The Implications of Parents’ Conjugal Histories for Children

Until recently, any young couple wishing to live together and to have children had no other option but marriage, the only acceptable framework in which to start a family. Conjugal and reproductive behaviour was regulated by law. Once married, individuals were expected to remain so, and divorce would only be granted in very specific circumstances. Just as forming a family was the object of rules and rituals, so would separation be required to respect certain formal arrangements.

In the present day, conjugal and familial matters don’t always participate in a legal structure. Presently, self-advancement is the focus of much of society, and the social pressure to have a partner and family don’t carry a lot of weight. This day and age is more about self fulfillment then ensuring a continuation of stability and structure. This change in values perhaps explains the ease with which both marriage and divorce are losing ground in Canada today. Living together and breaking up need not be entangled in legal procedures when the flexibility of conjugal ties becomes more desirable than the perpetuation of the family.

However, the growing popularity of cohabitation as the chosen context in which to give birth and raise a family does bring up some disturbing questions with regard to children in this context. Whether the parents were legally married or just living together, children’s needs must be taken care of following a separation, be there a divorce or not. How are matters settled when no legal procedures surround parental break up? Are the issues of custody, access and child support more likely to “escape” the judiciary when parents were just cohabiting? Does this influence the quality and quantity of contact that children maintain with both parents after separation? So the question remains as to whether a child can flourish and grow as a person when the parents have separated.

The data

The data come from the “Family History and Custody” section of the 1994-1995 questionnaire of the National Longitudinal Survey of Children and Youth (NLSCY). The sample is representative at the provincial level.

The changing family context

From marriage without cohabitation to cohabitation without marriage

The 1970s and 1980s have witnessed a small revolution in the matters of conjugal and parental life in Canada. The 1950s’ legacy to the 1960s had been the traditional family, based on legal marriage into which couples would commit themselves for life. In the aftermath of the 1968 Divorce Act and the tremendous rise of broken marriages that followed, the idea emerged and was put into practice that such an enduring commitment as marriage should be tested first. Cohabitation was progressively adopted by an increasing number of young couples, who either failed the test and were grateful to have avoided the rigidity of marital vows, or else moved on to marry in the blissful state of knowing their choice was the right one. So they thought, for the 1980s confirmed that what seemed at some point to be the perfect choice could later become an impossible alliance to maintain. Why then marry if only to risk going through the painful procedures of divorce? The answer became obvious in the 1980s, with increasing proportions of cohabiting couples deciding not to marry, even following the birth of a child, a living proof that cohabitation was becoming as durable as marriage in the 1950s. At least, this was what some thought.

Figure 1 summarizes the evolution of conjugal mores over the last four decades through the family context at birth of several cohorts of children. From an unquestionably monolithic pattern in the early 1960s, where over 90 percent of all children were born to parents who had married without cohabiting before, this picture changed dramatically over the 1970s and 1980s, with the result that fewer than 40 percent of children born in the early 1990s were brought into the world in such traditional marriages. Cohabitation has quietly grown in popularity, and by the early 1990s, a third of Canadian children were born to parents who had cohabited before marrying. The most striking result was that 20 percent of the youngest cohorts were born to cohabiting couples, a percentage that has not stopped growing to this day.


However, Table 1 shows that this trend did not happen at the same rhythm across the country. Of Canada’s five major regions, Quebec and Ontario present the two extreme cases as far as the adoption of cohabitation is concerned. By the time the 1993-1994 birth cohorts (the youngest in the NLSCY) came along, 43 percent of Quebec babies were born to cohabiting parents compared to only 12 percent in Ontario. On the other hand, 49 percent of all births in Ontario were still happening within “traditional” marriages as opposed to only 23 percent in Quebec. In the rest of the country, conjugal behaviour stood pretty much in between, but marrying directly without first cohabiting had become a minority choice. Official birth statistics confirm the trend observed here: in the last few years, increasing proportions of children (over 55 % in Quebec and 30 % in Canada as a whole) were born “out-of-wedlock,” that is to single mothers for a minority of children, and to cohabiting parents for a majority.


