the GSS,13 as had about 16% of those whose first spouse had died.

Canadians who married a second time averaged about 39 years old at the time of the wedding. Over half (55%) exchanged vows with someone who had also been married before, and more than one-third (37%) had already lived common-law with their new spouse.

At the time of the GSS, about 1.3 million of them (71%) were still married to their second spouse of almost 13 years. There are good reasons for believing that these marriages will continue to be successful. American research suggests that remarriages made after age 40 are more stable than first marriages.14 And the hazard model predicts that, all other factors being controlled for, Canadians who were in their 40s when they remarried face only half as great a risk of marital dissolution as those who were under 30. Even those who remarried in their 30s have a 27% lower risk of breaking up.

The reason dissolution risk falls as age at remarriage rises may be partly due to the partners' increased maturity. An American study reported that the quality of the relationship between the couple is better when both spouses are remarried; they scored higher on measures of intimacy-based reasons for marriage than other types of couples and lower on external reasons.15 As for the "psychological baggage" they may bring to their new marriage, evidence suggests that the effect of divorce on general happiness, depression and general health is significant but weak, once the effects of demographic variables are removed.16

The first failure may help to set the stage for the next one

However, over one in five of Canadians who remarried had left their second spouse within an average of 7.6 years. Why someone's subsequent marriage should end in dissolution is perhaps more puzzling than why their first one did.

Some of the theories social research has presented to explain remarriage failure include: a personal psychology that makes someone more likely to end relationships; learned behaviour, that is, they solved the previous marital problem with divorce; lack of social support for remarriages; and a smaller pool of suitable candidates available for remarriage, which reduces the likelihood of finding a compatible partner.17

The first two hypotheses suggest that previous conjugal history may help to explain why the subsequent marriage failed. As shown earlier, both first and subsequent marriages contracted at a young age are less likely to succeed, probably because failure tends to repeat itself if a person has not corrected their "marital style". Adults who are twice-divorced were 3 years younger than their still-married counterparts, both the first time they tied the knot (22 versus 25) and the second (about 36 versus almost 40).

Interestingly, though, living common-law – which is much more common among twice- than once-married people and is strongly associated with a first marital breakdown – is not a significant factor in the dissolution of a subsequent marriage, once all other variables are controlled for.

The importance of social support to the success of remarriage has been acknowledged by a number of researchers. The support received from family and friends plays a significant role in the quality of the marital relationship, especially in couples where both partners are remarried.18 In contrast, low levels of social support contribute to the psychological distress reported by people who have divorced, especially those who have left a marriage more than once.19

Being a member of a minority population is associated with subsequent marriage failure

The choice of a second marriage partner has interested sociologists long enough for them to produce two competing theories. The "learning hypothesis" proposes that a person looks for someone similar to themselves after the failure of a marriage to someone dissimilar; in contrast, the "marriage market hypothesis" argues that people end up with a dissimilar partner because of the limited number of candidates available for remarriage.20 Neither hypothesis has trumped the other, and the results of the GSS hazard model are equally inconclusive.

Although higher education is a prime protective factor against first marriage dissolution, it is much less important to subsequent marriage dissolution. This seems to suggest that there may be more educational similarity between partners in second marriages. This interpretation is supported by a Dutch study of recently remarried adults that shows both sexes tend to choose a second partner who is better educated than their first; men especially are more likely to remarry a woman whose education more closely matches their own.21

On the other hand, the model's results also seem to speak to the difficulty of finding a compatible partner the second time around if a person belongs to a small population group. Two variables that played no role in first marriage dissolution are significantly associated with the breakdown of subsequent marriages. First, the risk for a francophone living outside Quebec is 83% higher than that for an Anglophone, when all other factors in the model are controlled for. Second, being a member of a religious faith other than the predominant Catholic or Protestant churches increases the risk by 135%, compared with someone who has no religious affiliation at all.

It has become a truism that step­children are a prime contributor to the collapse of second marriages. The appeal of this idea is obvious, and teenagers especially can put any marital bond to the test, but studies are inconclusive: some find that they are a prime factor in remarriage failure22 yet others determine that they contribute to the marital satisfac­tion of the adults.23 The GSS model predicts that, when all other variables are controlled for, the presence of children in the household at the time of a subsequent marriage is not associated with marital dissolution.

The hazard model also shows that some factors associated with marital success or failure are simply not within a person's power to control. For example, women have the same risk of subsequent marriage dissolution as men, which is somewhat surprising because they had a significantly lower risk for a first marriage break-up. The answer may lie in women's attitudes to marriage, since a new story appears when attitudinal variables are removed from the model. If the predicted risk is calculated using only socio-demographic variables, women and men in a first marriage have an equal risk of dissolution; but in a subsequent marriage, women face a 30% higher risk than men. (Results of model not shown)

The third marriage

In 2001, according to the GSS, almost 137,500 Canadian adults had been legally married more than twice. They represented less than 1% of the ever-married population aged 25 and over. Virtually all of them had tied the knot three times.

Apart from their marriage habit, nothing much sets these serially-married Canadians apart, socio-demographically, from other married Canadians. They had entered their third marriage at an average age of almost 46, generally to someone who had also been married before. Over one-third (38%E) had lived with their third spouse before the ceremony.

And although 71% had recently celebrated their 8th anniversary with their most recent partner, almost one-quarter (23%E) had left their marriage after less than 4 years of matrimony.

Some researchers believe there is credible evidence that "...multiple marriers are different in personality and behavior (sic) from those who remarry only once."24 A 1990 U.S. study specifically of serial marriers agreed that both men and women married multiple times have higher levels of anxiety than those married only once or twice; multiply-married women also reported more psychological distress than other married women, even after controlling for their divorce history.25

Believing in Marriage produces a stronger marriage

This psychological profile – however brief – may help to shed some light on a rather counterintuitive finding from the GSS. One would expect that people who marry multiple times are keen believers in the value of marriage and family, but the data tell a different story.

Serial marriers are significantly less likely to claim that being married is important or very important to their happiness, at 69% versus 82% of people who married only once (including those divorced or widowed as well as those still married). Of course, deeply held beliefs can be altered by a person's experience, especially a severely negative experience such as the failure of their marriage. But this lack of commitment to the idea of marriage may become a self-fulfilling prophecy, since it is a key factor associated with marital collapse. People who do not believe that marriage is important for them to be happy have a predicted risk of both first and subsequent marriage failure 170% to 330% higher than people who feel it is very important, when all other variables are controlled for.

Similarly, serial marriers are almost twice as likely to say they would not stay in a bad marriage even for the sake of their children (50% compared with 28% of once-marrieds). Of course, this is probably a very hypothetical question for most once-marrieds, who may overstate their case, while serial marriers might have a more realistic idea of how much they are prepared to tolerate. Nevertheless, compared with those who believe they would stay in an irreparable marriage for the sake of their kids, the predicted risk of a first or second marriage dissolution is 69% to 116% higher for people who are prepared to leave.


Current events may suggest that the estate of marriage is in disarray. Some people would argue that society's acceptance of the individual's demand for personal fulfillment has freed irresponsible and hedonistic people to flit from one spouse to another.

However, marriage still seems to possess an aura that elevates it above a simple living arrangement. Most Canadians marry once and only once; less than one percent walk down the aisle more than twice. Married couples generally have "greater commitment and higher relationship quality" than partners in common-law unions,26 which suggests something about the transcendent nature of the marriage bond itself.

The factors associated with the break-up of a first marriage tend to be different than those that are significant risk factors for the dissolution of a subsequent marriage. In general, however, the predicted likelihood that their marriage will succeed is higher for people who marry in their 30s, did not live common-law before the wedding, have children, attend religious services, are university educated, and believe that marriage is important if they are to be happy.


