Shared Custody Arrangements:
Pilot Interviews With Parents

2004-FCY-5E

1. INTRODUCTION

1.1 Background

Since 1990, the Department of Justice Canada has undertaken a program of research on issues relating to the support and welfare of children in families that separate or divorce, including child support, and custody and access. The Department has commissioned several projects to examine different aspects of custody and access. These have included a pilot project to collect data on different types of custody and access orders from court files and to interview a small sample of parents about their parenting arrangements (Ellis, 1995). Another project analyzed findings from the National Longitudinal Survey of Children and Youth (NLSCY) and the General Social Survey (GSS) on custody, access and child support (Juby, Marcil-Gratton & Le Bourdais, in press; Le Bourdais, Juby & Marcil-Gratton, 2000; Lin, in press; Marcil-Gratton & Le Bourdais, 1999). Most recently, the Department commissioned two studies examining parenting arrangements. The first is a critical review of the literature on custody arrangements (Moyer). The second, undertaken around the same time, is a pilot study of shared custody arrangements. The second study is the subject of this report.

1.2 Scope and Objectives

The review completed by Moyer highlighted the need for more detailed information on the characteristics and impacts of differing custody arrangements. In response, the Department of Justice Canada supported a pilot study to examine one type of custody arrangement—shared custody—and to determine the feasibility of collecting information on parenting arrangements on a national level.

The pilot had two objectives:

  1. To explore shared custody arrangements in a small sample of parents. Examination of these types of arrangements in a small sample allowed for the generation of hypotheses and directions to explore in future research.

  2. To devise a methodology for conducting a study of shared custody arrangements that could extend to other parenting arrangements and national samples. The pilot study allowed for an initial examination of the limitations and practical methodological issues to be addressed in a national study, without the cost of a large survey.

For this pilot, the department decided to focus on shared custody arrangements. One of the early points raised in the review by Moyer was the importance of clearly defining different parenting arrangements. Terms such as "shared parenting," "shared custody" and "joint custody" are all used to mean a number of different things, both in the literature and in family law practice. In this study we used the federal government’s definition of shared custody in the 1997 Child Support Guidelines, as meaning that the children reside in two residences and that they spend a minimum of forty percent of time in the second residence (Department of Justice Canada, 1997). It is important in particular to distinguish shared custody in this sense from joint legal custody, in which parents share responsibility for key decision areas in the children’s lives, but may have any of several possible living arrangements.

1.3 Concepts Explored

1.3.1 Characteristics of the Families Engaged in Shared Custody

Few studies have explored the characteristics of families who share custody of their children after separation and divorce. Further, how these arrangements actually work out in daily life is unclear. There is likely a strong association between familial characteristics and the success of shared responsibility for the day-to-day care of children. For example, the number of children in the family and the age of these children may be a factor in determining shared arrangements. The age of the parents may also play a role. Finally, there has been some suggestion by researchers that parents with shared custody are from middle or upper-middle socioeconomic backgrounds, or are better educated/have better jobs than parents with other arrangements (Moyer, 18).

There has been little attention paid to conflict between parents with respect to shared custody arrangements. There is some preliminary evidence to support the hypothesis that parents with shared custody have lower levels of conflict, and that they did so even before the divorce was final (see Moyer). This finding makes intuitive sense, as parents who have shared custody likely come into contact much more often than those with other arrangements, and therefore need to be cooperative.

The relationship between the parents influences how shared custody works with respect to the interests of the children. It is important to determine the frequency and nature of contact between the parents, how often the parents discuss parenting issues and how those kinds of discussions go, what areas of disagreement arise, if any, and how the parents are able to maintain a positive parenting relationship. Further, although only peripherally related, the distance between residences of the parents also necessarily plays a role in the success of shared parenting arrangements. It is likely that parents who maintain shared custody live in close proximity to each other. Any real distance between households would likely negatively affect parents’ ability to make frequent changeovers.

1.3.2 Characteristics of the Shared Custody Cases

The potential stability or instability in arrangements for families with a shared custody arrangement is a very important consideration, particularly at a time when Divorce Act reforms are under consideration. It is generally believed that maintaining contact with both parents has a positive impact on children following separation and divorce. However, we know very little about shared custody arrangements in reality, particularly whether or not they are stable. According to Moyer, the literature suggests that living arrangements can change over time according to changes in parents’ employment and other circumstances. Finally, findings from the National Longitudinal Study of Children and Youth (NLSCY) suggest that: 1) the court orders do not necessarily reflect the reality of custody; and 2) over time, shared custody arrangements are not stable and the incidence decreases (Marcil-Gratton & LeBordais, 1999: 27).