Cohabiting parents and the increased risk for children to experience separation

By definition, flexibility of conjugal bonds implies that although couples that started living together might have been committed to one another, breaking up has become accepted and relatively easy to do. The fact that divorce and separation rates have continued to rise since the 1970s has much to do with this attitude, whether couples got married but expected to be able to divorce easily, or whether they simply cohabited and thought they could stop living together without ever resorting to any legal framework.

Having children does not seem to be a deterrent to these new attitudes. Figure 2 shows that over the last four decades, increasing proportions of Canadian children have experienced single parenthood mainly due to their parents’ separation. It also shows that this is happening at increasingly younger ages, because of the short life expectancy of their parents’ union.

Figure 2 below illustrates that approximately 25% of children who were conceived in the early 60’s, were born in a single mother home, or experienced their parents separation before reaching the age of 20. About 50% of the time, the child was at least 10 years old when the separation occurred. This would assume that a ’68 amendment to the Canadian Divorce Act,  which allowed couple to divorce more easily, was at least partly responsible. Twenty years later, 25 percent of children from the 1983-1984 birth cohorts had experienced single parenthood, but this time by the age of 10. The acceleration continues, and in the 1987-1988 birth cohorts, nearly 25 percent of children had experienced single parenthood as early as age six.


This trend shows no sign of slowing down, given that the risk for children to go through parental separation is directly linked to the type of union in which they were born. Figure 3 illustrates the cumulative percentage of children born in two-parent families, who have experienced parental separation according to the type of parental union. The figure is based on the 1983-1984 cohorts, that is, on children who had reached the age of 10 by the time the NLSCY was conducted.


The choice of parents to cohabit rather than marry has far-reaching consequences on the chances that a family will survive. Parents who cohabited before marrying are more likely to separate than parents who married without cohabiting, but they are clearly less likely to do so than couples that remained in a common-law union.

Children born into relationships where the parents haven’t entered into any legal agreement are at a much higher risk of having to go through the emotional roller coaster that is a separation. As such, an astounding 63% of children born in 83-84, had lived through a separation by their parents by the ripe old age of 10. However, those conceived in what would be considered a traditional family experienced a separation by the age of 10 at a rate of only 14%.

Given the higher propensity of cohabiting couples to separate and to do so early after the beginning of the union, it comes as no surprise that their children are overly represented among children from broken families.

Offspring that were alive and still residing with parents that were cohabiting when the survey was being conducted accounted for 13% of all children in Canada between the ages of 0 and 11 whose parents were together when the child was born. However, those children from from failed common-law marriages took up a rather substantial 34% of those who had experienced their parents separating. The question of whether or not married and cohabiting parents settle the issues of custody and child support differently is thus of the utmost relevance, since it concerns ever growing proportions of future citizens.

What are the consequences for children of their parents’ changing conjugal behaviours? Past research has shown that common-law partners are more egalitarian in the division of domestic labour (including the care of children) than married couples.[ 1 ] Can we thus expect unmarried parents to be more inclined than married parents to share the custody of the child following a separation? Furthermore, are married parents who chose not to legalize their break-up through divorce more likely to settle privately issues of custody and child support and to reach a common agreement than those who chose to divorce? The next sections address these questions.

Custody arrangements

After separation, how do parents negotiate custody arrangements? To shed light on this issue, the NLSCY asked parents if there “was a court order concerning the child’s custody” when they separated. Parents who answered affirmatively were further asked into whose custody did the court order the child to be placed.

Divorced parents are much more likely to have a court order for custody

Table 2 below helps illustrate the number of times court orders were pursued in order to gain custody of a child. An astounding 50% of parents who had split from their significant other reported they had, or were obtaining, a court order in regards to the children involved in the relationship.

However, this percentage varies according to the type of parental separation. Hence, divorced parents appear much more likely than separated parents to hold a court order for custody: 70 percent of the former either had a court order or were in the process of obtaining one, a percentage that is nearly twice as high as that observed for married parents who had not legalized their separation through divorce or for those who were cohabiting at the time of separation. This result is not surprising, given that the issue of custody generally constitutes a part of petitions for divorce.


The frequency of court orders is higher in Quebec and among parents reporting high levels of tension over custody and visiting rights


The probability that a court order exists is linked to factors other than the type of union and of separation. Our research has shown that the time elapsed since separation, the degree of tension existing between parents and the province of residence are all related to the process that follows separation and that leads parents to make decisions on how they will share child-care responsibilities.

Separating when children are involved is a complex process that evolves with time. If the physical split of a family can happen overnight, the resolution of all questions related to the legal and practical aspects of life after separation takes time. Division of property and decisions concerning children are often difficult, and the arrangements first established may prove unworkable in day-to-day life. So, the situation might evolve as the length of the separation increases. The process following separation is also influenced by the degree of tension that exists between parents over custody and visiting rights, and this, in turn, affects the likelihood that parents will turn to the court, this recourse being more frequent when the level of tension is high. Finally, the province of residence also has an impact on the way separated parents solve children’s issues, given that family courts are under provincial jurisdiction, and that changes in the patterns of family formation have not occurred at a similar pace across Canada.

Given these results, we systematically control for the effects of duration, level of tension and province of residence in our analysis of the various steps that separated parents take in making decisions about their children. The method of analysis we use is logistic regression, and the coefficients presented in tables 3, 5, 7 and 9 represent the net effect that any factor exerts on the probability of making a decision, once the effect of all the variables included in the equation has been controlled for. For example, Table 3 examines how the type of parental separation has affected the likelihood of parents obtaining a custody court order, after controlling for level of tension between parents, region of residence and duration since separation. Three types of parental separation are distinguished. One category is used as the reference group in the equation, and the coefficients attached to the two other categories compare the effect that the latter have in relation to the reference category. The coefficients presented in Table 3 are the odd ratios. A coefficient lower than 1.0 indicates that the category reduces the probability of having a custody court order, while a value higher than 1.0 indicates that the category increases this probability. More precisely, the results show that, even after controlling for the three factors mentioned, the type of parental separation remains clearly associated with the probability of parents having a custody court order (as shown in Table 2); separated parents, who are not divorced, are three times less likely to have a custody court order than divorced parents.


Custody is given to mothers in overwhelming proportions

In cases where a court order existed, mothers were given the custody of children in overwhelming proportions. In Canada as a whole, 80 percent of children under the age of 12 were placed in their mothers’ custody; only seven percent were placed in their fathers’ custody, with the remaining 13 percent in shared custody arrangements. Although the percentage of children who were put under their mother’s custody by the court is relatively high, one should note that it is lower than the percentage of children who were, in fact, living with their mother after the separation occurred. Family courts do not therefore appear to be solely responsible for giving mothers the exclusive custody of children, as media or associations of non-custodial parents sometimes imply. We examine the level of contact that fathers maintain with their children in a later section. For now, let us turn our attention to the issue of child support.

Child support agreement

This section examines how the issue of child support is settled following a separation. In the NLSCY, parents were asked “what type of agreement (none, private, court-order) was made regarding support/maintenance payments” when they separated or divorced. They were also asked if this agreement was made for child support only or if it also addressed the issue of spousal support, and how regular the maintenance support payments have been. No definition of the term “private agreement” was provided, so it remains difficult to interpret some of our results. For example, parents might have reached a truly private agreement between themselves, or they could have been referring to a written separation agreement, comparable to a court order, which can be enforced by, or registered with, an enforcement agency. One should keep this in mind when analyzing the data.

Children of divorced parents are more likely to be covered by some form of agreement

As can be seen in Table 4, for roughly a third of children who had experienced parental separation, the parents said there was no agreement concerning child support payments. For another third, the parents reported that they had a court order agreement or that they were in the process of obtaining one; and for the last third, parents declared having reached a private agreement. Children whose parents had actually divorced at the time of the survey were by far the most likely to be covered by some type of agreement: over half (57 %) were covered by a court order (including those in progress), and no agreement was reported for only 17 percent of children. In comparison, 42 percent of children from broken common-law unions were left without any form of child support agreement, and they were followed closely by children whose married parents had not yet divorced at the time of the survey (37 %). However, the latter were more likely to be covered by a private agreement than other children. Is this result due to a stronger desire of married parents who chose to separate “outside the law” also to solve privately the issue of child support, or is it simply due to the fact that these separations were more recent and that divorce had not yet been obtained?


