Introduction

There has been some discussion over the last several years about whether there is likely to be any change in the very low levels of payment and receipt of spousal support that have been found to exist in Australia (Parker, Parkinson & Behrens 1994, p. 546; Finlay, Bailey-Harris & Otlowski 1997, pp. 192-6). Both empirical and ideological factors tend to indicate that there will be a continuing decline in spousal support. Empirical factors include the extent of child support obligations reducing the capacity to pay spousal support in some cases, and women's increasing participation in the paid workforce reducing need. As regards ideology, the growing trend towards individualism works against situations where one adult is required to support another, and there may also be an associated increase in the desire of payees to achieve a 'clean break' from their spouse.

However, there are factors that point the other way. These include the increasing emphasis in government policy on private over public support (of which the child support scheme is evidence), increasing understanding about the continuing economic disadvantage that women suffer compared with men in the paid workforce (Funder 1993) and on marriage breakdown (Weston 1986, 1993), and a greater sense of the appropriateness of sharing the economic costs within marriage in the post-separation financial and property arrangements (Harrison, Funder & McDonald 1993).

To some extent at least, recent messages from the Full Court of the Family Court have emphasised the latter factors (we refer to these cases below), placing renewed emphasis on the use of spousal support as a means of compensating for the economic disadvantage women suffer on marriage breakdown. Such messages are likely to have varying degrees of impact on the actual practice of family law. Trial judges and legal practitioners may not notice them, or may disagree with them and ignore them; individuals privately negotiating arrangements are unlikely to be influenced by such messages to any great degree, at least not for a while; other factors may run counter to them (for example, the lack of availability of legal aid, or the desire to achieve a clean break).

Intermingled with the descriptive and predictive questions (What is happening? What is likely to happen?) are normative questions (What should happen?), and there is a continuing lively Australian and overseas debate about whether and in what circumstances there should be entitlements to spousal support (Bala 1994; Davies 1994; Deech 1977; Diduck & Orton 1994; Edgar 1996; Land 1984; Morgan 1996; O'Donovan 1982; Payne 1995; Sheppard 1995). The Family Law Council has recently established a Clean Break Sub-Committee to inquire into and report on a range of issues including the role of spousal support; it therefore seems likely that this debate will continue over the next few years.

The project

The data presented in this working paper are drawn from the Australian Divorce Transition Project, a random national telephone survey of 650 divorced Australians.⁽¹⁾ This survey, conducted in late 1997 by the Australian Institute of Family Studies, examined the divorce transition and its consequences for parents, and investigated the impact of divorce on an older cohort for whom the issues of employment, housing, superannuation and future security were important concerns. As part of this project, specific data were collected on spousal support. These data sought to measure the incidence, duration and amount of spousal support payments or receipts, and also to ascertain respondents' attitudes about whether, and for how long, spousal support should be payable.

Some definitional issues

We note here that the term 'spousal support' was used in the survey and is therefore used throughout this paper. The term used in the Family Law Act 1975 (Cwlth) is 'spousal maintenance'. Other terms are in fairly common use, including (particularly in the United States) 'alimony'. The designers of the interview instrument were of the view that the term 'spousal support' is more generally recognisable in the community, particularly as maintenance for children is now called 'child support', and this latter term is in general currency. The interview questions made clear that respondents were being asked to distinguish between child support and spousal support by instructing that 'spousal support is the money paid by a former spouse for the upkeep of the other spouse - not for the children'.

Also, spousal support is commonly thought of as payments of money on a regular basis over a period of time. It is often contrasted with property adjustment, in which a lump sum or some interest in property is transferred. In reality, the situation is rather more complicated than this: both forms of order can be given effect through periodic amounts, lump sums or property transfers (or a combination of these) and it is sometimes difficult to distinguish between spousal support and property division. While this leads to some methodological problems (discussed shortly), one of the unique aspects of this study is that it attempts to measure both periodic payments and other forms of payment of spousal support.

We also note that some State and Territory legislation enables a de facto partner or member of a 'domestic relationship' to apply for support.⁽²⁾ For some people at least, the policy issues around entitlements to support pan out quite differently depending on whether or not the claimant is or has been married to the payer. These issues have been discussed in law reform papers and other places (New South Wales Law Reform Commission 1983; Queensland Law Reform Commission 1992). The focus of the Australian Divorce Transition Project was on those who had experienced divorce, and therefore it was only claims that were brought under the Family Law Act, or settled in its shadow, that have been recorded.

Finally, we sometimes use gendered language in this paper. While spousal support is available under the terms of the legislation to both men and women (in contrast to the position under earlier law), all the evidence suggests that men are the vast majority of payers and women the vast majority of payees. These data support this evidence. As indicated in the final part of the paper, it is important not to lose sight of this in discussing possible future directions.

The legislative framework

The provisions in the Family Law Act that apply to the division of property and to spousal support are contained together in part VIII of the Act. In determining how property is to be divided between the parties, account must be taken first of the contributions of the parties (s. 79(4)), and second of the future needs and financial resources of the parties. So, for example, while the contributions of the parties may (depending on the circumstances) be regarded as roughly equal, a spouse who has been financially dependent on the other may get an extra percentage (perhaps 10 to 20 per cent) of the property to allow for her greater needs.

The factors that a court must take into account in determining spousal support claims (contained in s. 75(2) of the Act - in Appendix A) are the same factors that are required to be taken into account in this second stage (the future needs and financial resources stage) of determining the division of property. In the process of negotiating financial and property arrangements it will often not be clear whether additional property has been given to one party as part of the second stage of the property adjustment process, or as lump sum spousal support. Where a property transfer or lump sum order or registered agreement is made for spousal maintenance purposes, this should be specified in the order or agreement.⁽³⁾ However, this may not always occur (for example, one of the parties may have a strategic advantage in specification occurring or not occurring). Further, there is no such requirement with informal agreements. In any case, respondents might have regarded as spousal support an amount that was not specified because it was for future needs. Difficulties in determining the purpose of the transfer lead to some methodological problems (discussed below).

In order for an applicant's claim for spousal support to be successful, the threshold test in the legislation must first be satisfied. The threshold test is contained in s. 72 and requires both that applicants are unable to support themselves adequately for reasons that the legislation provides are relevant, and that the respondents are reasonably able to support them. Section 72 specifically mentions inability to support oneself because of care of children, and because of age or incapacity for appropriate employment, although these are not the only circumstances in which the threshold test can be satisfied; there is a requirement to have regard to all the s. 75 factors in determining whether the two limbs of the threshold test have been satisfied.

The question of whether particular applicants can be said to satisfy the threshold test for spousal support has received some attention in recent case law. In particular, in Bevan (1995) FLC 92-600, the Full Court stated that an applicant was not expected to use up a 'meagre capital sum' (in this case $155,000) in self-support (although the interest which she could earn from its investment was taken into account in assessing her needs). In Mitchell (1995) FLC 92-601, the Full Court found that the trial judge had erred in finding that the applicant could not satisfy the threshold test because she had not taken sufficient steps to obtain full-time employment. The Court took account of the difficulties faced by people who have a discontinuous work history and of material on the feminisation of poverty.

So, under what legal conditions is a claimant likely to get spousal support? Of course, spousal support can be paid by a spouse in any circumstances if they want to do so. However, a court would only order spousal support to be paid if the claimant is in need and the other spouse is able to pay. What is need and what is capacity to pay will depend on the circumstances of the case; for example, a claim is more likely to be successful where the claimant has limited capacity for work (particularly where that capacity has been limited by the circumstances of the marriage, such as by a gender based division of labour or the need to care for children), and where the spouse from whom support is sought has significant assets, income, or both. While matrimonial fault is not, of itself, to be taken into account under the legislative provisions,⁽⁴⁾ other 'justice' factors, such as support provided by the claimant to the other spouse while they obtained qualifications, are relevant. Further, it is possible that notions of fault may impact on decisions about spousal support in implicit ways. (We explore this further below.)

An additional point to make about the legislative framework is that where spousal support orders and registered agreements are made, amounts payable under them are able to be collected by the Child Support Agency under the Child Support (Registration and Collection) Act 1989. Thus, payments of spousal support can occur without any personal contact between the payer and the recipient. This may be especially important where spousal violence has been reported.

Why ask questions about spousal support?

The need to gather empirical data on the extent of payments and receipts of spousal support is fairly evident - quite simply, there is a need to get a better idea of what the law is on this issue. It is impossible to know this from the legislation and the reported case law. The legislation conveys a broad discretion on the Court to determine whether and to what extent to order spousal support. The reported case law often consists of exceptional, rather than everyday, factual situations or judicial decisions (Parker & Drahos 1990).

As a related point, attitudes often affect behaviour and are central to the interaction between public acceptance, compliance and the effectiveness of law. There is also intrinsic and practical worth in knowing what the divorced population thinks about particular areas of family law. This knowledge is important 'for public education, for appropriate sensitivity in the professions associated with family law, and for assuring provision and access to appropriate services (in the Court and elsewhere) for groups with demographic and value differences' (Funder & Smyth 1996, p. 6). Thus there is much utility in examining attitudes towards spousal support.

Structure of the paper

In Part I of this paper we report the basic incidence data and provide some relevant contextual discussion. Part II contains a cross-case analysis of the small numbers of cases involving payment or receipt of periodic spousal support in an attempt to profile the characteristics associated with these cash transfers. Part III of the paper reports the basic data on attitudes and then looks at the relationship between attitudes and behaviour. In Part IV we report on the results of an analysis of the data which identifies some possible correlates of attitudes to spousal support. Finally, in Part V we overview the paper and identify and discuss some policy and law reform implications of the study and some issues that need further exploration.

Method

The Divorce Transition Project: sample selection

Two independent samples (stratified by gender and geographical location) were drawn from the population of Australian households with telephones. Households from Western Australia, however, were not sampled due to some differences between the law on child-related issues in that State and the rest of Australia.

The first sample comprised 513 parents (284 women, 229 men) who:

• were divorced under Australian law;
• were separated after January 1988; and
• had a child under 18 years of age at the time of separation.

The second sample comprised 137 older women and men from long-term marriages (77 women, 60 men). The respondents in this sample were:

• divorced under Australian law;
• separated after January 1988;
• married for 15 or more years;
• from marriages where the age of the wife at separation was between 45 and 65 years.

Though the status of respondents as parents was not a specific criterion for inclusion in the long-term marriage sample, it is noteworthy that half the respondents in this sample had at least one child under the age of 18 years at the time of separation.

Respondents for both surveys were randomly selected from the Australian 1997 Telstra Electronic White Pages. This database lists Australian residential telephone numbers.

To obtain the target samples, a total of 72,308 telephone calls were made. Of these calls, 28 per cent (20,303) of households were not contactable (primarily because there was no reply or a disconnected or new business number). Of the 52,005 households that were contactable, 28 per cent of people refused to participate at the outset, and 70 per cent did not contain a person who met the screening criteria.

Of the 881 households that contained a person who fitted the sampling criteria, 74 per cent of respondents completed the interview (26 per cent refused to participate). However, since most refusals occurred before interviewers could determine the eligibility of respondents for the study, no accurate response rate could be calculated. When comparable base rates are used, these 'call results' nonetheless accord with a similar sample interviewed by telephone reported by Funder and Smyth (1996).

The Divorce Transition Project: demographic profile

Both samples were combined for the present study. The total sample thus comprised 650 divorced respondents: 361 women and 289 men (56 per cent versus 44 per cent) ranging in age from 22 to 76 years (mean (M) = 44 years, 2 months; standard deviation (SD) = 8 years, 11 months). Table 1 and Table 2 present the demographic profile of female and male respondents.

Notes. * Major metropolitan cities were defined in line with Australian Bureau of Statistics Statistical Divisions; 
^# includes those looking for work.

Table 1 shows that about 80 per cent of respondents were from the parent sample. As a result of the stratification of the sample for geographical location, 63 per cent of respondents lived in urban centres, and around two-thirds lived in either Victoria or New South Wales. Most of the others lived in either Queensland or South Australia.

Women and men were of a similar age (M=43 versus 46 years), had been married for an average of 14 years, and had been separated for 6 years on average (see Table 2). Both women and men were most likely to indicate that they had no post-secondary school qualifications (58 per cent versus 46 per cent). Women were less likely than men to report that they had received a diploma or vocational training (18 per cent versus 34 per cent), were less likely to be in paid work (68 per cent versus 78 per cent), and marginally more likely to indicate that they were tertiary graduates (24 per cent versus 20 per cent). 

Women were also more likely than men to rely on social security as their main source of income (33 per cent versus 12 per cent), to have significantly lower personal and household incomes (for example, median personal annual gross income $20,000 versus $30,000), and to be single (71 per cent versus 58 per cent). These differences are consistent with prior research conducted by the Australian Institute of Family Studies (for example, Weston 1986, 1993).

The Divorce Transition Project: survey content

The survey for the project contained 12 broad sections: (1) screener questionnaire; (2) marital history; (3) household composition; (4) parenting arrangements (including residence, contact and child support); (5) property division; (6) spousal support; (7) education, training and work history; (8) income; (9) housing; (10) personal wellbeing; (11) personal relationships; and (12) demographic information.

Text for the spousal support questions is provided in Appendix B. In brief, after an explanation of what spousal support is, respondents were asked three questions designed to ascertain whether they had ever paid or received regular spousal support (question (Q) 294), whether they were payers or receivers of that support (Q295), and whether they were still receiving or paying such support (Q296). They were then asked questions to measure the length of time they received or paid support (Q297A, Q297B) and the amount that they paid or received (Q298, Q299). They were then asked whether they received a larger share of the assets as part of spousal support (Q300). Finally they were asked the two normative questions: 'Do you think spousal support should ever be paid?' (Q302), and (if they answered affirmatively) 'For how long should spousal support be paid?' (Q303). In relation to this last question they were given options: for a limited period of time until the partner is on his or her 'feet'; for a limited period of time until a person repartners; indefinitely; other.

Methodological problems and conceptual issues

It is important to note a number of methodological problems that are a feature of studies of divorce based on retrospective self-reports. A major general problem relates to the accuracy of the data that depends upon the respondents' recall of information about events that might have occurred up to ten years before, and at a time of great stress and general difficulty. The questions on financial matters are likely to be the most difficult to answer accurately, as they require recall of detail and might have been matters of subsidiary importance to those that had to be made about, for example, children.

In addition, there is a particular danger with the spousal support data that respondents will either not have known, or will not be able to recall, whether they received spousal support or not, and particularly whether or not lump sum cash payment or property transfer was received as spousal support. This difficulty in knowing and recalling is compounded by the complex linkage between the property division and spousal support provisions of the Family Law Act, described above.

Furthermore, respondents might have had difficulty in drawing a distinction between child support and spousal support. To highlight this distinction, respondents were given the direction that 'spousal support is the money paid by a former spouse for the upkeep of the other spouse - not for the children'. This direction should have minimised the extent of error of interpretation, if not recall - though this distinction may be hard to make conceptually.

Finally, respondents might have taken a broader view of 'spousal support' than just the payment of money or transfer of property (for example, the provision of access to a car).

Three more general methodological limitations should also be noted. First, while the survey design can detect associations between spousal support and other factors, it cannot determine the direction of these associations - that is, causality. Thus no claim can be made that certain factors lead to the payment or receipt of spousal support or vice versa. Second, the unit of analysis was a spouse - not a couple - from a dissolved marital union; moreover scant information was obtained about respondents' former spouses. Thus there are limits to checking the accuracy of the data or understanding the cause of any mismatch that might have occurred. Likewise, the capacity to pay of female respondents' former husbands cannot be determined, nor can the need of male respondents' former wives. Finally, since not everyone is accessible by telephone, the omission of certain groups of people in the population available through telephone surveys sets limits on the generalisations that can be made from the data to the Australian population at large. Among those who are often systematically excluded in such surveys are the very poor, those with unlisted numbers, those who live in geographically remote areas, and those who have hearing or English language difficulties.

Nonetheless, few approaches enable nationally representative estimates to be obtained of both court-registered and privately arranged spousal support; court-registered agreements can be scrutinised from court files but knowing about agreements that are not registered depends upon self-reporting by those who enter into them.

Part I: Incidence, duration, amount: the basic data

Here we report on the basic data about incidence, duration and amount of spousal support, and explore change across time. Most tables are split by gender of respondent because gender is an important variable in its own right.

Incidence

In Australia, spousal support has been found to be rare, minimal, and brief. Population surveys commonly yield low rates of occurrence of between 2 and 8 per cent (although with higher rates in the immediate post-separation period) (Harrison 1993, p. 117; Harrison & Tucker 1986, p. 261). These estimates, however, typically refer to periodic payments and tend to exclude other indirect forms of payment - most notably property transfers. As indicated in the introduction, spousal support obligations can be fulfilled in these ways as well as, or in addition to, periodic payments. The estimates provided in Table 3 attempt to take account of these indirect transfers.

Notes. # N = 650; missing cases = 6 (1 respondent "couldn't say" if he had ever paid/received any spousal support; 
2 respondents who reported paying/receiving such support "couldn't say" if this was a payment or a receipt); and 3 cases of 'lump sum cash payment only ' ($5,000, $7,000 & $7,800) which were reported as periodic support were excluded from this analysis for conceptual clarity; 
* 53 respondents who had not yet finalised their property were not asked about property transfers; this did not affect the total percentages when these were disaggregated by property settlement status (settled vs. not settled).

Table 3 presents the number and percentage of divorced respondents who reported paying or receiving some form of spousal support, disaggregated by the type of support reported (periodic spousal support and property transfers). It indicates that 6 per cent of respondents had paid or received periodic spousal support since separation and 1 per cent reported paying or receiving both periodic support and a property transfer as support. ⁽⁵⁾ This estimate of those who received or paid some periodic support (7 per cent) is consistent with other incidence studies. An additional 10 per cent reported paying or receiving spousal support solely through a larger share of the assets at property division. Thus, when property transfers are factored in, around 16 per cent of the sample reported having paid or received some form of spousal support since separation. Expressed another way, about two-thirds of those who reported paying or receiving some form of spousal support did so through property adjustment, while about one-third of spousal support was reported to occur in the form of periodic support. ⁽⁶⁾ (It should be borne in mind that these incidence rates are based on respondents who have been separated across a period of ten years.)

Table 3 also shows that although the majority of respondents had never paid or received any form of spousal support, men were more than twice as likely as women to report that property transfer had occurred as part of spousal support (14 per cent versus 6 per cent). All things being equal, estimates given by women and men should be very similar. However, women and men often report divergent contributions and estimates across a range of domains (for example, housework contributions, control of finances) (Funder 1986), with men often overestimating the proportion of joint assets that their former spouses received, for example, and women underestimating their own share (see, for example, McDonald 1986).

It should be noted that, for clarity, Table 3 does not distinguish a small number of respondents (n=53) who had not yet finalised their property from those who had. When this distinction is made, however, an interesting pattern emerges: respondents who had not yet finalised their property were more than twice as likely as respondents who had settled their property to receive or pay periodic support (5 per cent versus 13 per cent). This finding suggests that, in some cases, periodic support might be acting as a kind of stop-gap 'bridging finance' to tide women over until property is settled.⁽⁷⁾

Has the incidence of spousal support changed over time?

One of the reasons for asking this question is that, as indicated in the introduction, the Full Court of the Family Court has, over the last several years, sent out some messages in its reported decisions which suggest that spousal support should be used more often - and particularly where there is little existing property to divide, and/or where the woman has been economically disadvantaged as a result of the sexual division of labour in the relationship.

The impetus for this change in some judicial attitudes seems to have been the Canadian case of Moge v Moge (1992) 43 R.F.L. (3d) 345, in which the Supreme Court of Canada used spousal support to provide compensation in these circumstances. In addition, the Full Court is increasingly taking judicial notice of certain 'social facts', and the empirical evidence about the economic disadvantages women compared with men suffer on marriage breakdown is one of those social facts sometimes referred to in its recent decisions.

The messages started in Australia with the case of Best (1993) FLC 92-418. The central issue on appeal in this case was the appropriate division of property in a context where the husband was a very high income earner, but there was little property to divide. The Full Court commented:

We should add one further factor to this discussion. In the wife's original application she sought lump sum support as an alternative or in addition to a property order under s. 79. That aspect did not seem to have been pursued in any significant way at the trial. Nor was it pursued with any vigour on this appeal.

In cases such as the present where there are minimal assets, but on the one side significant needs and on the other a significant future earning capacity, the power to order lump sum support, which may be met by annual payments over a period of years against that income or savings from it, may be an appropriate course. In such cases, and provided that the requirements of the Act are otherwise satisfied, it may be a mistake to conclude that where there are few assets they should be divided and that that is the end of the matter other than for periodic support.

Having regard to the manner in which the case has been conducted throughout on this issue it is unnecessary for us to consider this further. But it should be recognised that the 'clean break' concept may have been taken to extremes in the past and requires careful reconsideration in the light of changing economic and social circumstances and values and the benefit of experience over the past decade or so [our emphasis]: see, for example, the discussion in Moge's case.⁽⁸⁾

In subsequent cases⁽⁹⁾, various judges of the Court reiterated these messages. In particular, in Mitchell (1995) FLC 92-601 the Full Court, in finding that the trial judge was in error in refusing to grant spousal support on the basis of the wife's failure to take sufficient steps to find full-time employment, referred to the decision in Moge and stated that:

Importantly, and particularly in more recent times, there is the notorious circumstance that there is a significant gap between theory and reality for employment, especially for people in middle age, lacking experience and confidence, and who have been out of the skilled workforce for many years, and in the context of current high employment.⁽¹⁰⁾

And later:

Like Canada, Australia has a body of research indicating that mothers who are the primary carers of dependent children inevitably drop out of the paid workforce and consequently suffer financial deprivation which is exacerbated by marriage breakdown...In our view there are significant advantages to the Court being able to take judicial notice of research concerning the economic consequence of marriage and its dissolution.⁽¹¹⁾

The Full Court ordered a retrial of the spousal support issue in a context where the wife had received about $250,000 in property, was working part-time and the husband (a barrister) had significant capacity to pay support.

It was important to ask, therefore, whether these messages have had an impact on payment or receipt of spousal support between 1988 and 1995 (the last year for which data were available for the study). To explore this possibility, spousal support transfers where property had been finalised were compared by year of separation, as shown in Table 4.

Notes. # N=650: Missing cases = 91 (1 respondent "couldn't say" if he had ever paid/received any spousal support; 
2 respondents who reported paying/receiving such support "couldn't say" if this was a payment or a receipt; 
3 cases of 'lump sum cash payment only were excluded, as were 15 respondents who separated in 1996; 
8 cases in which a property adjustment was also reported as a form of spousal support in tandem with periodic support were also excluded; 
14 respondents did not report the year in which they separated); 
48 respondents who had not finalised their property were also excluded; p>.05.

Table 4 shows that there was no significant relationship between the year of separation and reported payment of spousal support. The rate of reported occurrence for periodic support appeared relatively stable across time (between 4 and 6 per cent). While there appeared to be a slight drop in reported property transfers as spousal support (from 13 to 8 per cent), this variation might have occurred by chance and might have been compounded by sampling issues.⁽¹²⁾

The lack of change through to 1995 is not really surprising. The significant decisions referred to above were reported in 1993 and 1995, and it is likely to take a while for messages of the sort sent out to filter down and affect solicitors' practices. Presumably it will take even longer for those messages to impact on individuals who are negotiating arrangements without legal advice. Further, as indicated in the introduction, other factors may counteract those messages; for example, if a legal culture has developed where spousal support is rarely sought, it may take legal action to obtain it. This is less possible in a context where legal aid funding for family law matters is declining.

Gender and periodic payments

So far we have been considering the incidence of spousal support, whether received as periodic payments or by other means. We now focus on periodic payments only.

To examine the gender distribution of periodic spousal support, the number and percentage of divorced male and female respondents who had ever received or paid regular financial support was calculated, as shown in Table 5.

Notes. # N = 650; missing cases = 3 (1 respondent "couldn't say" if he had ever paid/received any spousal support; 
2 respondents who reported paying/receiving such support "couldn't say" if this was a payment or a receipt); 
includes 8 cases where both periodic support and property transfer was reported and treats 3 cases of 'lump sum cash payment only' as 'neither payment nor receipt'); 
* <1%.

Table 5 shows that 7 per cent of males reported paying spousal support and 6 per cent of females reported receiving it.⁽¹³⁾

Significantly, only one female reported paying spousal support, and only one male reported receiving it. This is what would be expected. While spousal support is available under the terms of the legislation to both men and women (in contrast to the position under earlier law), all the evidence suggests that men are the vast majority of payers and women the vast majority of payees. (These data support this evidence.) This is explained at least in part by the continuing sexual division of labour within relationships and by the economic disadvantages women suffer in the workforce. While it is certainly true that women are increasingly participating in the paid workforce, there is no doubt that work patterns for women are different from those for men, and that discontinuous workforce participation characterises women's experience (although economic and labour market changes have made life-long employment with one employer for men far less common than it used to be). Women are often employed part-time and/or casually. Further, women tend to be concentrated in lower paid occupations, and women in careers tend to have their progress affected by family responsibilities (ABS 1992; Funder & Harrison 1993; Harrison, Funder & McDonald 1993; Scott 1984).

While these factors are in part a product of historical and labour market factors, they also reflect the fact that while women are increasingly assuming the burden of breadwinning, there is still a strong division of labour along gender lines within the home (Bittman 1991).

The gender difference so far as payment and receipt is concerned also contributes to the gender politics on questions about whether spousal support should be payable (discussed below).

Duration and amount

Table 6 summarises the average amount paid or received, and the average duration of periodic payments reported by these respondents.

Before interpreting Table 6, note the small number of respondents in the sample for whom duration and quantum data were available (n=38 and n=28 respectively).⁽¹⁴⁾ Caution should thus be exercised in attempting to generalise these results to the broader Australian divorced population who have paid or received spousal support.

Note: missing cases: duration = 6, amount = 16 (including one periodic amount of $52 p.a. which was deemed an outlier and removed); 
note that the values for the male receiver and female payer were included because each had a similar socio-demographic profile with the other payers or receivers; 
the tables also includes 8 cases where both periodic support and property transfer was reported - these periodic amounts did not differ significantly from those in which no property transfer was reported (p>.05).

Table 6 shows that spousal support was reported to last for about two years, and had a median value of $6,640 per annum ($128 per week). Divorced females reported receiving about $10,400 per annum (median), whereas divorced males reported paying half this amount (median = $ 5,200 per annum).⁽¹⁵⁾ There was, nonetheless, considerable variability among individual payers and receivers (evidenced by the relatively large standard deviation about the mean values). In the extreme, one respondent received periodic support for three months while another received it for seven and a half years. Another respondent received about $520 in periodic support each year, while another reported receiving $52,000 each year.

Part II: Profiling payers and receivers

It is of both intrinsic and immediate practical interest to know who pays and who receives periodic spousal support. If the incidence of support were not so rare, this kind of profiling exercise would best be served through statistical modelling. However, given the low occurrence of this form of support and the small sample, such modelling is inappropriate and requires a different analytic tack along the lines of descriptive cross-case analysis.⁽¹⁶⁾ Thus the approach we have taken with the incidence data is to examine a list of relevant variables for both those who are current payers or receivers of spousal support and those who were ever payers or receivers of spousal support, and to report any striking patterns among these clusters. No claim is made that these patterns generalise to, or reflect, the Australian divorced population who have ever paid or received spousal support. Nonetheless, this approach may provide some clues about who pays and who receives support.

The variables selected for the analysis were ones that we speculated might have an impact on payment/receipt rates either because they are indicators of relevant factors under the legislation (for example, capacity to pay, opportunity costs of marriage and future needs) or because they may otherwise enhance the perceived 'merit' of a claim (for example, matrimonial 'fault'). There are other factors that would have been worth including, but about which we had no data or poor quality data. (See the discussion of property outcomes and asset levels below.)

Since spousal support is typically transient, any distinction between those who are currently paying or receiving spousal support and those who have ever paid or received such support is artificial: with time, the current payers will become 'evers'. This division nonetheless provides a degree of conceptual clarity because of the contemporaneous nature of the data for current payers and receivers.

In examining the profile of previous payers and receivers, the picture becomes ambiguous because the temporal nature of the data restricts what can be said. For instance, we only have data about current income and not about income when spousal support was being paid or received. Ideally, information about both spouses should be collected before, during and after the period in which spousal support was occurring. This was not done because of the broad-brush approach adopted in the Australian Divorce Transition Project.

For this reason we report here the detail of the profiling exercise for current payers and receivers and only summarise the results of the exercise for those who have ever received or paid spousal support.

Of the 650 divorced respondents, only 32 reported ever receiving (n=13) or paying (n=17) periodic spousal support, while another 14 reported that they were still receiving (n=6) or paying (n=8) support. The following profiles emerged for those who were current receivers or payers of periodic support.

Current receivers of periodic support

All six respondents who were still receiving spousal support were female sole parents (four of whom had three children).⁽¹⁷⁾

• Five respondents were also receiving child support from the person from whom they were receiving spousal support. (The other respondent did not have a child under 18 years at the time of separation but did count as a 'sole parent' because age of the child is irrelevant to that.)
• All six respondents had been in relatively long-term marriages to the payer: three had been married for more than 20 years, and three had been married between 9 and 15 years.
• The four respondents who had finalised the division of their assets had done so in formal ways. (Two respondents had their case adjudicated by a judge and two others had registered written agreements in court.)
• While four of the six respondents were employed, only one respondent reported that her wage was her main source of income; four others said they relied mainly on Department of Social Security (DSS) benefits or child support. (One respondent did not provide any information on income.) The five respondents who stated their income had personal gross annual incomes (including social security) below $25,000, which were supplemented (substantially in four cases) by child support, spousal support, or both.
• Five of the six respondents had no post-school education.
• Four of the six respondents had spent a sizeable percentage of the time they were married to the payer (40 per cent or more) out of paid work.
• Five of the six respondents reported that they took most or complete responsibility for several household tasks during the marriage.⁽¹⁸⁾ (The other respondent said she and her former spouse shared equally most tasks.)
• Three of the six current receivers reported that a former spouse's affair was the main reason for marital dissolution. (This was more than twice the proportion in the overall sample of women who had been separated for less than three years.)
• Three of the six respondents had been separated for less than two years; of these three, two had not yet finalised their property.

In summary, the six female respondents who were currently receiving spousal support (a) had been in relatively long-term marriages; (b) had used legal processes to settle property; (c) were sole parents; (d) reported making important indirect, non-financial contributions to the marriage; (e) had, in several cases, spent a sizeable amount of the marriage out of paid work, and (d) appeared to be in economic need. ⁽¹⁹⁾ These characteristics (with the exception of work discontinuity and the domestic division of labour) were more marked in this subsample than in the overall sample of women who had separated in the last three years. But this might have occurred by chance.

Given the dominant rationale and legislative base for spousal support, it would be expected that those who received spousal support would demonstrate needs linked to the circumstances of the marriage. Such needs are more likely to exist in longer marriages where there is a gendered division of labour, discontinuous workforce participation and continuing responsibility for young children. It is, therefore, somewhat surprising that these characteristics were not more marked in this group of receivers of spousal support than in the overall sample. Further, there are clearly many who are 'in need' as result of the marriage who do not receive spousal support.

An interesting observation is what seems to be an association between an affair being the perceived main cause of marriage breakdown and the receipt of spousal support. While the legislative scheme makes no allowance for consideration of such factors, the perceived 'justice' of a claim for spousal support may be greater in such cases. This may affect the extent to which the claimant presses the claim for spousal support, the extent to which lawyers and other advisers suggest this course of action and/or support this person in their claim, the willingness of the other party to agree to pay, and the willingness of judges and other decision-makers to make such an order.

Again, the fact of formal contact with the court system is what might have been expected. Although spousal support awards appear to be rare and only supported in any circumstances by just over a half of the divorced population, they occur in a context where there may be increasing interest within some parts of the legal community in spousal support. Outcomes producing spousal support payments are, therefore, more likely to arise where there has been such formal contact with the system and, in particular, where claims have been litigated.

On the other hand, many of those going through a divorce who have contact with the legal system do not pay or receive spousal support. It is likely that many legal practitioners have not absorbed the messages in the Full Court decisions referred to earlier, and the 'clean break' philosophy may accord with dominant values in the profession and lead to applications for spousal support being discouraged. As indicated below, this may be a partial explanation for the mismatch between attitudes in the divorced community towards spousal support and levels of payment and receipt of it. It is likely, then, that there are considerable differences in the views of practitioners on the subject of spousal support and that these are having a variety of impacts on outcomes.

It is also noted that of the six current receivers, three had been separated for less than two years, and two of these three had not yet finalised their property. It may be that in these cases support is being provided for a short time to enable the recipient to 'get back on her feet',⁽²⁰⁾ again the rationale for the award of spousal support which is most acceptable in the divorced population. It is also likely that in many cases spousal support is paid until a property settlement occurs; this may account for the fact that half the current receivers had only been separated for a short time.

Current payers of periodic support

Current payers of periodic support were predominantly male (seven men and one woman.

• All eight respondents who were still paying spousal support had been in marriages of less than ten years duration.
• All had young children from the relevant marriage. (None of these children was older than nine years at the time of separation or older than 12 years at the time of interview).
• Six of the eight respondents (including the one female) were still paying child support. (One respondent was a resident father and another had a shared parenting arrangement.)
• Five of the eight respondents (including the one female) had no children living in their household. (Four lived alone and one lived with a partner.)
• Five of the six respondents who had finalised the division of their assets had done so with some involvement with the legal system. (One respondent had his case determined by a judge and four others had registered written agreements in court.)
• Seven of the eight respondents (including the one female) were currently employed, six of whom had personal and household gross annual incomes less than $52,000.
• Seven of the eight respondents (including the one female) had continuous employment throughout the marriage.
• Four of the eight respondents (including the one female) had separated within the past three years. To sum up, the eight respondents who were currently paying spousal support (a) were typically non-resident parents of young children (and thus were still paying child support); (b) settled their property predominantly by registering an agreement in court; (c) had continuous employment throughout the marriage; and (d) had been in marriages of less than ten years duration. These characteristics, apart from work continuity, were somewhat more pronounced in this subset compared with the overall sample of men who had separated within the past three years. Once again, these differences might have been a function of sampling error.

Again, with the exception of the length of the marriage, one would expect that the small number of men who pay spousal support would have these characteristics, and they mirror the findings for current receivers.

Interestingly, only one of the eight respondents who was currently paying spousal support appeared to have a strong capacity to pay support. (Though the other payers were employed, their incomes were not exceptionally high.) This might suggest that, for some men, 'values' as opposed to 'capacity to pay' might partly drive financial support for a former spouse. Five of the six other men in the sample who had very high incomes (that is, $100,000 plus) (and therefore, presumably, high capacity to pay) had never paid spousal support. (We do not, however, have data about whether their former wives were 'in need').

In retrospect: previous payers and receivers

The seventeen respondents who had previously received periodic spousal support (a) appear to have been in marriages of more than ten years duration; (b) were typically resident parents of dependent children (both at separation and interview); (c) were characterised by work disruption to care for children; (d) settled their property predominantly by registering an agreement in court; and (e) did not appear to be very affluent at the time of interview.

While this descriptive profile bears an affinity with that of the six current receivers of spousal support, it does not differ markedly from that of the overall female sample, except for the tendency for the previous receivers to have been in marriages of more than ten years duration.

The thirteen male respondents who had once paid spousal support (a) had been married for more than ten years; (b) had young children at the time of separation; and (c) had made use of legal processes to divide their property. This profile, however, does not differ markedly from that of the overall male sample, except for the tendency for the previous payers to have gone to court to settle property (55 per cent versus 25 per cent).

Profiling 'property transferors'

Aside from noting the gender breakdown of the 62 respondents who reported spousal support in the form of a property transfer only (see Table 3), the profile of these respondents was not analysed any further. The main reason for this exclusion was that these data seemed too sketchy. First, the designers of the survey did not ask the direction of the spousal support property transfer, unlike periodic support. Though we would expect this direction to have occurred along gender lines, the often complex nature of property division, requiring an adjustment of pre-existing legal interests, militated against making this assumption. Second, respondents were not asked to quantify the value of the share of assets that were reportedly transferred as part of spousal support. (This estimation is as much a conceptual issue as it is a methodological one.) The relative importance of this form of spousal support is thus unclear. Third, similar to the profile for previous payers and receivers of periodic support, the picture is somewhat ambiguous for profiling 'property transferors' because much of the data collected have a current frame of reference.

Summary

The main thrust of the preceding analysis has been to identify and understand some of the factors that might underpin the occurrence of periodic spousal support. Obviously such an exercise is fraught with difficulty when dealing with such small numbers and with missing information on some variables. To this end, no claim can be made that these factors generalise to, or reflect, the Australian divorced population who have ever paid or received spousal support.

Nonetheless, the upshot of these descriptive profiles is that, aside from gender differences in child-rearing, work continuity, and earning capacity, the two main factors that seem to underpin the payment of periodic spousal support are marriage duration and use of legal processes to settle property. Interestingly, however, while receivers appear to be characterised by 'economic need', the male payers do not, in the main, seem to be 'high flyers'.

Part III: Attitudes: the basic data

Here we look at the basic data on attitudes and then at the relationship between attitudes and behaviour.

Normative attitudes

To explore the extent to which divorced people believe that a former spouse should provide direct financial assistance to the other spouse (that is, spousal support), the divorced respondents were asked the question 'Do you think spousal support should ever be paid?'

Table 7 shows that a little over half of the sample (54 per cent) believed that spousal support should ever be paid. This is quite a startling figure and indicates a significant incongruence between the attitudes of a substantial minority of the population (and a majority of men) and the provisions of the legislation (which provide an entitlement to support in certain circumstances). This incongruence may go some way towards explaining the low levels of payment and receipt of spousal support. (We explore the links between attitudes and behaviour below.) The respondents were not questioned further about why they held the attitude they did towards spousal support or about the circumstances in which it should or should not be paid.

Notes. # N=650 (71 respondents "couldnít say and 1 other refused to answer this question). The gender difference was statistically significant: p<.0001.

Divorced females were more inclined to be of the view that spousal support should be paid (62 per cent) than divorced males (43 per cent). This difference may be partly explained by the fact that receivers of spousal support (mostly women) are more likely to be in favour of it than those who have to pay it (mostly men). On the other hand, only a small proportion of this sample had actually paid or received spousal support, so it is likely that gender politics are important background here. This is not to suggest that there is, for example, a consistent 'feminist' line on spousal support. While it is a consistent theme in feminist writing to point to the economic disadvantages women suffer on the breakdown of relationships, it is only for some that spousal support provides a part of the solution in some circumstances. 'Liberal feminists' reject the dependency this conveys (Deech 1977). For other feminists, dependencies are state rather than private responsibilities (Fineman 1995). Views about spousal support also cut across political lines; there is surprising agreement about the need for spousal support among conservative economic analysts of law (Posner 1986) and critical feminist writers (Minow 1990).

Although respondents were not asked in what circumstances spousal support should be paid, respondents who believe that support should be paid were also asked the duration for which this support should last.