Shared custody implies more than just the dual residence of children. It assumes that there will be an agreement between parents about the sharing of other parenting responsibilities, about decision-making with regard to their children and about the expenses associated with raising the children. There is little information regarding how parents with shared physical custody delegate decision-making responsibilities. However it is important to determine the extent to which these areas are agreed upon by the parents, the types of agreements in place, the stability of agreements, and whether parents are satisfied with the decisions made and responsibilities assigned. Regardless of where the children live after a separation, there are divisions of responsibility to be worked out as long as both parents are still actively involved in the children’s lives. In shared custody arrangements there are often more decisions to make about the division of responsibilities because both parents are likely to be actively involved in the day-to-day lives of the children. The research, while limited in this area, suggests that individual areas of responsibility are more likely to be shared equally in shared custody arrangements than in other arrangements (Moyer, 27).

There has been some evidence to suggest that shared custody arrangements are associated with compliance with child support payments, although the causal direction of the effect is undetermined. However, Moyer notes that parental income has rarely been controlled in these studies, yet parents with shared custody are usually from higher income strata (those more likely to be compliant regardless of the arrangements made).

The costs of maintaining two households is recognized as a significant additional burden on separating parents, especially when children are involved. In shared custody arrangements these new costs can be especially significant since there is a need, at least to some extent, to duplicate the home environment for the children. Shared custody is found in the research to cost more than sole custody, although one Australian study found that some of the additional costs, such as having a bedroom for the children, are also often borne by non-custodial parents in sole-custody arrangements (see Moyer). However, the actual costs of shared custody, in itself or in comparison to other custody arrangements, have not been well established.

1.3.3 Parental Satisfaction with Shared Custody Arrangements

It is reasonable to expect that how parents feel about their arrangements not only affects their success in maintaining shared custody, but could also affect the adjustment of their children. One might expect that parental satisfaction with the arrangements could affect longer-term stability.

Parents with shared custody may have fewer areas of dispute than other cases. There is preliminary evidence to support higher, lower, and no differences in rates by the type of custody arrangement (see Moyer). Determining whether shared custody arrangements lead to different patterns of court use and relitigation may have implications for the use or encouragement of such arrangements in the future.

1.3.4 Children’s Adjustment and Outcome

In Canada, the primary consideration for judges in deciding on child custody for separating and divorcing parents is what will be in the best interests of the children. The courts also review parental agreements, where those agreements come before the court, to ensure that the best interests of the children are being met. Much of the focus of federal family law policy is also on the impacts of separation and divorce on the welfare of children. Thus, in examining how shared custody arrangements are put into practice, an important interest is how the arrangements affect the children involved.

Research on divorce and separation focuses to a large extent on the relationship between parents, and their ability to cooperate in matters relating to parenting, as having a profound effect on how children cope with the break up of the family. Moyer states that different types of custody arrangements are not necessarily associated with different outcomes for children (p. 34). It appears that variables relating to the relationship between the parents (e.g. level of conflict) and characteristics of children (e.g. adaptability) have a stronger effect on outcome than the arrangements for children’s residence. Most of the existing literature on outcomes for children indicates that custody arrangements, in themselves, do not have a determining affect on children’s well-being. However, research looking at aspects of custody arrangements, such as cooperation between the parents and parental relationships, has found some outcome differences by custody arrangement.

1.4 Pilot Study Methodology

A sample of 199 cases was drawn from an initial list of divorces registered in Edmonton and Calgary, Alberta from fall 1998 through spring 2000.[1] The divorce order for all cases included a shared custody arrangement. Files for the selected cases were available to the research team at the two court locations. The files were reviewed to ensure that they were, in fact, shared custody cases, and contact information was drawn from the files, including the names of the two divorcing parents and any information as to their place of residence. No other information originated from the court files.

Where no address or phone number information was available, there was no attempt to contact the respondent (see Figure 2.1 for flow chart of recruitment). For 80 files, this was the case. In 84 cases, one of the parents had no address or phone number available, so no contact was made. For 35 cases (70 parents), contact information was on file for both parents. Of those cases, 50 people were interviewed, 3 declined to be interviewed, and 17 were contacted but were unavailable.

Determination of a response rate for this study is not clear. There are several levels to consider (see Figure 2.1):

  1. The number of families for whom at least one address was available: 59.8 percent (119/199).
  2. The number of families for whom two addresses were available (one for each parent): 17.6 percent (35/199). This requirement put a significant limitation on the sample, taking potential cases from 119 to 35.[2]
  3. The number of parents for whom contact information was available for both parents, and who agreed to participate: 71.4 percent (50/70). This is a conservative number because only three parents actually refused to participate.

Figure 2.1 Flow Chart of Participant Selection


Flow Chart of Participant Selection

[ Description ]

Information from each of these levels has an impact on determination of samples for future research (for discussion, see the feasibility section).

There were fifty interviews conducted, half with mothers and half with fathers. Thirty-eight of the fifty interviews were with "matched parents."[3] The remaining 12 interviews were with parents for whom the other partner was unavailable. The inclusion of "matched parents" was designed to enable the researchers to obtain information about the expenses of maintaining two households in shared custody cases, and to be able to compare responses of the two parents on some issues.

