When Parents Separate: Further Findings from the National Longitudinal Survey of Children and Youth

2004-FCY-6E

III  SHARING CUSTODY

One of the most challenging issues faced by separating parents is how to share the physical care and economic support of their children. Are the children to live with the mother or the father after separation, or will each parent create a family home and have the children with them part of the time? Who will be responsible for the financial cost of raising the children now that the household income that previously supported one household is shared between two? Reaching an agreement in the emotionally charged climate that surrounds most couples in the throes of separation is not easy. Nonetheless, many parents come to an agreement between themselves, with or without what may be costly legal involvement. At other times, they are unable to agree and the court imposes a solution based on what is perceived to be in the best interests of the child.

The divorce statistics collected by Statistics Canada, the only national data source on how parents organize care after separation before the NLSCY, suggest a trend towards greater diversity in custody arrangements. Based on this information, Figure 3.1 illustrates how custody arrangements have evolved for minor children whose custody was included in a court order from just after the 1968 Divorce Law up to the year 2000.

  • Despite minor fluctuations during the 1970s and 1980s, the proportion of children put into mother, father or joint custody in 1990 was almost identical to that of 1970, with sole mother custody granted for just under three-quarters of children, and the rest divided fairly evenly between father and joint custody.
  • The 1990s saw a rapid rise in the popularity of joint custody; by the year 2000, well over one-third of children (37%) were placed in the custody of both divorcing parents. Sole father custody, on the other hand, actually declined during the period, from around 15% to under 10%. These changes meant that, by the year 2000, sole mother custody was awarded for only just over half the children concerned.

Figure 3.1  Distribution of minor children for whom there was a custody order at divorce, according to the type of arrangement, Canada, 1970-2000

However, divorce statistics provide an indication of legal custody rather than actual living arrangements. Although joint custody means that separated parents continue sharing all the major decisions concerning their children, it does not necessarily mean that they share their children's physical custody. Consequently, unlike sole custody, where children generally live with the custodial parent, joint custody covers a variety of living arrangements. These statistics are also limited in terms of the information they provide on the way parents share children's living arrangements after separation in two ways:

  • First, by referring only to divorce, they provide no information on custody arrangements reached by cohabiting couples or by married couples who separate but do not divorce.
  • Second, not all divorces including minor children are covered by these statistics; custody arrangements settled privately prior to divorce are not always entered on the divorce record.

By collecting information on all types of separation, and on both legal custody arrangements and actual living arrangements, the NLSCY made it possible to fill in these gaps (see Marcil-Gratton and Le Bourdais, 1999). One drawback with the Cycle 1 data used in earlier analyses, however, was that only children up to the age of 11 years were included. By the third cycle, in 1998-99, the oldest cohorts of children had reached the age of fifteen years which, coupled with samples of young children added at each cycle, made it possible to include a wider age range of children (0-15 years). In this section, we use this sample to extend the analysis of the evolution of custody arrangements through time, with particular reference to shared custody; in the rest of the text, shared custody refers to shared physical custody, and is used synonymously with "shared living arrangements."

At the NLSCY, separated parents were asked whether there was a court order concerning the child's custody; they were then asked specifically if the court order had placed the child in shared physical custody. However, an earlier analysis of Cycle 1 data (Marcil-Gratton and Le Bourdais, 1999, Table 7) showed that this still did not mean that children had shared living arrangements, either at separation or later on. In fact, among children aged 0-11 years, under one-quarter of children with shared custody as part of the court order actually lived part of the time with each parent. More than two-thirds lived exclusively with their mother at separation, and 11% exclusively with their father. The analysis also showed that:

  • Under half the children had a court order[7] concerning custody arrangements.
  • Shared living arrangements were more frequent when custody was settled privately.

In Table 3.1, these figures are compared with estimates made for all separations experienced by the broader sample of children, aged 0-15 years in 1998-99. Among children with custody arrangements contained in a court order, shared custody increased (up from 12.6% in 1994-95 to 17.1%), as did the proportion of these children actually sharing living arrangements (from 2.6% to 5.6%). These figures confirm, nonetheless, the tenuous relationship between shared physical custody included in a court order and actual living arrangements:

  • Only one-third of children for whom shared custody was granted actually shared living arrangements between parents at separation.
  • Children were considerably more likely to live in shared custody when parents came to an agreement without legal intervention (11.6%) than with it (5.6%).