Table 5 examines the effect, net of other factors, that the conjugal history of parents exerts on the likelihood that they report either type of agreement concerning child support. It is based on a multinomial logistic regression. The results presented in Table 5 predict the likelihood of separated parents obtaining a court order agreement (including those under progress) or reaching a private agreement, as opposed to having no agreement.


After controlling for the existence of a custody court order, in addition to the variables introduced earlier into the analysis, the type of parental separation remains strongly associated with the type of agreement reached over child support. Hence, children whose parents were not divorced at the time of the survey were clearly less likely to be covered by some type of agreement: children whose parents separated after their marriage or common-law union had collapsed had approximately a third the probability of having a court order agreement, and between a third and a half the probability of having a private agreement, as those whose parents were divorced. Therefore, married parents who had separated but were not yet divorced at the time of the survey did not appear more inclined to solve privately the issue of child support, as was suggested in Table 4.

The type of child support agreement is linked to existence of a custody court order

Not surprisingly, the type of agreement concluded for child support is linked to the way in which custody issues were settled by separated parents. Children who were covered by a custody court order were by far the most likely also to be covered by a court order agreement concerning child support payments: they were 13 times more likely to be covered by a court order agreement than by no agreement, compared to children for whom there was no custody court order. By contrast, they were less likely (20 percent less) to be the object of a private agreement.

These results indicate that children who were born to married parents who divorced are more likely to be “protected” by decisions made by the courts: not only will the courts have more often approved or imposed the choice of their custodial parent, they are also more likely to have confirmed or imposed a child support agreement. On the other hand, children whose parents have not divorced are more likely to be covered by a private agreement or by no agreement at all. What are the implications of these findings? Are children “protected” by court decisions better off than children who are not? Do they benefit from better relationship with their non-custodial parent, both in terms of contacts and financial support? Let us now turn to these issues.

Frequency of contact with the non-custodial parent

As mentioned earlier, in overwhelming proportions, mothers do get the sole custody of children following a separation or divorce. Table 6 shows that 85 percent of children under the age of 12 were living exclusively with their mother when the NLSCY was conducted in 1994-1995; half of the remaining percentage were living with their father only, and the other half were under shared custody. Among the children living with their mother, half visited their father on a regular basis (i.e., 25% every week and 17.5% every two weeks); close to 30 percent (25% out of 85%) of these children saw their father irregularly (once a month, on holidays or at random), and 20 percent never visited him.


The pattern of contact varies somewhat according to the type of parental separation. On the one hand, children from broken common-law unions were found to be more likely to live with their mothers at the time of the survey than children whose parents had been married; they were also less likely to live in a shared custody arrangement (4 % compared to around 10 %). On the other hand, children whose married parents had separated but not yet divorced at the time of the survey were far more likely to have close contacts with their father: nearly 60 percent of them were either living with their father, sharing residence with both parents or visiting their father weekly. To what extent is this result due to the fact that these separations were more recent? In other words, will the patterns of separated married parents resemble those of divorced parents as time goes by and as divorce is obtained by a growing proportion of parents? And more important, will the recourse to courts reinforce the ability of fathers to maintain contacts with their children?

To answer these questions, we conducted a logistic regression that aims to predict the propensity of fathers to maintain any contact — regular or irregular — with their children. Counted as having contacts with their father are children who were either living with their father only, under shared custody or visiting their father every week or every two weeks, or visiting irregularly. The results presented in Table 7 show the net effect of each variable, once the others are taken into consideration.


Children of broken common-law unions are less likely to maintain contact with their father

Even after controlling for a series of variables, such as the ways in which custody and child support matters were resolved, the likelihood of fathers maintaining contact with their children remains significantly associated with the type of parental separation. Among all children who experienced parental separation, those whose parents were still in a common-law union when they broke up had less chance of seeing their fathers than those whose parents were divorced (coefficient of 0.567). On the contrary, children whose married parents were separated but not divorced were almost twice (coefficient of 1.701) as likely to have maintained contact with their father. This result suggests that married parents who chose to separate “outside the law” have different behaviours than divorced parents in regards to maintaining contact with their children. The former are perhaps less likely than the latter to have engaged in a new union, and to have established a new family. Indeed, research conducted in the United States has shown that existing relationships between fathers and children tend to weaken as fathers enter new unions and form new families.[ 2 ]

The ways in which custody and child support issues were settled also influence the probability that fathers maintain contact with their children. On one hand, other things being equal, children covered by a custody court order (including those under progress) have approximately 60 percent as much chances as those without a custody court order to see their father regularly or irregularly. On the other hand, the existence of some type of child support agreement clearly increases the probability of children maintaining contact with their father: a court order agreement doubles their chances of having any contact with their father. The conclusion of a private agreement exerts an even stronger effect: children in these cases are eight times more likely to see their fathers.

One should note that the propensity of fathers maintaining contact with their children decreases as the length of separation increases. Hence, fathers who had separated two to four years before the survey were only roughly half as likely to maintain contact with their children, compared with those whose separation was less than two years old; with the probability dropping slightly further after five years or more of separation (results not shown). This result confirms the necessity of controlling for the time elapsed since separation when studying the relations existing between separated fathers and their children.

These results are quite fascinating. On one hand, children whose parents chose to start and break their family within a legal framework are more likely to maintain contact with their father. This finding thus tends to indicate a greater commitment of divorced fathers toward their children, following a separation. On the other hand, recourse to the tribunal to determine custody or child support arrangements is not associated with as high a probability of a continued father-child relationship. Contact between father and child is much more likely where parents have reached a private agreement. This is probably partly due to the fact that courts often have to intervene in the most complicated cases, where relations between parents are already very difficult.

Payment of child support

Let us now turn to the more practical expression of fathers’ fulfilment of their responsibility toward their children, namely the payment of child support. To what extent is the frequency of contact that fathers maintained with their children associated with the promptness with which they provide financial support? To what extent is the way child support agreement was reached related to the regularity with which fathers fill their financial obligations? These are closely related issues that deserve further analysis. Unfortunately, the question concerning the regularity of child support payments was not posed to parents who mentioned that no agreement existed regarding support payments. Thus, we cannot compare how these separated parents fulfil their financial obligations toward children in relation to parents who have reported some form of agreement. However, we can compare parents who have obtained a court order agreement with those who concluded a private agreement. Children living with their father on a full-time or part-time basis were also excluded from the analysis, since there are usually no child support payments in such situations. The following analysis is thus based on children who were living exclusively with their mother at the time of the survey and who were covered by a child support agreement.

The regularity of payments is influenced by the type of child support agreement

Table 8 shows that, in general, a larger percentage of children who were covered by what their parents described as a private agreement received regular payments than children whose parents reported having a court order agreement. Two thirds of children under private agreements benefited from regular support payments, compared to 43 percent of children whose parents stated they had a court-ordered agreement. Moreover, cases where there had been no payments in the six months preceding the survey were found to be much more common where a court order agreement existed than where support payments were dealt with through a private agreement (30 % versus 12%).


This trend holds true regardless of the type of parental separation. Hence, 68 percent of children whose divorced parents concluded a private agreement received regular child support payments, and only eight percent received no payment in the last six months. In the case of court-ordered agreements, the percentage of children benefiting from regular payments drops to 43 percent while that receiving no payment increases to 25 percent. Children from broken common-law unions fared much worse in the case of private agreements: only slightly more than half enjoyed regular payments, and 22 percent were left without any support payment in the six months prior to the survey. This result suggests that unmarried fathers are perhaps not as committed to ensure children’s well-being as we first expected; on the contrary, flexible conjugal ties would seem to be associated with a lower commitment to children, both in terms of contact, as we have seen above, and of financial matters, as we shall now see.