Due to the decision to interview as many matched parents as possible, there were essentially three levels of analysis for the participants in the current study:

  1. Total sample (N=50 individuals);
  2. Nineteen pairs of "matched parents" (mother and father from a former union; N=38 individuals); and
  3. Thirty-one parents (one mother or father from each family).

Analyses from the total sample were mostly those describing participants. Analyses from the "matched parents" were those in which information from both members of a former union was required: for example, the expenditures for maintaining two households in a shared parenting arrangement. The "matched parent" responses were of interest because they offered insight into varying perspectives between the two parents. Finally, the sample of parents was used for most analyses. This smaller sample was created to equalize the weighting of former unions.[4]

The decision to collect information through interviews with parents was made for several reasons, including the ability to obtain: 1) more detailed and accurate information than from file reviews, including parental perceptions and satisfaction with arrangements; 2) details on changes in arrangements; and 3) indication of the type of information that should be incorporated into an instrument for further research.

The interview guide was a collaborative effort between the Department of Justice Canada research officers assigned to the project, the primary researcher from the firm Alderson-Gill & Associates, and a second independent consultant with knowledge in the area of child custody and access. The issues to be covered, the analytic limitations of a small sample, the sensitivity of the subject matter, and the duration of the interviews were all considered in the development of this study. Based on the literature review by Moyer, the interview guide had eleven sections corresponding to areas of inquiry of interest to the research. These included:

  • Family history;
  • Arrangements at the time of separation;
  • Arrangements at the time of the divorce;
  • Arrangements as they are currently (at the time of the interview);
  • Parental responsibilities;
  • Expenses associated with shared custody;
  • Relationship between the parents;
  • Outcomes for the children;
  • Parental satisfaction with the shared custody arrangement;
  • Demographic information; and,
  • Inquiry about possible future interviews with children.

For the pilot study, the interview schedule included a total of 75 questions and 38 follow-up questions. All but one of the initial questions was closed-ended, with a pre-set list of responses, and, in most cases, the possibility of specifying an "other" response. Of the 38 follow-up questions, 30 were open-ended, typically asking respondents to elaborate on their initial response (see Appendix A for the Interview Guide).

We sought basic information about the living arrangements for the children at the time of the parents’ initial separation, at the time of the divorce, and at the time of the interview (which was usually two to three years after the divorce). Thus, although the pilot was based on a single set of interviews at one point in time, we built a retrospective look at how custody arrangements changed in the three time periods into the interview guide. These questions did not have the power of longitudinal assessment and relied on parents’ ability to recall previous periods of time. However, the questions were not of a nature to require detailed memory of events. For example, they required that parents remember whether there was a shared custody arrangement or some other living arrangement. The questions also asked respondents to recall other aspects of the arrangement, such as the process for making child-rearing decisions and the division of parenting responsibilities.

A standard letter from the Department of Justice Canada went out to all parents for whom an address was available. The letter provided a brief description of the purpose of the research and expressed the hope that they would be interested in participating through a telephone interview. The letter also gave the selected parents an option to contact the Department of Justice Canada at a 1‑800 number, or the primary researcher at his office phone number, to obtain further information or to decline to participate. Those who did not decline were later telephoned, asked to participate, and a time was arranged for the interview.[5] Because of this "negative option" approach, no follow-up letter was issued to prospective respondents.

Interviews were conducted by telephone. A female researcher with interviewing experience on family law policy issues conducted the majority of the interviews. The primary researcher performed the remaining interviews as a way to test the flow of the questions and make adjustments to the guide. He continued to conduct interviews on an ongoing basis to better equip him to interpret the findings. After five interviews, a small number of minor adjustments were made to the interview guide to improve the wording of a few questions, but the nature of the questions did not change.

The interviews lasted between 45 minutes and two hours, with an average length of about one hour. The great majority of respondents agreed readily to participate, showed an interest in the issues being raised and offered thoughtful responses. With very few exceptions, respondents agreed to answer all the questions they were asked, including information about their employment, income and education. One respondent declined to provide income information, and one other respondent declined to provide estimates of expenditures associated with the shared custody arrangement. For a very small number of questions, respondents said they did not know the answer. Otherwise, response rates for individual questions were 100 percent.

1.5 Structure of the Report

The remainder of this report is organized into three sections. Section 2 provides and discusses the results of the pilot survey interviews. Section 3 examines the methodology of the study and discusses various considerations for conducting a larger study on parenting arrangements in Canada. Section 4 contains a short wrap-up of the findings and a feasibility portion. There are several appendices included in this report:

  • Appendix A: Shared Custody Parent Interview Guide
  • Appendix B: Reported Areas of Disagreement Between the Parents, and Reported Reasons For Ability to Avoid Disagreements
  • Appendix C: Reasons For Parent Satisfaction or Dissatisfaction With Living Arrangements
  • Appendix D: Reasons Reported For Children Being Happy or Unhappy With Living Arrangements
  • Appendix E: Reported Changes in Children’s Behaviour, and Parent Attribution