Table 8 shows that almost two-thirds (65 per cent) of those who believed that spousal support should be paid thought that support should continue until a former spouse is 'on her or his feet'. Another 20 per cent believe that support should continue for a limited time until the former spouse repartners. Only a minority of respondents (5 per cent) held the view that spousal support should be paid indefinitely.

Somewhat predictably perhaps, divorced women were more likely than divorced men to believe that spousal support should be paid until the former spouse is 'on her or his feet' (69 per cent versus 57 per cent), whereas divorced men were more likely than divorced women to believe that financial support should last until the former spouse repartners (30 per cent versus 14 per cent).

Notes. #N=650; missing cases = 357 (17 respondents "couldn't say"; 
the 268 respondents who said spousal support should never be paid were not asked this question, nor were the 72 who couldn't say if it should ever be paid); 
* 'other' includes 6% who said while kids are dependentí and 4% who said ë it dependsí the gender difference was statistically significant: p<.01.

To some extent the different views on this question reflect differences on the purposes that spousal support should serve (although views on this question would have been teased out more effectively by asking in what circumstances spousal support should be paid). There is no consistent philosophy about this underlying the legislation or the case law. Historically, support obligations were justified in terms of rights and duties rather than the utilitarian policies that generally lie behind contemporary spousal support laws (Parker, Parkinson & Behrens 1994, p. 519). The idea that a spouse should have a lifelong entitlement to spousal support (which existed, at least in theory, under the old law) clearly has little support within the divorced population given the small number of respondents who thought that spousal support should be paid indefinitely.

The alternative functions that can be served by spousal support were summarised in the Family Law Council's (1989) paper Spousal Support as rehabilitative support, custodial support, compensation for past contributions, lifestyle support, punitive damages, and reduction in welfare expenditure. In recent case law (discussed above), the emphasis has been on using spousal support as a means of compensating for the economic costs of the division of labour within relationships. Those respondents who reported that spousal support should be paid until the receiver is 'on her feet' (the largest percentage of respondents) are presumably in favour of support for rehabilitative, and possibly custodial, support purposes. Respondents who reported that spousal support should be payable until repartnering are probably recognising a variety of purposes for spousal support, but are not endorsing an approach that would provide compensation for the economic costs of the division of labour within the relationship irrespective of whether or not the payee has repartnered.

Relationship between attitudes and behaviour

When Tables 3 and 7 are compared, a possible discrepancy is apparent between the attitudes of divorced women and men towards spousal support and their actual behaviour. While over half of the sample (54 per cent) believed that spousal support should be paid in some circumstances, only a small proportion (7 per cent) had ever actually paid or received periodic spousal support, with an additional 10 per cent of divorced respondents reporting that they had paid or received a larger share of the assets as spousal support.⁽²¹⁾ This apparent disparity was more pronounced for women than men.⁽²²⁾

A number of explanations for what seems a likely disparity suggest themselves. A voluminous social-psychological literature demonstrates that attitudes do not necessarily predict behaviour (Sears et al. 1988). The need to be seen in a favourable light by others (that is, social desirability) often drives this disparity. Favourable attitudes to the payment of spousal support thus may be another case in point.

Other factors to note are that there may be a mismatch between capacity to pay (particularly where there is a child support obligation and little property) and willingness to do so.

There may also be a mismatch between the law about, and the legal system's attitudes towards, spousal support, and broader community attitudes. Thus, while a payer may be willing to pay and a receiver keen to receive spousal support, their negotiations will take place in the 'shadow of the law', which includes the legal advice they are given. If the advice is that we do not usually pay or receive spousal support (although, given the recent Full Court decisions mentioned above, this advice may be starting to change in some legal communities at least), this may impact on outcomes irrespective of individual attitudes. However, the profiling of payers and recipients of spousal support suggests that those paying or receiving spousal support were more likely than those in the sample as a whole to have had formal contact with the legal system and we speculated there that this may be explained by some increasing interest in some parts of the legal community in spousal support and by the need to litigate novel claims.

Interestingly, the attitudes of those who had ever paid or received periodic support (n=44) showed an even greater degree of assent for the idea of support than the overall sample, as well as a larger gender effect. Most females (90 per cent) who had ever received support believed in the principle of spousal support, compared with over half (56 per cent) of men who had paid support. A similar pattern held when property adjustments reported as spousal support were included (81 per cent versus 58 per cent). Behaviour thus appears to reinforce attitude (though it is possible that those with a stronger attitude were more determined to obtain support).

While payers and receivers of spousal support were more likely than those in the sample to believe spousal support should be paid, it is notable that around half of those men who actually paid support might not have been happy about this (either at the time or in retrospect). Further, 10 per cent of women who received spousal support believed that it should never be paid. A higher proportion of the respondents might have felt that spousal support should not have been paid or received in their case; the data did not enable us to measure this

Part IV: Possible correlates of attitudes towards spousal support

Attitudes towards spousal support may be influenced by many factors. Understanding the factors that lie behind community attitudes to laws will help policy-makers and legislators to determine what impact, if any, those attitudes should have on law reform and policy directions, and also the shape of any reform. As an example, if women who have experienced violence are less supportive of spousal support than those who have not, this might lead us to think carefully about reform directions that would make a history of violence one ground for the award of spousal support. It would certainly lead us to critique any proposal that required an application for spousal support as a prerequisite to receiving social security unless there was an exception in the case of violence.

We used the statistical technique of logistic regression to determine the relative importance of some possible key factors. The outcome measure was the question 'Should spousal support ever be paid?' (yes versus no). Twelve predictors were specified in the model:

• age of respondent;
• region (urban versus rural);
• education (no post-school versus post-school);
• current relationship status (single versus repartnered);
• children in the household (at least one versus none);
• year of final separation;
• legal process used to settle property (formal versus informal);
• proportion of the marriage out of paid work (women: less than 33 per cent versus 33 per cent or more; men: some versus none);
• employment status at separation (in paid work versus not in paid work);
• employment status at interview (in paid work versus not in paid work);
• violence in the relationship (yes versus no); and
• initiator of the decision to separate (respondent alone versus former spouse alone or joint).

Number and selection of variables in the logistic model

Too many variables in a statistical model relative to the number of respondents limits the extent to which it can be generalised to the population under examination (Tabachnik & Fidell 1989). After the deletion of 214 cases with missing values (125 women, 89 men), data from 236 women (65 per cent) and 200 men (69 per cent) were available for analysis. Consequently, we restricted the number of variables to 12 (that is, a minimum respondent-to-variable ratio of 20:1 as suggested by Tabachnik and Fidell 1989).

Variables were selected on both theoretical and practical grounds. For instance, we were interested to see whether there was a generational difference in attitudes to spousal support. Older respondents are probably more likely than younger ones to have had traditional marriages (that is, male breadwinner, female homemaker) and to have been brought up with a contractual view of marriage; these are two factors that may result in greater support for spousal support. Furthermore, for many women who have been subjected to violence in their marriages, the existence of an ongoing obligation owed to them by their abuser may be dangerous, frightening and/or inappropriate. It is for this reason that recipients of social security who are in fear of violence are often excused from the requirement to apply for child support from their abuser. We were, therefore, interested to explore the relationship between attitudes to spousal support and the experience of violence in the marriage in question.⁽²³⁾

Other variables were also worthy of inclusion, but sizeable amounts of missing information, substantial overlap with other variables in the model, or little variability of responses restricted the inclusion of these variables in the statistical model. For example, many respondents (especially females) could not provide property estimates, and there was a strong correlation between the age of the respondent and the age of children (r(610)=.8, p<.001). Moreover, because the original model contained several series of preliminary binary variables, an initial regression was computed to flag redundant variables. These variables were then dropped from further analyses for economy.

Since divorced women and men differed significantly in their attitudes towards spousal support, separate direct regression models were computed for each gender. Appendix C lists the parameter estimates (b), standard error estimates, significance, and log odds ratios (eb) for the 12 predictors for females and males. Details of the coding of each predictor are provided in Appendix D.

Correlates of women's attitudes

The statistical model comprising the above 12 predictors was statistically reliable for women (X² (12)=48.99, p<.0001), indicating that the predictors (as a set) reliably distinguished between women who believed support should be paid and women who did not.

Of the 12 predictors, women's age represented the single most important factor contributing to their views about the appropriateness of spousal support. The odds ratio (e^b=2.29; p<.001) indicates that for each ten-year increase in a woman's age, she is twice as likely to believe that support should occur. Thus women of 60 years of age would be six times as likely to think support should occur than women aged 30 years. We acknowledge that the interpretation of this variable is a little more complex than described here because age may be acting as a 'surrogate' for several other variables, including intergenerational differences in values and experiences, earning capacity, and the age of children.

The next most important factor for women was the initiator of the decision to separate. Women who made a unilateral decision to separate were half as likely to believe that support should be paid than those whose husband had made the decision or who had made the decision jointly (e^b=.46; p<.05). While 'matrimonial fault' is not legally relevant to decisions about spousal support, for some women it may be more appropriate that spousal support be awarded as a form of compensation (or retribution) under conditions of perceived fault (for example, where the husband made the decision to leave).

The third most important factor for women was the proportion of the marriage that they had spent out of paid work - typically to care for children. Women who had spent less than one-third of the marriage out of paid work were twice as likely to believe that spousal support should be paid than women who had spent more than one-third of the marriage out of paid work (e^b=2.02; p<.05). This result is surprising, since one of the purposes of spousal support is to provide compensation for women for their indirect, non-financial contributions to marriage (especially that related to caregiving).

Funder (1986) found that women (and men) undervalued non-financial contributions to marriage. It follows, then, that women who have spent a substantial proportion of the marriage out of paid work to care for children may believe that they do not deserve spousal support. Alternatively, women who have worked throughout the marriage may believe that spousal support should be available to other women who are not in their circumstances. (We do not know whom the respondents were thinking of when they answered this question: themselves or other women.)