The regularity of child support payments is closely linked to the frequency of contact

To what extent does the association observed between regularity of payments, and the type of parental separation and support agreement hold when one controls for the other variables included previously into the analysis? To answer this question, we conducted a logistic regression that examines the propensity of fathers making child support payments. The results of this analysis are presented in Table 9.


The ability of fathers to maintain contact with their children strongly influences their propensity to provide child support payments, and the impact of this variable remains important, even after controlling for the type of parental separation and the type of custody and child support arrangements. As can be seen in Table 9, children who saw their father were significantly more likely to receive payments than those who had no contact with him — twice as likely if they saw their father irregularly, and nearly seven times as likely if they saw him regularly.

Interestingly, the propensity of children to receive child support payments also remained strongly associated with the type of parental separation. Children coming from a broken common-law union were only approximately 25 percent as likely to obtain support payments as were those whose parents had divorced.


The rise in conjugal instability, partly related to the trend toward common-law unions, is affecting a growing proportion of children. Our objective was to explore the relationship between this new phenomenon and the way parents are taking responsibility for their children after separation. The results show that parents who broke from a common-law union are indeed behaving differently from other parents. Our analysis of the diverse steps of the process that follows separation sheds light on the fact that separated parents, in particular those who were in a common-law union, tend to be less likely to go to court when decisions have to be made about child custody and support agreement than are divorced parents. Moreover, the results show that fathers who had children in a common-law union are less likely to maintain contact with their children and, similarly, also less likely to support them financially than parents who were married. Our analysis illustrates another interesting fact: parents who can reach a support agreement privately seem more likely to develop relationships that favour the continuation of father-child contact and support payments. Recourse to the legal system for marriage and separation is associated with a greater commitment of parents toward their children, but recourse to the courts for custody arrangements is not, probably because courts are often called upon to find solutions for couples who have greater difficulties than others.

From a policy point of view, the most meaningful finding is the fact that private arrangements between separated parents seem to be more favourable to the child’s well-being. This finding suggests that priority should be given to the development and reinforcement of mediation services; in this manner, parents could be helped to find conditions of custody and support that make the interests of the children the ultimate goal.

In terms of the court system itself and how it intervenes, many questions remained unanswered. Once a separated couple asks the courts to act, is the process such that it will prevent any improvement in the deteriorating relationship, or will it favour agreement? Much more research needs to be done on that particular, and delicate, aspect of the problem.

More broadly, education about parental responsibility seems to be at the core of finding a new equilibrium within the family system, in particular with regard to fatherhood. In a general context in which institutional constraints and supports are less valued, only public debate and reflection about the evolution of the relationship between parents and children can bring a new sense of commitment toward children.

Read more:


*   Nicole Marcil-Gratton, Céline Le Bourdais and Évelyne Lapierre-Adamcyk are Professors at the Centre Interuniversitaire d’Études Démographiques, Université de Montréal and at the Institut National de la Recherche Scientifique (INRS). This paper is derived from a report (Marcil Gratton and Le Bourdais, 1999) published by the Child Support Team of the Department of Justice Canada. An earlier version of the paper was presented in session 302, “Family Disruption and Children’s Well-Being” of the 1999 Annual Meeting of the Population Association of America, held in New York. The authors wish to thank Paul-Marie Huot and Julie Archambault for data analysis and the preparation of figures.

 1.  C. Le Bourdais and A. Sauriol, La part des pères dans la division du travail domestique au sein des familles canadiennes, Montreal, INRS-Urbanisation, serie “Études et documents,” no. 69, 1998; B.A. Shelton and D. John, “Does Marital Status Make a Difference? Housework Among Married and Cohabiting Men and Women,” Journal of Family Issues, Vol. 14 (1993), pp. 401-420.

 2.  E.C. Cooksey and P.H. Craig, 1999. “Parenting From a Distance: The Effects of Paternal Characteristics on Contact Between Nonresidential Fathers and Their Children,” Demography, Vol. 35 (1999), pp. 187-200; B. A. Shelton and D. John, “Does Marital Status Make a Difference? Housework Among Married and Cohabiting Men and Women,” Journal of Family Issues, Vol. 14 (1993), pp. 401-420.