Table 3.1  Distribution of children with separated parents in 1994-95 and 1998-99, according to type of custody and living arrangements at separation, NLSCY, Cycles 1-3

Children aged 0-11 years in 1994-95
Court order No court order Total
Custody
order
Living arrangements Living arrangements
· Mother 80.8 89.5 86.1 86.8
· Father 6.6 7.9 5.4 7.0
· Shared 12.6 2.6 8.5 6.2
Total 100.0 100.0 100.0 100.0
N 1215   1730 3276 a

Children aged 0-15 years in 1998-99
Court order No court order Total
Custody
order
Living arrangements Living arrangements
·Mother 77.7 88.0 82.0 85.0
· Father 5.2 6.4 6.4 6.5
· Shared 17.1 5.6 11.6 8.5
Total 100.0 100.0 100.0 100.0
N 2060   2142 4407 b

Although children are still far more likely to remain in their mother's physical care when their parents separate, this comparison suggests some movement towards a greater equality in sharing children's physical care after separation. In the next section, we look more closely at the question of whether the trend towards increased father involvement after separation, suggested by the rise of joint custody in the divorce statistics, is also reflected in children's living arrangements after separation. We will then explore which forms of shared living arrangements (alternating weeks, etc.) are most commonly selected by parents sharing custody, and how these arrangements evolve over time.

Are more children living in shared custody when parents separate?

The divorce statistics presented earlier indicate a rapid increase in joint custody throughout the 1990s; the information in Table 3.1 suggests that this trend extends to actual living arrangements, though to a lesser extent. Figure 3.2 compares living arrangements made at separation for children whose parents separated in the early 1990s (1992-94) with those separating in the late 1990s (1996-98); results appear to confirm this trend. In the four-year period dividing the two interviews, the proportion of children in shared living arrangements rose from 9% to 13%, while the percentage of children in their mother's custody dropped from 85% to 78%.

Figure 3.2  Distribution of children whose parents separated in the two years preceding Cycles 1 (1994-95) and 3 (1998-99), according to the living arrangements at separation, NLSCY

However, with mother custody most common among younger children, and other forms of custody more common later on, it is possible that the older age of the 1998-99 sample explains this evolution. We used multinomial logistic regression techniques to evaluate how far children's age, as well as a number of other variables, explains the apparent trend towards greater diversity of living arrangements. The model assesses the likelihood that children live in shared custody or with their father rather than residing with their mother when parents separate, and the results, in the form of odds ratios are presented in Table 3.2.[8]

Table 3.2  Impact of given variables on the probability that children (aged 0-15 years in 1998-99) live in shared custody or with their father rather than remaining with their mother when parents separate, NLSCY, Cycle 3 (Multinomial logistic regression odds ratios1—N = 4377)

Year of separation
Variables Living arrangements (with mother)
Shared With father
Child's age at separation 1.143 *** 1.198 ***
· (Before 1991) 1.000 1.000
· 1991-1994 1.153 .858
· 1995-1998 1.487 * .650 *

Type of agreement
Variables Living arrangements (with mother)
Shared With father
Child's age at separation 1.143 *** 1.198 ***
· (Court order, or in progress) 1.000 1.000
· No court order 1.874 *** .958

Sex of child
Variables Living arrangements (with mother)
Shared With father
Child's age at separation 1.143 *** 1.198 ***
· (Boy) 1.000 1.000
· Girl .930 .861

Region of Canada
Variables Living arrangements (with mother)
Shared With father
Child's age at separation 1.143 *** 1.198 ***
· (Canada, excluding Quebec) 1.000 1.000
· Quebec 2.450 *** 1.334 *

Reference category given in parentheses.