The fourth most important factor for women's attitudes towards spousal support was the reported experience of violence. Women who indicated no violence in the marriage were almost twice as likely to believe support should be paid than women who reported experiencing spousal violence (e^b=1.93; p<.05). The experience of spousal violence thus seems to promote the notion of a 'clean break'. It is also possible that the low self-esteem that often accompanies the experience of violence may foster a sense of 'undeservedness'.

Finally, women who did not have children in their households were more than three times as likely to believe support should be paid than women who had children in their households (e^b=3.33; p=.08). This result approached statistical significance. Once again, it suggests the undervaluation of non-financial contributions to marriage by those who make them, as well as the possible impacts of intergenerational differences (as outlined above). It is also possible that women with children think that child support is the more crucial issue when there is a limit to how much a former spouse can pay.

Correlates of men's attitudes

We used the same 12 predictors for the men's model, except that the 'time out of paid work during marriage' variable was dichotomised as 'none' versus 'some'. (Only 26 per cent of men had spent some time not in paid work during marriage.)

The logistic model for men was statistically reliable (X² (12)=23.02, p<.05), indicating that the predictors (as a set) reliably distinguished between men who believed support should be paid and men who did not.

As with women, the single most important factor for men contributing to the attitude that spousal support should be paid was men's age. The odds ratio (e^b=1.64; p<.01) indicates that for each ten-year increase in a man's age, he is 1.6 times as likely to believe that support should occur. Thus men of 60 years of age would be nearly five times as likely to think support should occur than men aged 30 years. Again, this may reflect different views between generations about marriage and the role of law. This finding may also reflect the fact that the 'opportunity costs of marriage' for a woman will be greater the longer the marriage is. Older men may be better able to see and react to this experience on the part of their former spouse than younger men from shorter marriages.

The only other factor in the model that appeared to influence men's attitudes towards support was whether men lived in urban areas or in rural areas. Men who lived in urban areas were more than twice as likely to believe that spousal support should be paid than men who lived in rural areas (e^b=2.27; p<.05). This is somewhat surprising, given that one might expect more traditional views about marriage to be held by men from rural areas (just as we have suggested this as an explanation for the disparity in attitudes between older and younger men). On the other hand, urban men may have more progressive political views (particularly on gender issues) than rural men and may thus be more willing to accept redistribution of wealth to redress the economic disadvantage flowing from separation.

Supplementary analysis: property pool

As noted earlier, property estimates were not incorporated into the model because of a large amount of missing data - especially for female respondents. These missing data would have substantially reduced the predictive power of the other variables in the model. Nonetheless, a supplementary analysis was conducted of property estimates in relation to attitudes towards spousal support. This consisted of a simple cross-tabulation of one variable against the other for those who provided sufficient information on their assets (65 per cent of men, 44 per cent of women).

To distinguish substantial joint assets at separation from modest assets, couples' net personal assets after property division (CPI adjusted to 1996 dollars) were subjected to the Department of Social Security Assets Test cut-off (currently $265,500 for non-home-owner couples⁽²⁴⁾ ). Table 9 presents these data.

Notes. # N=650: missing cases = 291 (100 males, 191 females).

Table 9 shows that respondents who had relatively large net joint assets at separation were more likely to support the notion of spousal support than respondents with a smaller pool of joint property (that is, assets below the DSS Assets Test). These differences were statistically significant for both women (p<.01) and men (p<.05). This finding is not surprising given that those parties with greater asset wealth tend to be older and to have formalised agreements on property matters; these latter two factors appear to be associated with positive attitudes towards spousal support.

Summary

For women, four of the 12 factors in our statistical model appeared to foster the belief that spousal support should be paid. These factors (in decreasing importance) were increasing age, joint or no participation in the decision to separate, a relatively small percentage of the marriage out of paid work, and no violence in the marriage. For men, only two factors appeared to foster the belief that support should be paid: increasing age, and living in the city (as opposed to living in the country).

These findings should be viewed as exploratory, however, because methodological issues limited the inclusion of all relevant variables. ²⁵ Though they warrant confirmation among other divorced samples, these findings may nonetheless serve as useful markers for consideration.

Part V: Overview, policy issues and further research agendas

Overview

This working paper seeks to estimate the incidence of spousal support in Australia and to assess the degree to which divorced Australians believe this form of support should be paid.

Our data suggest that periodic spousal support continues to be rare, minimal and brief. Over the last decade, it appears to have occurred in less than 7 per cent (±2 per cent) of divorces, typically lasts for about two years, and averages about $128 per week ($6,640 per annum). There is, nonetheless, considerable variability among individual payers and receivers. One in ten respondents reported a property transfer as spousal support (though men were twice as likely as women to report that this form of spousal support had occurred).

While the vast majority of 'payers' were male and the vast majority of 'receivers' female, both shared at least two characteristics: they used formal legal processes to settle property matters (for example, went to court) and had been in relatively long-term marriages of at least ten years duration. Whereas receivers appeared to be characterised by 'economic need', payers did not in the main appear to be 'high flyers' in terms of their current earnings. For some payers, values and attitudes might partly drive support.

The picture on attitudes is difficult to describe. On the one hand, it could be said that a high degree of assent was given to the principle of spousal support. Over half the sample (particularly women) believed that spousal support should be paid in at least some circumstances, and that it should last until a former spouse was 'on her (or his) feet'. For some men, however, support seemed to be conditional on their former spouse not repartnering. Described in another way, however, these findings indicate that a substantial minority of the divorced population thinks that there should NEVER be entitlement to spousal support.

Older respondents (both female and male) were more likely than younger respondents to think support should be paid; wealthy respondents with large net joint asset pools were more likely than less wealthy respondents to think support should be paid. Likewise, respondents who had paid or received periodic support were more likely to think it should be paid than those who had not paid or received support.

In contrast, women who had initiated the decision to separate, who had experienced physical violence in the marriage, or who had spent a fair amount of the marriage out of paid work (mainly to care for children) were less likely than other women to think support should be paid. These attitudes appear to be underpinned by a loose amalgam of feelings of fault and guilt, and the undervaluation of non-financial contributions. That women who reported experiencing violence prefer a 'clean break' is understandable.

We suggest that one interpretation of the data collected as a whole is that there is a substantial mismatch between attitudes to spousal support (a majority in favour), and behaviour (very low levels of incidence). One explanation for this apparent disparity is the potential mismatch between the capacity to pay and the willingness to do so. There may also be a mismatch between the law about spousal support and the legal system's attitudes towards it, and broader community attitudes. Thus while a payer may be willing to pay spousal support and a receiver keen to receive it, their negotiations will take place in the 'shadow of the law', which includes the legal advice they are given. If the advice is that we do not usually pay or receive spousal support, this may impact on outcomes irrespective of individual attitudes.

Policy and reform issues

This study provides empirical support for the proposition that spousal support issues are fundamentally gender issues. Those who receive spousal support are almost always women, and those who pay it almost always men. Further, women are more likely to support the existence of spousal support obligations than are men. These findings provide an indication that decisions about reform and policy in this area hold important implications for women's equality before the law. On the one hand, as indicated in the recent case law, the payment of spousal support is one limited mechanism through which to address women's economic disadvantage on marriage breakdown. If, as this study finds, such obligations are created only rarely, then redressing that disadvantage depends upon changing those patterns of payment and/or looking to other broader strategies (including social security and taxation reform).

One picture that can be drawn from the data and analysis carried out for this study is of a legislative framework for the creation of obligations and entitlements to pay and receive spousal support which rarely results in the actual payment or receipt and which only slightly over half the divorced population (and a minority of men) believes should ever be imposed. This mismatch should feed into the debates about the use of spousal support in contemporary social and economic conditions.

The fact that legislative provisions are rarely used is not necessarily of concern. Provisions might have been enacted to deal with unusual cases. (The provisions making step-parents liable in some circumstances for child maintenance is a family law example.) Provisions may also be designed to be norm-setting in a way that does not necessarily require their use as such. (Many of the provisions on children under the reformed part VII of the Family Law Act fill this function.) However, the provisions on spousal support do not clearly fall into either of these categories. Moreover, they may lead to a misleading picture that stifles or subverts other policy and reform agendas. For example, the presence of legislative spousal support entitlements might be relied upon to defeat arguments for reform of matrimonial property law, or for putting more resources into state support in situations of economic need, or for encouraging women to remain economically independent in their relationships. Further, the provisions may create unrealistic expectations among 'users' of the Family Law Act which may then affect post-separation negotiations. To some extent at least, these expectations may affect their behaviour within marriage (for example, decisions by women about whether to remain in the paid workforce or not).

If it is accepted that the mismatch between the presence of the provisions and their actual use is a problem, there are two possible reform and policy responses. The first is to accept that the spousal support provisions are appropriately framed and that the problem is with other factors that impact on the limited way in which they are used; it is these factors that need to be addressed in a reform agenda. These may include the attitudes of lawyers and judges, the difficulties of process, the lack of legal aid, and so on. The second is to suggest that the law needs to be brought more closely in line with the practice. In the case of the spousal support provisions, this may require them to be drawn more narrowly to reflect the circumstances in which spousal support tends to be paid and received currently.

The following are some options for reform which were suggested by the Family Law Council in 1989:

1. Do nothing. That is, allow spousal maintenance to be awarded occasionally and randomly on the variety of principles already contained in the Family Law Act 1975.
2. Leave the legislative provisions unchanged but seek to raise the consciousness of the legal profession and judicial officers that in certain specified fact situations, more generous and consistent awards of spousal maintenance ought to be made.
3. Abolish spousal maintenance altogether (either retrospectively or prospectively from some future date after a period of notice).
4. Amend the Family Law Act 1975 to provide that other than in exceptional circumstances, spousal maintenance (based on need and ability to pay) should only be awarded for a limited period (e.g., three years).
5. Amend the Family Law Act 1975 to provide that other than in exceptional circumstances, spousal maintenance should only be awarded for a limited period for the purpose of enabling the payee to undertake specified retraining.
6. Amend the Family Law Act 1975 to provide that other than in exceptional circumstances, spousal maintenance should only be awarded (for limited or long-term periods of time) for the losses of the payee's earning capacity 'caused by' or 'arising out of' the marital relationship. (p. 38)

Suggestion 2 reflects the first policy/reform response identified above (namely, attempting to change non-legislative factors to bring practices in line with the legislative provisions). In terms of the alternative option - namely, attempting to achieve a match between what happens currently (in the limited way identified in this study) and the legislative provisions - none of the options proposed by the Family Law Council captures what we know from the profiling exercise. Many of those who receive spousal support would not come under one of the more limited categories identified above. The analysis in this project provides some indications of those circumstances, although it certainly does not paint a clear enough picture to enable drafting of legislative provisions. One response to the findings of this profiling exercise is that it is important to leave the legislative provisions as broad as possible to provide maximum flexibility. (It is this flexibility that the profiling exercise suggests is being taken advantage of). 

However, if legislative reform more carefully specifies the circumstances in which spousal support is payable, we might be concerned about the lack of consistency and uniformity in the operation of the law (of which this study provides an indication). One of the reasons we might be concerned about this is that it may make obtaining private orders more difficult; if the shadow the law casts is very unclear, negotiating may take longer and be more costly in all sorts of ways (Dewar 1998).

Family law in Australia has, since 1975, been based on a highly discretionary model. The current provisions on spousal support reflect this model. It has been suggested that we are moving into a new stage of family law characterised by more rules (less discretion), more rights (less utilitarian impetus), and more normative content (less 'leave it to the experts') (Dewar 1998). An example of law reform that reflects this new stage includes the adoption of a child support scheme where discretion is only able to be exercised after the application of a formula to calculate the extent of the obligation. Given the inconsistent circumstances in which spousal support is and is not paid, such a system might be desirable in the area of spousal support too. On the other hand, deciding what the content of such rules or a formula should be would clearly be very difficult, and could not be done in isolation from property law reform and child support.

The focus in this commentary so far has been on the mismatch between legislative entitlements and receipt or payment of spousal support. Other possible mismatches identified in this study are between the attitudes of the divorced population and the legislative provisions, and between the attitudes of the divorced population and their behaviour.

It is a notable finding that only just a majority of the divorced community (and a minority of men) support there being any spousal support entitlement at all in any circumstances. As already indicated, it is hard to know whether to describe this from the point of view of the glass being half full (a majority support an entitlement) or half empty (a substantial minority do not support it). In any case, legislation does not, and should not always, follow community attitudes, although where it does not this may create compliance difficulties. Also, there are many examples, particularly in the family law area, of legislation being consciously used in an attempt to change community attitudes and behaviours (for example, the amendments on child support and parenting decisions). On the other hand, legislation is unlikely to be a very effective tool for achieving such attitudinal changes, and certainly needs to be accompanied by broader strategies. In addition, as pointed out by Funder and Smyth (1996), mismatches need to be noted as they may reflect a potential source of discontent.

If community attitudes to spousal support (and more generally to the financial obligations of spouses to each other after marriage breakdown) are a problem, then community education programs and other interventions may be appropriate. In particular, the idea of marriage as a partnership, the economic costs of which should be shared by the parties on marriage breakdown,⁽²⁶⁾ might be promoted. One way in which this might occur is through the adoption in part VIII of the Family Law Act of an 'objects and principles' provision similar to that recently inserted into part VII (on children).⁽²⁷⁾ This would provide more guidance about the circumstances in which financial and property adjustments should be made, and could provide a limited educative function.

If the problem is seen to be not the attitudes but the mismatch, then it might be appropriate to look at reform agendas that attempt to reduce that mismatch. The basic data gathered in the Australian Divorce Transition Project and reported in this working paper indicate that the greatest level of support for spousal support being payable is expressed for the situation 'until a person is on his or her feet'. While legislative provisions could embody this position, support for such a position is only evident in a minority of respondents to the survey (but a majority of those who support spousal support in some circumstances).

In terms of the mismatch between attitudes and behaviours, the point here is that while over 50 per cent of the divorced population say spousal support should be payable in some circumstances, divorced people only actually pay it or receive it in a small minority of cases. Various explanations have been proffered for this mismatch in the body of this working paper. To the extent that it may reflect a difference in attitudes between the divorced population and those in the legal system (as indicated above), then education programs for the profession and other systemic changes may be appropriate. To the extent that this mismatch may reflect a disparity between willingness and capacity, the links between child support, spousal support, property adjustment and other forms of support may need to be explored.

More specifically, the material on attitudes might lead us to reflect on more specific policy questions, including the following:

• Women who have experienced violence in their relationships are less inclined than other women to support the existence of spousal support obligations (though they may be unaware of the role of the Child Support Agency as a collector and buffer in this process). This might be taken as evidence to use against legislative reform, which makes the existence of violence specifically relevant to spousal support claims (cf. Behrens 1993). On the other hand, women who have experienced violence may have greater rates of low self-esteem and be less inclined to value their contributions highly; to this extent the problem may be one of changing attitudes while recognising the inappropriateness of forcing women to claim for spousal support.
• Women who have spent a considerable amount of time out of the workforce were less likely to give responses in favour of spousal support. This is a worrying finding in the sense that these women are more likely to be in a position of economic dependence on separation and, objectively, to 'need' spousal support. In addition, as indicated previously, it may be further evidence of devaluation by women of their contributions, which is likely to be reflected in property settlements as well as in decisions about spousal support. This may be an area where community education programs are warranted, although such programs might be 'pushing the pram uphill' because of the high value placed on paid work in our society.

Further research agendas

There are a number of issues around spousal support that have not been the subject of this study, or about which we have not been able to reach any conclusions because of the nature of the study and the data received. While all aspects of the study warrant further exploration, the following issues appear to be particularly germane. Cross-validation All data sources have their own strengths and weaknesses. The estimates reported in this paper should be cross-validated with data from other sources (for example, court statistics, Child Support Agency records, lawyer surveys) as well as with other samples (such as separating couples with no dependent children at separation). We noted earlier that the Family Law Council is currently conducting its own research into legal practitioners' experiences of, and attitudes towards, spousal support. Its preliminary findings, reported in Boland 1998, appear to be consistent with the findings presented in this paper regarding the relatively low incidence of spousal support, and the relevance of practitioner attitudes to whether spousal support is sought.

Cohabitation Non-marital cohabitation is an increasingly widespread family type and is afforded equivalent legal status to marriage in many areas of the law. In some States and Territories, people who have been in de facto or domestic relationships can claim spousal support from the other party. For methodological reasons, this paper has focused on those who have divorced under the Family Law Act or settled in its 'shadow'. The extent to which the incidence of spousal support and attitudinal patterns for the divorced population hold for de facto and domestic relationships nonetheless warrants investigation.

Indirect transfers Most estimates of spousal support typically refer to periodic payments and tend to exclude other forms of support - most notably property transfers. Though our data sought to estimate the frequency of less direct forms of support, they did not allow estimation of the value of these forms of support. Reliable data on the size of lump sum cash or property transfers as spousal support would be useful, not just to get a better sense of the relative importance of each type of support but also because indirect transfers (if sizeable) might partly bridge the apparent gap between support for the principle of spousal support and its apparent low occurrence. But partitioned data may be hard to obtain, and we would argue that while the law might seek to partition out the different components of maintenance, the boundaries may be quite fuzzy in people's minds.

Temporal and regional variations Though there was some evidence of a slight increase in the occurrence of periodic support in recent years, this increase was not statistically significant. Presumably it will take some time for judicial messages to impact on individuals who are negotiating arrangements without legal advice. Future research should thus be conducted in two and five years to assess the extent to which the rate of occurrence might have changed across time. Moreover, the present investigation was unable to explore regional variations by States or Territories due to the small number of applicable cases. This would be another useful area to pursue, as earlier studies have indicated considerable regional differences (Family Law Council 1989, p. 21, para 8.4).

Supporting children and a former spouse Like child support, spousal support is one possible cash or asset transfer in a complex mosaic of post-separation transfers. Little is known, however, about the interaction between spousal support and other post-separation financial arrangements - especially its interaction with child support.

While the above issues have a quantitative bent, qualitative research would be useful in a number of areas. Some possible research questions that might benefit from a semi-structured, in-depth, contextual approach using open-ended questions are:

• What do members of the general population know about spousal support?
• Do separating individuals receive advice from legal practitioners about spousal support and, if so, what advice do they receive?
• At a personal level, do legal practitioners and other legal professionals think spousal support should ever be paid and how does this impact on the advice they provide?
• Why do respondents pay support outside of court-mandated support?
• Why do respondents think support should or should not be paid?
• What type of spousal support would receivers like to get and why?
• Do former spouses help each other in ways not formally measured that might constitute spousal support (for example, Higher Education Contribution Scheme fees, rent, home loan, car payments, car maintenance)?

Given the continuing debate about spousal support and its potential importance in redressing the economic inequalities that follow marriage breakdown, these issues are worthy of further research.

References

• Australian Bureau of Statistics (ABS) (1992), Work Patterns for Women, Victoria, October 1991, Catalogue no. 6204.2, Australian Bureau of Statistics, Melbourne.
• Bala, N. (1994), 'Spousal support law transformed - Fairer treatment for women', Canadian Family Law Quarterly, vol. 11, pp. 13-56.
• Behrens, J. (1993), 'Domestic violence and property adjustment: A critique of "no fault" discourse', Australian Journal of Family Law, vol. 7, pp. 9-28.
• Bittman, M. (1991), Juggling Time: How Australian Families Use Time, AGPS, Canberra.
• Boland, J. (1998), 'Spousal maintenance - The how, what, when and why of spousal maintenance', Paper presented to the 8th National Family Law Conference, October 24-28, Hobart, Tasmania, in Conference Handbook: The Challenge of Change, Television Education Network, Melbourne, pp. 267-96.
• Davies, C. (1994), 'Compensatory support: New beginnings or a return to the past?' Canadian Family Law Quarterly, vol. 11, pp. 129-45.
• Deech, R. (1977), 'The principles of maintenance', Family Law, vol. 7, pp. 229-33.
• Dewar, J. (1998), 'Reducing discretion in family law', Australian Journal of Family Law, vol. 11, pp. 309-26.
• Diduck, A. & Orton, H. (1994), 'Equality and support for spouses', The Modern Law Review, vol. 57, pp. 681-702.
• Edgar, D. (1996), 'The certainty of uncertainty: Let's aim for equity and live with that', in The 7th National Family Law Conference: The 3 R's: Relationships, Rights, Responsibilities: Conference Handbook, Television Education Network, Melbourne, pp. 1-18.
• Family Law Council (1989), Spousal Maintenance Discussion Paper, AGPS, Canberra.
• Fineman, M. (1995), The Neutered Mother, the Sexual Family and Other Twentieth Century Tragedies, Routledge, New York.
• Finlay, H., Bailey-Harris, R. & Otlowski, M. (1997), Family Law in Australia, 5th edn, Butterworths, Sydney.
• Funder, K. (1986), 'His and her divorce', in P. McDonald (ed.) Settling Up: Property and Income Distribution on Divorce in Australia, AIFS, Melbourne, pp. 224-40.
• Funder, K. (1993), 'Women's post-separation employment and reliance on social security', in K. Funder, M. Harrison & R. Weston (eds.) Settling Down: Pathways of Parents After Divorce, AIFS, Melbourne, pp. 83-115. 
  
  Funder, K. & Harrison, M. (1993), 'Drawing a longbow on marriage and divorce', in K. Funder, M. Harrison & R. Weston (eds.) Settling Down: Pathways of Parents After Divorce, AIFS, Melbourne, pp. 13-32.
• Funder, K. & Smyth, B. (1996), Family Law Evaluation Project 1996: Parental Responsibilities: Two National Surveys, AGPS, Canberra.
• Harrison, M. (1993), 'Patterns of maintenance payment over time', in K. Funder, M. Harrison & R. Weston (eds.) Settling Down: Pathways of Parents After Divorce, AIFS, Melbourne pp. 116-34.
• Harrison, M., Funder, K. & McDonald, P. (1993), 'Principles, practice and problems in property and income transfers', in K. Funder, M. Harrison & R. Weston (eds.) Settling Down: Pathways of Parents After Divorce, AIFS, Melbourne, pp. 192-209.
• Harrison, M. & Tucker, T. (1986), 'Maintenance, custody and access', in P. McDonald (ed.) Settling Up: Property and Income Distribution on Divorce in Australia, AIFS, Melbourne, pp. 259-78.
• Huberman, A. M. & Miles, M. B. (1994), 'Data management and analysis methods', in N. K. Denzin & Y. S. Lincoln (eds.) Handbook of Qualitative Research, Sage, California, pp. 428-44.
• Land, H. (1984), 'Changing women's claims to maintenance', in M. Freeman (ed.) The State, The Law and The Family, Tavistock Publications, New York, pp. 25-35.
• McDonald, P. (1986). 'Property distribution: The shares of each partner and their determinants', in P. McDonald (ed.) Settling Up: Property and income Distribution on Divorce in Australia, AIFS, Melbourne, pp. 171-93. 
  
  Minow, M. (1990), Making All the Difference, Cornell University Press, Ithaca. 
  
  Morgan, S. (The Hon. Justice) (1996), 'Equity does not necessarily mean equality: Spouse support and other techniques as a means of redressing economic inequalities in marriage', in The 7th National Family Law Conference: The 3 R's: Relationships, Rights, Responsibilities: Conference Handbook, Television Education Network, Melbourne, pp. 43-76.
• New South Wales Law Reform Commission (1983), Report on Defacto Relationships, New South Wales Law Reform Commission, Sydney.
• O'Donovan, K. (1982), 'Should all maintenance of spouse be abolished?', Modern Law Review, vol. 45, pp. 424-33.
• Parker, S. & Drahos, P. (1990), 'Closer to a critical theory of family law' Australian Journal of Family Law, vol. 4, pp. 159-75. 
  
  Parker, S., Parkinson, P. & Behrens, J. (1994), Australian Family Law in Context: Commentary and Materials, Law Book Company, Sydney.
• Payne, J. (1995), 'Spousal and child support after Moge, Willick and Levesque', Canadian Family Law Quarterly, vol. 12, pp. 261-99.
• Posner, R. (1986), Economic Analysis of Law, 3rd edn, Little Brown & Co., Toronto.
• Queensland Law Reform Commission (1992), Defacto Relationships, Working Paper no. 40, Queensland Law Reform Commission, Brisbane.
• Scott, H. (1984), Working Your Way To the Bottom: The Feminisation of Poverty, Pandora Press, London. 
  
  Sears, D. O., Peplau, A., Freedman, J. L. & Taylor, S. E. (1988), Social Psychology, 6th edn, Prentice-Hall International, New Jersey.
• Sheppard, C. (1995), 'Uncomfortable victories and unanswered questions: Lessons from Moge', Canadian Journal of Family Law, vol. 12, pp. 283-329. 
  
  Stake, R. E. (1994), 'Case studies', in N. K. Denzin & Y. S. Lincoln (eds.), Handbook of Qualitative Research, Sage, California, pp. 236-47.
• Tabachnik, T. B., & Fidell, L. S. (1989), Using Multivariate Statistics, 2nd edn, Harper & Row, New York.
• Weston, R. (1993), 'Income circumstances of parents and children: A longitudinal view', in K. Funder, M. Harrison & R. Weston (eds.) Settling Down: Pathways of Parents After Divorce, AIFS, Melbourne, pp. 135-55.
• Weston, R. (1986), 'Changes in household income circumstances', in P. McDonald (ed.) Settling Up: Property and Income Distribution on Divorce in Australia, AIFS, Melbourne, pp. 100-30.
• Wright, R. E. (1995), 'Logistic regression', in L. Grimm & P. R. Yarnold (eds.), Reading and Understanding Multivariate Statistics, American Psychological Society, Washington, DC, pp. 217-44.

Appendix A: Relevant legislative provisions

Family Law Act 1975 - section 75

Matters to be taken into consideration in relation to spousal support 75.

1. In exercising jurisdiction under section 74, the court shall take into account only the matters referred to in subsection (2).
2. The matters to be so taken into account are:
   • a. the age and state of health of each of the parties;
   • b. the income, property and financial resources of each of the parties and the physical and mental capacity of each of them for appropriate gainful employment;
   • c. whether either party has the care or control of a child of the marriage who has not attained the age of 18 years;
   • d. commitments of each of the parties that are necessary to enable the party to support: 
     (i) himself or herself; and 
     (ii) a child or another person that the party has a duty to maintain;
   • e. the responsibilities of either party to support any other person;
   • f. subject to subsection (3), the eligibility of either party for a pension, allowance or benefit under: 
     (i) any law of the Commonwealth, of a State or Territory or of another country; or 
     (ii) any superannuation fund or scheme, whether the fund or scheme was established, or operates, within or outside Australia; and the rate of any such pension, allowance or benefit being paid to either party;
   • g. where the parties have separated or the marriage has been dissolved, a standard of living that in all the circumstances is reasonable;
   • h. the extent to which the payment of support to the party whose support is under consideration would increase the earning capacity of that party by enabling that party to undertake a course of education or training or to establish himself or herself in a business or otherwise to obtain an adequate income;
   • i. the extent to which the party whose support is under consideration has contributed to the income, earning capacity, property and financial resources of the other party;
   • j. the duration of the marriage and the extent to which it has affected the earning capacity of the party whose support is under consideration;
   • k. the need to protect a party who wishes to continue that party's role as a parent;
   • l. if either party is cohabiting with another person - the financial circumstances relating to the cohabitation;
   • m. the terms of any order made or proposed to be made under section 79 in relation to the property of the parties;
   • n. any child support under the Child Support (Assessment) Act 1989 that a party to the marriage has provided, is to provide, or might be liable to provide in the future, for a child of the marriage; and
   • o. any fact or circumstance which, in the opinion of the court, the justice of the case requires to be taken into account.
3. In exercising its jurisdiction under section 74, a court shall disregard any entitlement of the party whose support is under consideration to an income tested pension, allowance or benefit.

Appendix B: Spousal support questions

ASK EVERYONE:

Q294. Have you EVER paid or received regular spousal support?

IF RECEIVED SPOUSAL SUPPORT, ASK:

Q295. Was that PAID or RECEIVED spousal support?

Q296. Are you still receiving or paying spousal support?

Q297A. How long have you been receiving/paying, or did you receive/pay spousal support?

IF EVER PAID/RECEIVED SPOUSAL SUPPORT, ASK:

Q298. How much was actually received or paid?

IF AMOUNT GIVEN AT Q298:

Q299. (Was that per week, per fortnight, per month or per year?)

IF ALL ASSETS SETTLED, ASK:

Q300. Did you pay or receive a larger share of the assets as part of spousal support?

Q302. Do you think spousal support SHOULD EVER be paid?

IF SPOUSAL SUPPORT SHOULD EVER BE PAID, ASK:

Q303. For how long SHOULD spousal support be paid?

Appendix C: Logistic regression tables

Notes. n=361: missing cases = 125; ns = not significant.

Notes. n=289: missing cases = 89; ns = not significant.

Appendix D: Coding scheme for logistic regression model

List of tables

• Table 1 Demographic profile (percentages) by gender: divorced respondents
• Table 2 Demographic profile (averages) by gender: divorced respondents
• Table 3 Type of spousal support paid or received by gender: divorced respondents
• Table 4 Type of spousal support by year of separation: divorced respondents who had finalised their property
• Table 5 Payment or receipt of periodic spousal support by gender
• Table 6 Duration and annual amount of periodic spousal support by payment or receipt
• Table 7 Whether spousal support should ever be paid: reports of divorced female and male respondents
• Table 8 Length of time spousal support should be paid: reports of divorced female and male respondents
• Table 9 Normative attitudes towards spousal support by net joint assets (CPI adjusted to 1996 dollars) at separation (above or below DSS assets test): divorced respondents