  1. Cherlin, A. 2004. "The Deinstitutionalization of American Marriage." Journal of Marriage and Family 66, 4: 848 to 861.
  2. Milan, A. 2000. "One hundred years of families." Canadian Social Trends 56 (Spring 2000): 2 to 12; Statistics Canada. 2003. "Update on families." Canadian Social Trends 69: 11 to 13.
  3. Statistics Canada. The Daily, March 9, 2005.
  4. Statistics Canada, 2003.
  5. Clark, W. 2000. "Patterns of religious attendance." Canadian Social Trends 59: 23 to 27.
  6. Burnham, T. and J. Phelan. 2000. Mean Genes. Penguin Books: NY, NY.
  7. Le Bourdais, C., G. Neill and P. Turcotte. 2000. "The changing face of conjugal relationships." Canadian Social Trends No. 56: 14 to 17.
  8. Milan, A. 2003. "Would you live common-law?" Canadian Social Trends No. 70: 2 to 6.
  9. Burnham and Phelan, 2000. Wright, Robert. 1994. The Moral Animal: Why we are the way we are – The new science of evolutionary psychology. Vintage Books: New York.
  10. According to 2003 data, 50% of marriages in Quebec ended in divorce within 30 years. Statistics Canada. The Daily, March 9, 2005.
  11. Adams, M. 2003. Fire and Ice: The United States, Canada and the myth of converging values. Penguin Canada; Toronto: 77 to 102.
  12. Bartlett's Familiar Quotations: Fifteenth and 125th Anniversary Edition. 1980. Emily Morrison Beck (ed.). Little, Brown and Company; New York: 354 (22).
  13. The rate of remarriage in the U.S. was about 75% in the 1990s. Coleman, M., L. Ganong and M. Fine. 2000 "Reinvestigating Remarriage: Another decade of Progress." Journal of Marriage and the Family 62, 4: 1288 to 1307.
  14. Coleman, Ganong and Fine; 2000.
  15. Intrinsic reasons for marriage are couple-based and relate to level of intimacy, e.g. "We are close and intimate. We have special ways of demonstrating affection and letting each other know how we feel." Extrinsic reasons for marriage were more instrumental: "He/she knows what I want. He/she meets my needs. He/she is someone my parents would approve of. He/she is well-liked by my friends. People are impressed by my choice." Kurdek, L.A. 1989. "Relationship Quality for Newly Married Husbands and Wives: Marital History, Stepchildren, and Individual-Difference Predictors." Journal of Marriage and the Family 51, 4: 1053 to 1064.
  16. L.A. Kurdek. 1991. "The Relations between Reported Well-Being and Divorce History, Availability of a Proximate Adult, and Gender." Journal of Marriage and the Family 53, 1: 71 to 78.
  17. Coleman, Ganong and Fine, 2000; Vemer, E., M. Coleman, L.H. Ganong and H. Cooper. 1989. "Marital Satisfaction in Remarriage: A Meta-Analysis." Journal of Marriage and the Family 51, 3: 713 to 725. Kurdek, 1989; Gelissen, J. 2004. "Assortative mating after divorce: a test of two competing hypotheses using marginal models." Social Science Research 33: 361 to 384.
  18. Kurdek, 1989.
  19. Kurdek, L.A. 1990. "Divorce History and Self-Reported Psychological Distress in Husbands and Wives." Journal of Marriage and the Family 52, 3: 701 to 708.
  20. Gelissen, 2003.
  21. Gelissen, 2003.
  22. Coleman, Ganong and Fine, 2000.
  23. Vemer, Coleman, Ganong and Cooper, 1989; Kurdek, 1989.
  24. Vemer, Coleman, Ganong and Cooper, 1989.
  25. Kurdek, 1990.
  26. Wu. Z. and C.M. Schimmele. "Repartnering after First Union Disruption." Journal of Marriage and the Family 67 (February 2005): 27 to 36.
Full article in (PDF)


Warren Clark is senior analyst and Susan Crompton is Editor-in-Chief of Canadian Social Trends. They can both be reached at

You need to use the free Adobe Reader to view PDF documents. To view (open) these files, simply click on the link. To download (save) them, right-click on the link. Note that if you are using Internet Explorer or AOL, PDF documents sometimes do not open properly. See Troubleshooting PDFs. PDF documents may not be accessible by some devices. For more information, visit the Adobe website or contact us for assistance.

Home | Search | Contact Us | Français Top of page
Date modified: 2008-11-21 Important Notices