1 Coefficients significant at: ***p<.001; **p<.01; *p<.05.

The model confirms the great importance of children's age at separation in the choice of living arrangements, with both shared and sole father custody significantly more likely among older children. It also confirms an increase in the popularity of shared living arrangements during the 1990s. Even controlling for age, children whose parents separated in the second half of the decade were significantly more likely to share time between the homes of both parents than those whose parents separated during the 1980s. This was all the more likely when parents reached a private custody agreement. The move towards increasing diversity in custody arrangements is particularly marked in Quebec. Children were significantly more likely to share living arrangements in Quebec than in the rest of the country. They were also more likely to live with their father after separation, going against the general movement away from sole father custody during the period. Finally, and perhaps surprisingly, the child's sex does not appear to have a significant impact on the type of custody.

As we have seen, shared custody is an arrangement that is not easily defined, and not only because there is no direct equivalence between a shared custody agreement and shared living arrangements. Shared living arrangements are flexible and, unlike sole custody, tend to evolve into other living arrangements with the passage of time since separation (Juby, Le Bourdais & Marcil-Gratton, 2003). In the following sections, we will explore what insight the NLSCY can give into the way parents who decide on shared physical custody of their children actually arrange this care, and how these arrangements evolve through time.

What types of sharing patterns are most commonly selected by parents who share custody?

For parents who decide to share the physical care of their children at separation many choices are open. The most popular options selected by parents[9] at the time of separation are shown in Figure 3.3. The fact that close to a quarter came to an arrangement other than those offered in the questionnaire indicates that there are many more possibilities open.

Figure 3.3  Distribution of children, aged 0-15 years in 1998-99, in shared custody at separation, according to the type of shared living arrangements, NLSCY, Cycle 1-3 (N=373)

  • Alternating each week between the parents is the most common form of sharing, with three children in ten in shared custody moving from their mother's to their father's home each week.
  • Much less commonly chosen at separation (8%) is a similar arrangement, but with children changing residence every two weeks.
  • Spending alternate nights with each parent was chosen even more rarely (5%), and is probably only workable when parents live in very close proximity to one other.
  • Almost one child in five (18%) spent the weekend with one parent, weekdays with the other. This may be the most practical option when one parent lives closer to the school than the other.
  • Interestingly, 17% of children declared living in shared custody spent only one weekend in two with one of their parents, and the rest of the time with the other. Time-sharing here is very similar to that arranged for many children considered to be in the sole custody of their mother or father.

Does the choice of arrangement depend on the child's age when parents separate?

Although all types of arrangements are found at all ages, there appear to be some age-related patterns.[10] The most common arrangement, alternating weeks, was chosen most frequently when children were aged between 3 and 11 years when their parents separated, less often for very young children, and rarely for adolescents. Interestingly, among adolescents, only one of the "standard" arrangements was selected with any frequency: spending weekends with one parent, weekdays with the other. Most of the others have an "other" arrangement, perhaps reflecting the greater autonomy of this age group both in terms of designing and implementing more diverse and flexible types of sharing. The standard sharing patterns are also less common among very young children. In fact, approximately 30% of children under the age of three in shared living arrangements lived primarily in one household, spending only one weekend in two with the other parent.

How do shared living arrangements evolve through time?

Only 40% of children who were in shared custody at separation were still alternating between their parents' homes by Cycle 3 and, of these, one quarter had a different arrangement from that first put in place (Figure 3.4). Half the children were living full-time with only one parent, 32% with their mother and 17% with their father. The rest (11%) were living with both parents in the same household after the latter had decided to get back together again at some point during the period. Two types of shared living arrangements were particularly likely to transform into sole custody:

  • Alternating every two weeks. Four-fifths of children with this arrangement were living in the sole custody of their mother or father, in equal proportions, by 1998/99. It is possible that this arrangement was selected at the start by parents whose circumstances were not particularly favourable to sharing custody.
  • One weekend in two. As might be expected, the majority of children were declared living full time with their mother by 1998/99. It is possible that the frequency of contact with their father did not change during the period; the change in declaration may simply reflect a change in the perception of the respondent.

Figure 3.4  Living arrangements by 1998-99 for children (aged 0-15 years) who were in shared custody at separation, NLSCY, Cycle 3 (N=318)

The move from shared custody to another form occurs gradually over time.Among children whose parents had been separated for under two years in 1998-99, 85% were still in some kind of shared custody arrangement. This proportion dropped rapidly with the time elapsed since separation—to 37% for those whose parents had been separated for 2-3 years, and to 13% after 4‑5 years. Among children whose parents had been separated six years or more, only 8% were still living in two households.

From shared to sole custody—a multivariate analysis

The importance of the duration since separation comes out clearly in the multivariate analysis presented in Table 3.3, which evaluates the impact of various factors on the likelihood that children moved into sole mother or sole father custody by 1998-99 rather than remaining in shared custody. This analysis shows:

  • The longer the time since separation, the more likely the move into sole mother or sole father custody becomes.
  • Shared living arrangements put in place for children whose parents separate when they are of secondary-school age are significantly less durable than those made for primary-school aged children. This result is at first surprising. One might assume that the greater independence of older children would facilitate the movement between parents' households, making them more likely to remain in shared custody. This does not appear to be the case. Having spent a larger part of childhood with both parents in a single residence, are they less willing to put up with the inconvenience of two homes? Does their greater autonomy in fact facilitate frequent contact with both parents, whether or not they actually live with them?
  • Shared custody is also more likely to develop into sole father custody when children are very young at separation. This result was also unexpected, given the greater propensity towards mother custody among young children. Could it be that, at times, shared living arrangements are chosen for young children by couples in which the father took the main responsibility for the children because the mother found it hard to cope, and is unable to cope alone after separation? More detailed, qualitative data are necessary to understand this process, as they are to understand the circumstances surrounding father custody more generally.
  • Girls are significantly less likely than boys to move from shared custody to father custody.
  • Certain types of shared custody arrangements also influence the move to father custody: children who alternated between parents every two weeks, or who were in some other type of shared arrangements, are more likely to be with their father later on.
  • Children living in Quebec in 1998-99 were significantly less likely to move from shared living arrangements to living exclusively with either their mother or father. Not only is shared custody more common in Quebec than elsewhere in Canada (see Table 3.2), it is also more durable.

Table 3.3  Impact of given variables on the probability that children (aged 0-15 years in 1998-99) in shared custody at separation are living with their mother or father in 1998-99 rather than remaining in shared custody, NLSCY, Cycle 3 (Multinomial logistic regression odds ratios1—N = 269)

Child's age at separation
Variables Living arrangements at C3
(shared custody)
With mother With father
Duration since separation 2.145 *** 2.274 ***

Child's age at separation
Variables Living arrangements at C3
(shared custody)
With mother With father
· (6-11 years) 1.000 1.000
· 0-5 years .968 3.361 *
· 12-15 years 4.570 * 14.371 ***

Sex of child
Variables Living arrangements at C3
(shared custody)
With mother With father
· (Boy) 1.000 1.000
· Girl 1.832 .162 ***

Custody arrangement at separation
Variables Living arrangements at C3
(shared custody)
With mother With father
· (Alternating weeks) 1.000 1.000
· Alternating two weeks 1.077 8.194 *
· Alternating nights .721  
· Weekdays / weekends .469 1.647
· One weekend in two 1.844 .222
· Other .606 5.067 *

Region of Canada
Variables Living arrangements at C3
(shared custody)
With mother With father
· (Canada, excluding Quebec) 1.000 1.000
· Quebec .398 * .264 **

Reference category given in parentheses.

1 Coefficients significant at: ***p<.001; **p<.01; *p<.05.

Beyond this, however, the question remains—why are shared custody arrangements so likely to transform into sole custody over time? Is it too difficult to maintain as an arrangement? Where does the impetus for change come from? From the parents, because of some external change, in the place of work or new family commitments, or from the child who finds moving between two homes unsettling or complicated? Unfortunately, this sort of question cannot be answered by the NLSCY, which provides little information on the circumstances of the family unit that does not contain the principal respondent for the child. Even the distance separating the homes of separated parents is unknown. Nonetheless, one important fact can be gleaned from the data: however shared custody may evolve in the years following separation, it appears to be associated with a higher level of children's continued involvement with both parents after separation. Even when mothers and fathers no longer shared the physical custody of their children, over 60% of the "non-resident" parents maintained at least weekly contact with their children; more than three-quarters saw them at least every two weeks, and fewer than one in twenty had not seen their child in the twelve months preceding Cycle 3. This frequency of contact is considerably higher than that found between non-resident parents and children living in sole custody from the moment of separation.


Date modified: