Developing Spousal Support Guidelines in Canada: Beginning the Discussion


This section of the paper provides some social context for the development and operation of spousal support guidelines. An important issue in developing spousal support guidelines is the question of potential impact. Data on the actual incidence of spousal support is important in establishing a benchmark against which to assess the potential impact of guidelines. Length of marriage is likely to emerge as a central structural component in any proposed Canadian guidelines, as is the presence of dependent children. A review of existing data on the nature of marriages which end in divorce may assist in structuring guidelines (for example, help us determine what is a long marriage). To the extent this data reveals what percentage of divorces will fall into different categories, it will also assist in assessing the impact of any proposed guidelines.

What follows is a review of the information available with respect to incidence of spousal support and the characteristics of marriages which end in divorce. Unfortunately, as will become apparent, the information is in many cases limited and of questionable reliability.

A. Incidence of Spousal Support

Current law appears to offer a very expansive basis for spousal support—an expansion that began with Moge in 1992 and intensified with Bracklow in 1999. Reliable data on the actual incidence of spousal support does not exist, but the scant available suggests that spousal support is present in only a small percentage of divorce cases—ranging from the low twenties at best, to the low teens at worst.

The best data comes from the federal government's 1988 Evaluation of the Divorce Act,[119] where data based on an examination of court divorce files showed that spousal support was sought in 16% of divorce files, and in only 19 percent of cases where there were dependent children. Data in the same study drawn from surveys of individuals who had gone through the divorce process, a slightly more reliable source, showed only a slightly higher incidence: 22 percent of wives reported that they had sought spousal support at some time since the separation; while 30% of the husbands reported that their ex-wives had sought spousal support.[120] One of the problems with this data is that it dates from 1988, and hence does not take into account any of the shifts in the law as a result the Moge, let alone Bracklow.

The more recent data is even more limited in scope. One source of data is provincial maintenance enforcement programs. As part of the Maintenance Enforcement Survey, which is not yet fully implemented, information has been collected on two provinces, Saskatchewan and British Columbia, for orders registered at March 31, 2000.[121] One has to extrapolate from the survey figures. Those conducting the survey estimate that most provincial maintenance enforcement programs are only dealing with 40 to 50% of support orders, with the emphasis being on those orders where there is difficulty with enforcement. As well, welfare recipients are obligated to register their orders for enforcement, thus skewing the data towards low-income recipients.

With respect to Saskatchewan, spousal support only was found in 4.1% of Divorce Act orders registered in the program and combined spousal and child support orders represented 7.4% of Divorce Act support orders (so a total of 11.5% of all support orders).[122] For British Columbia, spousal support only was found in 4.1% of Divorce Act orders registered in the program and combined spousal and child support orders represented 4.8% of Divorce Act support orders (for a total of 8.9% of all Divorce Act support orders).[123]

Another source of data, the survey of child support awards under the Divorce Act database, shows spousal support awards in approximately 13% of cases where there are orders for child support.[124] This data source is limited because it only includes cases in which child support is awarded.

All of these numbers seem low, but American data reveal the same pattern, with percentages in the high teens.[125]

What explains the low incidence of spousal support in the existing studies? One possibility, of course, is that the studies do not reflect the actual incidence of spousal support. Given that spousal support is a higher-income phenomenon, it is likely that a significant amount of spousal support is found in agreements rather than court orders, and hence does not get captured in court-based data.

The other possibility is that uncertainty about the spousal support obligation prevents claims from being made. The 1988 Divorce Act evaluation asked women why they had not asked for spousal support: 63% felt that they were self-sufficient or had other sources of support; 24% did not believe in it or wanted a clean break; another 11% wanted support but did not think support would be granted, or if granted, be paid.[126] These reasons reflect the strong influence of ideas of clean break and self-sufficiency and the limited basis for spousal support in 1988.

One might have expected to see shifts in the willingness and desire to claim spousal support with the expansion of the obligation as a result of Moge and Bracklow. However, spousal support remains highly discretionary and very uncertain. It is still not a clear entitlement. We know little of what actually goes on in the negotiation of separation agreements, but it may be that spousal support is the first claim to be "pulled off the table" or whittled down in the give and take of negotiation.[127]

If there is some accuracy to the studies showing a relatively low incidence of spousal support, the development of guidelines might have a fairly significant impact in increasing incidence and making spousal support a relatively standard part of many divorce claims, at least in middle and high income cases. This would be consistent with the existing legal framework, but might constitute a fairly dramatic shift in practice.

B. Length of Marriage

Any guidelines, however constructed, will likely have to take marital duration into account. What do we know about how long marriages last?

In 2000, the median duration of marriage, calculated from date of marriage to date of divorce was 11 years.[128] This figure has remained relatively constant for over decade. Given that most spouses will have been separated for at least one year before divorce, the median period of actual cohabitation could be less than 10 years.[129]

The majority of divorces, (60.8%) involve couples who have been married for less than 15 years. A sizable number (19%) of these are marriages that end within four years.

Only 12% of marriages that end in divorce last 25 years or more. Even if one extends the definition of long marriage to include marriages that last at least 20 years, the number only rises to 22%.[130]

What are the implications of this for structuring guidelines. Under some the guidelines reviewed in Part 1V above, income-sharing does not kick in until the relationship has lasted a minimum of 5 years. Based on Canadian data, that would exclude 19% of the cases. Long marriages are probably the easiest to deal with; depending on one's definition that would include only either 12% or 22%.

The vast majority of marriages that end in divorce are in the middle in terms of duration: 23.5% lasted 5-9 years; 18.5% lasted 10-14 years; and 13.8% lasted between 15 and 19 years. Even within this middle range, there is a much higher incidence of divorce in marriages of shorter duration, with the percentages declining after 15 years. Many of these marriages are likely to still have dependent children (see below). Whatever is done under guidelines for marriages of medium duration will have an impact on a significant proportion of the divorcing population.[131]

C. Marriages with Dependent Children

Guidelines may provide for different treatment of marriages involving dependent children. Existing studies show that approximately half of divorces involve dependent children.

Data collected under the Divorce Act registry indicates that, in 2000, 42.6% of divorces involved dependent children.[132] This number has slowly been going downward since 1991, when the figure was 53.5%. There is no data, unfortunately, on the age of the dependent children. Falling in the category of marriages without dependent children would be both marriages in which there were children, but they are no longer dependent, and childless marriages.

This data is not complete, however, and must be approached with some caution. It only includes cases where children are noted on the record of divorce. In cases where parties have, for example, reached an agreement about custody beforehand, and the court did not have to adjudicate, the children may not be entered on the record of divorce. Consequently, divorce registration data underestimates the number of cases of divorces involving dependent children.

It is possible, therefore, that well over 50% of marriages involve dependent children. Guidelines need to be developed with the recognition that such cases may well constitute the majority of spousal support cases.

Date modified: