Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

5.1 Introduction

As in most other western developed countries, marriage breakdown has increased in Australia, particularly since the end of World War 2. While the increase in the rate of divorce in Australia has slowed since the 1980s and may have even stabilized and started to decline, the nature and characteristics of divorcing couples continue to change. It is very likely that these changes in divorce trends are underpinned at least in part by the rise of unmarried, or de facto, cohabitation (henceforth cohabitation) as an alternative or ‘stepping stone’ to marriage. Cohabiting relationships are less stable than marital relationships, but we know little about the stability of cohabiting relationships from official statistics. Thus, official statistics underestimate the true extent of relationship dissolution in the Australian population. In this chapter we document historical trends, explore changes in the nature and characteristics of divorce in Australia and examine differences in the dissolution of cohabiting and marital relationships using survey data.

5.2 Historical Trends

Rates of divorce in Australia have increased considerably over the last century. Figure 5.1 reports the crude divorce rateFootnote 1 in Australia since 1901.Footnote 2 At the turn of the twentieth century divorce was virtually non-existent in Australia, with only 398 divorces granted in 1901 and a crude divorce rate of less than 0.1 (ABS 1971). The rate then increased gradually from the mid-1960s until 1975.

Fig. 5.1
figure 1

Crude divorce rate, Australia 1901–2011 (ABS 1971, 2005a, 2012b)

Full size image

In 1976 no-fault divorce was introduced with the implementation of the 1975 Family Law Act and the crude divorce rate spiked to 4.6 per thousand head of population aged over 15 (Fig. 5.1). The new Family Law Act 1975 sought to establish a law based upon two pillars: ‘the support for marriage and family; and the right of a party to leave a marriage upon its irretrievable breakdown, the latter being evidenced by 12 months separation of the parties’ (Australian Parliament House of Representatives Standing Committee on Legal and Constitutional Affairs 1998: 95). The 14 grounds of divorce were replaced by one – irretrievable breakdown. Within a few years the crude divorce rate dropped to around 2.6 per thousand head of population over the age of 15 and has oscillated between 2.5 and 3.0 since the late 1970s. The introduction of the 1975 Family Law Act, and with it no-fault divorce, dramatically and permanently changed the rate of divorce in Australia.

Some have argued that the easy access to divorce provided by the Family Law Act was a major cause of the substantial increase in divorce in Australia from the mid-1970s. The data indicate, however, that the rise in the crude divorce rate following the introduction of the Family Law Act was relatively short-term. Within five years of the Act being introduced crude divorce rates had settled to a rate that reflected linear trends established in the mid-1960s (Ozdowski and Hattie 1981). It is likely that the spike in divorce was primarily a response to pent-up demand from couples that had separated but not divorced in the late 1960s and early 1970s. There is some survey evidence for this. Burns (1980a, b) conducted a study on separation and divorce in late 1975, prior to the introduction of no-fault divorce, and found that some separated respondents were waiting for the introduction of the Family Law Act to divorce legally. Despite minor yearly fluctuations the steady increase in the crude divorce rate evident prior to 1976 has ceased and there has been little change since the early 1980s. Since the year 2000, the trend suggests a decline in divorce (see Fig. 5.1); in 2008 divorce rates were at their lowest in 20 years (ABS 2009).

5.3 Continuity and Change Since No-Fault Divorce

Despite the plateau and decline in the crude divorce rate, divorce continues to be a pervasive feature of Australian social life. Thirty-two percent of current marriages are expected to end in divorce and this is predicted to increase to 45 % over the next few decades, with younger marriage cohorts more likely to divorce (Carmichael et al. 1996). Further, there is widespread government and community concern about divorce and its consequences as evidenced by recent government policy and legislative reforms (Australian Parliament House of Representatives Standing Committee on Legal and Constitutional Affairs 1998; Kaspiew et al. 2009).

Ongoing changes in divorce in Australia are more clearly revealed if we use measures other than the crude divorce rate. The crude divorce rate indicates the rate of breakdown in the total Australian population aged over 15, including those who are married and unmarried. Given that rates of marriage have also declined since the late 1970s, the crude rate may be under-estimating marriage breakdown because its denominator is not restricted to the married population (de Vaus 2004). An alternative indicator is a divorce rate which uses the married population as the denominator. Figure 5.2 shows the divorce rate of the married population in Australia between 1981 and 2001.Footnote 3

Fig. 5.2
figure 2

Divorce rate, Australia 1981–2001 (ABS 2005a)

Full size image

Compared to the crude divorce rate, this divorce rate is much higher. While the divorce rate shows a very similar pattern to the crude divorce rate, the peaks and troughs are more pronounced. The rate of divorce has varied from a low of 10.6 per 1,000 married men or women in 1987 to a high of 13.1 in 2001. Data from the 2006 Census indicate that this figure had dropped to 12.0, and the 2011 Census data indicate that it had further declined to 11.6 (ABS 2012c). These declines are consistent with the general decline in the crude divorce rate since 2000.

Other characteristics of divorce in Australia such as age at divorce, average time to divorce and number of dependent children involved in divorce have also changed since the 1980s. These changes reflect broader social and demographic changes in relationship formation and fertility timing in Australia over the last three decades. Figure 5.3 illustrates that since the introduction of the Family Law Act in 1976 the median age at divorce has increased from 36.1 in 1977 to 44.5 in 2011 for men and from 33.0 to 41.7 over the same period for women.

Fig. 5.3
figure 3

Median age at divorce for men and women, Australia 1977–2011 (ABS 1979–1993, 1994–2001, 2005c, 2012b)

Full size image

It is likely that the median age at divorce is increasing due to two factors. First, people are marrying at older ages. In 1977 the median age at first marriage was 23.8 years for men and by 2011 this had increased to 29.7 years. Similarly for women the median age at first marriage increased from 21.4 years in 1977 to 28.0 years in 2011 (ABS 2005d, 2012c).Footnote 4 Second, the median duration of marriage to separation and divorce has increased. Figure 5.4 shows that the main increase occurred between 1997 and 2006, when the median duration of marriage to separation increased from 7.7 to 9.9 years, and of marriage to divorce from 11.1 to 12.5 years. There was also an increase in the time between separation and divorce, from 2.7 years in 1981 to 3.5 years in 2011, with most of that increase occurring during the 1990s.

Fig. 5.4
figure 4

Median duration of marriage to separation and divorce (years), Australia, 1977–2011 (ABS 2005c, 2012b)

Full size image

The proportion of divorces involving children under the age of 18 has also changed over time. Figure 5.5 illustrates a decline in the proportion of divorces involving dependent children from 63 % in 1977 to around 50 % by 2003, and this proportion has dropped below 50 % since 2007 (ABS 2012b). This reduction in the proportion of divorces with dependent children is due in part to delayed child bearing (see Chap. 9).Footnote 5 Even though the proportion of divorces involving children has dropped since the early 1980s, the actual number of children whose parents divorce each year has remained fairly constant at around 50,000 children (ABS 2001, 2012b).

Fig. 5.5
figure 5

Proportion of divorces involving children under the age of 18, 1977–2011 (ABS 2005c, 2009, 2012b)

Full size image

In summary, despite a plateau and recent decline in the divorce rate, the nature and composition of the divorcing population has continued to change with increases in age at divorce and time to divorce and a decline in the proportion of divorces involving children. When considering these trends in marriage breakdown the limitations of official statistics also need to be taken into consideration.

First, official divorce statistics tend to underrepresent marriage breakdown at any given point because many marriages end in permanent separation and never proceed to divorce or do not proceed to divorce for several years; the median time from separation to divorce was 3.5 years in 2011 (ABS 2012b). In these circumstances marriage breakdown is not officially recorded until divorce is awarded (ABS 1999, 2000).Footnote 6

In Table 5.1, we present the results of marital history information on those who had separated or divorced from their first marriage in wave 1 of HILDA (2001).Footnote 7 We find that approximately 18 % of those who had separated from their marriage had not gone on to divorce by the time of survey. The average duration of separation of those people who had separated but not legally divorced was 5.7 years. This average is 2 years longer than that reported by official divorce statistics in 2011. This is because the ABS divorce statistics are recorded when a couple divorces. While the majority of separated people had only recently separated in the HILDA sample (63 % of them having separated less than 2 years before the survey), about 20 % of the separated people had been separated for 10 years or more without divorcing.

Table 5.1 Distribution (number and per cent (%)) separated and divorced from first marriage (HILDA 2001)
Full size table

The second major limitation of official divorce statistics is that they significantly under-represent the true extent of relationship dissolution in Australia, because they do not take into account the increasing number of cohabiting relationships. In the remainder of this chapter we examine differences in the dissolution of cohabiting and marital relationships.

5.4 Marriage and Cohabiting Relationship Dissolution: Evidence from HILDA

Arguably, many of the changes in the timing of divorce and composition of the divorcing population since the 1980s are underpinned by changes in family and relationship formation and in particular the increasing number of couples who are in cohabiting relationships (see Chap. 2). While the rise of cohabitation is contributing to changing patterns of divorce, the contribution of cohabitation to overall rates of relationship dissolution is not captured by official divorce statistics. Previous Australian and overseas research has indicated that cohabiting relationships tend to be less stable than marital relationships (Qu et al. 2009), but we know little about the pattern and nature of the differences in dissolution between the two types of relationships. To better capture the extent of relationship dissolution in Australia from both cohabiting and marital relationships, we need survey data.

The majority of previous research on cohabitation and relationship dissolution has concentrated on the dissolution of marriage after a period of cohabitation. Most studies find that a period of cohabitation prior to marriage increases the risk of subsequent divorce (Bennett et al. 1988; Teachman and Polonko 1990; Axinn and Thornton 1992; DeMaris and Rao 1992; Bracher et al. 1993; Hall and Zhao 1995; Lillard et al. 1995; Berrington and Diamond 2000; Dush et al. 2003; Hewitt et al. 2005). Far fewer studies have investigated the dissolution of cohabiting relationships that do not proceed to marriage (see Schoen 1992; Thompson and Collela 1992 for notable exceptions).

In this chapter we are not only interested in what happens after marriage (preceded by cohabitation or not), but also in what happens with cohabiting relationships that do not proceed to marriage. There are three potential pathways cohabiting relationships can follow: couples can continue to cohabit, become legally married or separate (Qu et al. 2009). To investigate relationship dissolution among cohabiting and marital relationships, we differentiate between three mutually exclusive relationship groups, those who are: (1) married without prior cohabitation, (2) cohabiting only, and (3) married after a period of cohabitation. So that we are not comparing cohabiting relationships with long-term marriages we restrict our examination to first marriages only and to relationships formed since 1995. Our sample is respondents in HILDA Waves 1–10 (2001–2010).

In Table 5.2 we show the overall proportion of respondents in the abovementioned three relationship groups, for relationships commencing between 1995 and 2010. The final column in the table provides the total proportion of each relationship type observed over that time. The most common relationships were cohabitating only relationships (42 %), followed by cohabitations that resulted in marriage (38 %), with the fewest number of people marrying directly (19 %). The small proportion of those marrying directly is consistent with ABS data indicating that the proportion of people cohabitating prior to marriage has increased from 67.2 % in the late 1990s to 78.2 % in 2011 (ABS 2007, 2012c). The middle column of the table indicates that the majority (69 %) of the relationships that ended in HILDA between 1995 and 2010 were cohabiting only relationships.

Table 5.2 Relationship type and outcome (column per cent) for relationships formed between 1995 and 2010 (HILDA 2001–2010a)
Full size table

While this information provides us with a summary of relationship dissolution across these relationship groups, there are a number of limitations to this approach when examining relationship dissolution. Relationship dissolution is a time-dependent event (Heaton et al. 1985; Heaton 1991; Heaton and Call 1995), where the risk of dissolution may increase or decrease over the duration of the relationship. To better understand the nature and extent of differences in the time dependency of relationship dissolution for these relationship types, we use retrospective and prospective relationship information from the first 10 waves of HILDA.

We examine relationship survival and the hazards of dissolution over the first 15 years of the relationship, restricting our analyses to relationships formed after 1995. First we examine the survival function, which tells us the proportion of respondents surviving relationship breakdown at each year. Figure 5.6 plots the survival function for separation from first marriages, cohabitating relationships and first marriages preceded by cohabitation in the sample. The 15-year survival of first marriages formed since 1995 in our sample is 92.6 %, and the first 5 years of marriage for this group are very stable. This differs from previously published Australian research on marriage dissolution (see Hewitt et al. 2005: 173), which indicated that approximately 82.8 % of marriages survived the first 15 years and that many marriages ended within the first 5 years. However, the previous study included marriages that had formed in the 1930s and 1940s when divorce and cohabitation were rare, as well as marriages that were formed in the 1960s and 1970s when cohabitation was relatively rare, but divorce was increasing. Thus the earlier figures represented an average over all marriages irrespective of the year of marriage. The results here suggest that for more recent marriages formed since 1995, early marriage is relatively stable.

Fig. 5.6
figure 6

Fifteen-year survival of de facto cohabitations and first marriages formed after 1995 (HILDA 2001–2010, see Appendix 5.1)

Full size image

The 15-year survival of marriages preceded by cohabitation is marginally higher at 93.1 % than of those not preceded by cohabitation. Finally, Fig. 5.6 shows that cohabiting relationships that have not proceeded to marriage have much lower survival rates at all relationship durations, with very small numbers of cohabiting relationships reaching 15 years duration (numbers not shown) and only 64.7 % of these relationships surviving at 15 years duration.

An alternative way of looking at the timing of relationship dissolution is the hazard rate. The hazard rate represents the likelihood of experiencing relationship dissolution given that the relationship did not end in the previous year (Yamaguchi 1991: 9). In other words the hazard indicates the proportion of relationships that ended in separation for each time interval, given that the respondent was still in their relationship at the previous time interval. In Fig. 5.7, the hazards of relationship dissolution for each group are presented. The graph shows that the hazards of relationship dissolution are similar for those who are married with or without a period of cohabitation and are relatively low. There is an overall trend of increasing hazard of dissolution over time, with marriages preceded by cohabitation having a slightly elevated risk of dissolution over the 12 years of relationship duration. However, additional analysis indicates that there were no significant differences in the hazards of relationship dissolution for direct marriages and marriages preceded by cohabitation. This finding is consistent with recent research that suggests the increased risk of divorce for those who cohabited before marriage has diminished or disappeared for younger cohorts (Klijzing 1992; Schoen 1992; de Vaus et al. 2005; Hewitt and de Vaus 2009).

Fig. 5.7
figure 7

Hazards of relationship dissolution for cohabiting and first marriages formed after 1995, HILDA 2001–2010 (see Appendix 5.1)

Full size image

Figure 5.7 also shows that cohabiting relationships that do not proceed to marriage have a higher likelihood of dissolution at all relationship durations. The U-shaped pattern of likelihood of dissolution from cohabitations is very different from the gradual increase for those who married (either with or without a period of cohabitation). The U-shape distribution indicates the likelihood of dissolution in the first couple of years of a cohabiting relationship is very high, then stabilises once the relationship reaches 3 years in duration and increases quite dramatically again after 10 years. It should be noted that the number of cohabiting relationships at 10 years was relatively small and therefore the hazard estimates are less reliable. Therefore these results for cohabitations of longer durations should be treated with some caution. We restrict Fig. 5.7 to 12 years’ relationship duration.

These patterns of relationship dissolution for marital and cohabiting relationships formed since 1995 in HILDA are interesting for their departure from patterns recorded by previous generations and the ways in which they reflect more recent trends in relationship formation. Many couples use cohabitation as a ‘trial’ marriage (Seltzer 2000; Manning and Smock 2002; Qu et al. 2009). It appears that many of the marriages that might once have ended in the first few years of marriage may have been replaced by cohabiting relationships. This has resulted in a lower risk of divorce early in marriage for more recent marriage cohorts than in previous marriage cohorts. As in previous generations, Australians continue to form relationships that are relatively unstable in their early years, but in more recent generations those relationships are less likely to be legalised with marriage.

5.5 Why Is Cohabitation Less Stable?

With the increase in cohabitation as either a prelude or alternative to marriage, a large and growing body of work comparing and contrasting cohabitation and marriage has emerged. Understanding differences between couples that choose to cohabit or marry is important for explaining why cohabiting relationships tend to be less stable. Arguably, the most prominent recent explanation for differences between cohabiting and married couples is commitment theory. According to commitment theory the motivation for cohabiting rather than marriage is based on a lack of personal dedication to a partner and constraint commitment (Stanley et al. 2004).

Personal dedication refers to interpersonal commitment associated with a strong desire for the relationship to last into the future (Rhoades et al. 2011). Some research indicates that cohabiters as a group tend to value individual freedom more than their married counterparts (Axinn and Thornton 1992; Thompson and Collela 1992). Other research finds that cohabiters tend to have lower levels of relationship commitment and fewer moral constraints to stay in their relationship than married couples (Nock 1995; Brown and Booth 1996). These differences suggest that cohabiters have lower levels of interpersonal commitment to their partner and to being in a relationship than married people.

Constraint commitment refers to the costs of ending or leaving a relationship including financial constraints (i.e. access to income, home ownership), social pressure and concerns for children (Stanley et al. 2006: 503). Overall, cohabiting relationships have lower levels of constraint commitment, in that partners are more likely to keep their money separate (Vogler et al. 2006), less likely to own a house together (Mulder and Wagner 2001; Baxter and McDonald 2004) and less likely to have children in the relationship (ABS 2012a); although it should be noted that a significant number of children are now being born to couples who are not married.

Interestingly, this argument also highlights the fact that the transition from cohabitation to marriage may not necessarily indicate a greater level of interpersonal commitment. Rather, once involved in a longer term cohabiting relationship, the costs of leaving may be a more important determinant of the stability of the relationship or the transition to marriage than personal dedication to one’s partner (Stanley et al. 2006). Some long-term cohabiters with high levels of constraint commitment, such as children or co-ownership of a house, resemble married couples. For example, Willets (2006) finds that long term cohabiting relationships with high levels of constraint commitment have similar levels of relationship quality to marital relationships. However, long-term cohabiting relationships of a highly committed nature are still relatively rare (Kiernan 2002; Seltzer 2004; Qu et al. 2009).

This research suggests that, overall, cohabiting couples have lower levels of dedication to the relationship with their partner and fewer structural constraints to ending the relationship when compared to married couples. These factors are likely to strongly influence decisions that partners make about whether to remain in the relationship or to end the relationship. Using the Generations and Gender Survey (see Sect. 7.3.1) to compare and contrast cohabiting and married couples across eight European countries (Bulgaria, France, Germany, Hungary, Norway, Romania, Russia, and The Netherlands), Wiik et al. (2012) find that cohabiters are more likely to have plans to break-up than married couples.

5.6 The Consequences of Relationship Dissolution

Of primary concern to researchers and policy makers are the consequences of relationship dissolution for individuals, families and children. The growth in marriage breakdown is significant because there are substantial short and medium term, social, psychological and economic costs for spouses and children (Amato 2000), as well as very significant costs to the national economy (Australian Parliament House of Representatives Standing Committee on Legal and Constitutional Affairs 1998). Marital dissolution is not only an emotionally stressful event for individuals, but results in changes in many areas of life including employment, household income, and household composition (Wood et al. 2007). Despite the dramatic rise in cohabitation, and the instability of cohabiting unions, few studies have investigated the consequences of relationship dissolution for those in cohabiting compared to marital unions.

While cohabitation seems to offer some similar advantages to marriage, the important differences outlined in the previous section suggest that when relationships end we might expect that separating from cohabitation may have less impact on people’s lives than separating from marriage. Two outcomes that have been investigated are the consequences of relationship dissolution for income and health and wellbeing.

5.6.1 Income

Previous research in Australia (Smyth and Weston 2000), the United States (Bianchi et al. 1999) and Europe (Poortman 2002; Uunk 2004; Aassve et al. 2007; Andreß and Bröckel 2007) finds that men do better financially after separation than women. Typically after marital separation men’s household income remains relatively stable and women’s decreases (Andreß and Bröckel 2007). These differences are likely due to gender differences in changes in household composition combined with gender differences in earnings. For men, the average number of people in their household diminishes after relationship dissolution as they are less likely to have primary responsibility for the care of children, but their household income does not decline dramatically as men typically contribute the majority share to household income before the relationship ends (Bianchi et al. 1999; Smyth and Weston 2000). In contrast, women’s household size decreases less after relationship dissolution because they are more likely to have greater care responsibilities for children, but their household income decreases more dramatically as they tend to contribute less to household income. We know little about what happens when cohabiting relationships break down.

To compare and contrast the consequences of relationship dissolution from cohabiting and marital relationships we use a measure of household income that includes any tax transfers, government benefits, private transfers (such as the payment of child support) and income from salary, wages, and business. We use this measure as it captures the total income available in the household for consumption or savings. We also equivalise our income measure because the financial needs of households change with each additional member, and equivalised income better captures people’s actual standard of living as it takes household composition into account. Due to large gender variations in the household composition of spouses after separation the most appropriate measure of household income is equivalised household income. In the HILDA Waves 1–10 sample we found that women who separated from marriage had the largest average household size after separation (2.4 persons) and cohabiting men who separated had the smallest (1.01 persons) while women who separated from cohabiting relationships (1.7 persons) and men who separated from marriage (1.5 persons) were in between.

In Fig. 5.8, we show the predicted equivalised household income for men and women after separation from marriage and cohabitation. We plot equivalised household income at three time points: in the year prior to relationship dissolution; in the year of dissolution and in the year after dissolution. In the left panel we present the predicted equivalised income for men. The graph shows that men’s equivalised household income increased after separation. There were no differences in the household income of men who were married compared to men who were cohabiting before or after relationship dissolution. The picture for women is quite different. Not surprisingly women in cohabiting and marital relationships have similar equivalised household incomes to men. After relationship dissolution, however, cohabiting women’s equivalised household income increased in a similar pattern to that for men. In contrast, equivalised household income for women who separated from marital relationships remained stable and was not significantly different from equivalised household income when they were married.

Fig. 5.8
figure 8

Equivalised annual household income after separation from cohabitation and marriage, by gender (HILDA 2001–2010) (Models control for union duration, age, employment status and highest level of education. See Appendix for more information on modelling approach used (Hewitt and Poortman 2010))

Full size image

The main finding that cohabiting women have a stronger financial position after separation than married women is consistent with previous research in two main ways. First, cohabiting women tend to contribute a higher share of household income during the relationship than married women (Kalmijn et al. 2007). Second, couples in cohabiting relationships are less likely to have children than couples in marital relationships (Hewitt et al. 2010), and therefore cohabiting women are less likely to have dependent children to care for after separation. Even though a significant proportion of children are currently born in cohabiting relationships, the majority is born within marital relationships. Together these two factors likely contribute to the stronger financial position of cohabiting women than married women after separation.

5.6.2 Health

It is well documented that intimate relationships are important to health (Carr and Springer 2010). A large number of studies spanning decades show that, compared to being unmarried, being married is associated with better physical and mental health and well being (Gove and Shin 1989; Wade and Pevalin 2004; Williams and Umberson 2004; Willitts et al. 2004; Strohschein et al. 2005; Bennett 2006; Zhang and Hayward 2006) and lower rates of mortality (Grant et al. 1995; Nagato et al. 2003; Brockman and Klein 2004; Dupre et al. 2009). A handful of studies have compared the health profiles of people in marital and cohabiting relationships, and the findings of these studies are mixed. In general no differences are found in the physical and mental health of cohabiting versus married people (Horwitz and White 1998; Wu et al. 2003); if differences are found cohabiters tend to have poorer health than married couples (Brown 2000).

People who are separated, divorced or widowed have worse health than their partnered or never-married counterparts (Bierman et al. 2006; Wood et al. 2007), which suggests that marital loss may be particularly consequential for health. Far fewer studies have investigated what happens to health when cohabiting relationships end. While cohabitation seems to offer some similar health advantages to marriage, there are some important differences in the experiences and conduct of cohabiting relationships that may indicate differences in the health consequences of ending such relationships; although the scant evidence to date suggests that there are no differences in the health consequences of separation for married and cohabiting couples (Wu et al. 2003).

We examine the consequences of relationship dissolution from cohabiting and marital relationships for physical and mental health. Figure 5.9 shows the physical health consequences of separation for men and women from marital and cohabiting relationships. For men, there were no physical health differences by union type or stability, although the graph suggests a decline in health for cohabiting men leading up to separation, followed by a return to previous health levels by one year after separation. For women, those who separate from cohabiting or marital relationships have a small improvement in their physical health (although physical health scores are similar to those recorded before the relationship ended).

Fig. 5.9
figure 9

Physical health (SF-36) after separation from cohabitation and marriage, by gender (HILDA 2001–2010) (Models control for relationship duration, age, number of children under 18 in the household 50 % or more of the time, household income, employment status, highest level of education and health status at the previous wave. See Appendix 5.1 for more information on modelling approach used (Hewitt et al. 2012))

Full size image

In Fig. 5.10, we show the mental health consequences of relationship dissolution for men and women in cohabitation and marital relationships. These graphs show similar patterns for men and women. First, those who experienced separation from a relationship had poorer mental health before and after the transition. This is consistent with previous research which suggests that prior to a relationship ending men and women experience low levels of relationship quality which negatively affect mental health (Kalmijn and Monden 2006). Second, the results indicate that the mental health consequences of separation from marriage are significantly worse than for separation from cohabitation. Finally, we see that within a year or two after separation, mental health has recovered to levels similar to those recorded prior to separation, and for women are slightly higher than prior to separation. Thus the consequences of relationship dissolution for mental health also appear to be short-lived.

Fig. 5.10
figure 10

Mental health (SF-36) after separation from cohabitation and marriage, by gender (HILDA 2001–2010) (Models control for relationship duration, age, number of children under 18 in the household 50 % or more of the time, household income, employment status, highest level of education and health status at the previous wave. See Appendix 5.1 for more information on modelling approach used (Hewitt et al. 2012))

Full size image

These results indicate that relationship dissolution has a stronger and more negative association with mental health, though not long-lasting, than for physical health. There are also clear negative mental health implications for those separating from marriage compared to those separating from cohabiting relationships, and these findings are consistent for men and women. There are, however, some important gender differences for household income. For men, equivalised household income improves and there are no differences in the consequences of relationship dissolution for men who are cohabiting or married. In contrast, married women have a much lower equivalised household income after separation than cohabiting women after separation. On balance, our results suggest that separation from cohabitation has far less severe consequences for finances and health than separation from marriage.

5.7 Discussion

The goal of this chapter was to illustrate continuity and change in the nature of relationship dissolution in Australia and to provide insights into recent trends and outcomes. Over the last century in Australia divorce has gone from being virtually non-existent to becoming a common feature of family life by the mid-1970s (Hewitt et al. 2005). While this sparked a moral panic about a crisis in ‘the family’ late last century, there is little evidence that such a crisis has occurred. Since the early 1980s the rate of divorce has slowed, stabilised and from the year 2000 is showing a slight decline. In addition, the nature and characteristics of divorcing couples continue to change, with increases in the median age at divorce and time to divorce and decreases in the proportion of divorces involving children. These trends are consistent with the stabilisation of the overall rates of divorce and suggest that fewer children are being affected by divorce now and in the future. However, marriage has also transformed and one factor that may partially explain these trends in the legal dissolution of marital relationships is the increasing number of cohabiting relationships that are not captured in official statistics. This suggests that some unstable marriages have been replaced by cohabitations.

Using data from the HILDA survey we compared and contrasted the stability of married and cohabiting relationships. Consistent with broader trends shown by official statistics, which indicate that marriage has stabilised, we find that marriage, and in particular early marriage, is relatively stable. In contrast, our examination of cohabiting relationships provides good evidence that Australians are not necessarily experiencing relationship dissolution at lower rates than in the past. In fact, if anything, they are possibly experiencing higher rates of overall relationship dissolution, but in cohabiting relationships rather than marriage.

It is well documented that on average the nature and circumstances of cohabiting relationships differ from those of marriages (Stanley et al. 2004). These differences, such as lower average levels of emotional as well as structural commitment amongst cohabiters, provide strong insights into why cohabiting relationships are less stable. These differences also suggest that in the case of relationship dissolution the consequences for cohabiters may be less severe. However, few studies have tested this idea. In this chapter we contrasted the financial and health consequences of relationship dissolution for cohabiters compared to those who are married. We find that while relationship dissolution does tend to have a negative impact on financial and mental well being, the consequences are stronger for married people.

These results on the consequences of dissolution for cohabiters and married respondents in Australia are not entirely consistent with previous research in the field. We find significant mental health differences for cohabiting and married respondents who experience relationship dissolution, but a Canadian study found no significant differences in the mental health consequences of relationship dissolution for married or cohabiting respondents (Wu and Hart 2002). We also find that married women fare significantly worse financially than cohabiting women after separation, even though the financial position of married women after separation relative to their position prior to separation is not significantly worse. Previous Australian research suggests that this is largely due to the flow of government transfers into separated women’s households (Hewitt and Poortman 2010). However, a US study using the Longitudinal Survey of Youth concludes that women whose cohabiting relationships end have similar financial standing as previously married women (Avellar and Smock 2005).

The overall picture of relationship dissolution in the Australian context, provided by this chapter, is relatively positive. Officially, the trends suggest more stable and potentially lower divorce rates in the future. Even though Australians are experiencing high rates of relationship dissolution from cohabiting unions, the evidence presented here suggests that the emotional, social and financial effects of separation from cohabiting relationships are less severe than they are from marriages. Most couples whose relationships end are able to progress with their lives and those with children often renegotiate their post-separation relationship in positive ways (Funder 1996; Smart and Neale 1999; Smart 2000). Nevertheless in the short term there are major social, emotional and financial implications for both men and women experiencing relationship dissolution from cohabitation and marriage (Amato 2000). It is thus important to maintain social and financial supports for Australian couples who have experienced relationship dissolution, whether from cohabitation or marriage, and to continue to monitor trends and outcomes given the rapid rate of change in patterns of family formation and dissolution.