Domestic divisions of work and couple dissolution in the UK

Abstract

This paper analyses the effect of the domestic division of work on the risk of dissolution of married and cohabiting couples in the UK. The division of work is usually categorised in terms of equality or traditionalism, depending on whether housework is done mostly by women or not. Equity has received less attention in the literature. We propose an exploratory measure to include both equity and equality/traditionalism in our operationalisation of divisions of work, estimating the total number of hours that each partner spends doing any type of work (paid and unpaid). We hypothesise that the risk of dissolution will increase for cohabitors if they depart from an equitable and egalitarian model, whereas marriages will be protected by more traditional arrangements. We use an event-history approach to analyse data from 16 waves of the British Household Panel, from 1992 to 2008. Our results support the hypotheses only partially and suggest that for cohabitors both equality and equity are important for stability.

Keywords

Cohabitation domestic work equity gender separation

Women's massive entrance into the labour market has reshaped family formation patterns all over the world, and nowadays a great proportion of couples follow the dual-earner model, where both partners are active in the labour market (Fahlén, 2016). The literature has researched extensively the potential effects of this change in women's roles on family outcomes, and more specifically on marriage stability and divorce (Esping-Andersen and Billari, 2015). The findings of this literature are mixed: some studies expect wives' employment to have a positive effect on stability (Oppenheimer, 1997; Sayer and Bianchi, 2000), but other studies find that the risk of marital dissolution decreases when the husband is the main earner (Kalmijn, 2007), or when wives are in paid work or work for long hours (see Özcan and Breen, 2012, for a review).

But the diffusion of dual earner couples does not only affect the public sphere, it also entails changes in the traditional division of unpaid work. Even if domestic work continues to be gendered, there have been recent changes towards less traditional arrangements, with women decreasing their participation in domestic chores and men slightly increasing theirs (Altintas and Sullivan, 2016). Conflicts about housework are also likely to influence marital satisfaction, and a smaller number of studies have considered the role of unpaid work in family dissolution, showing that women's dissatisfaction with the division of housework increases the risk of separation (Frisco and Williams, 2003). Recent research has suggested that it is not the amount of unpaid work, but the coherence between partners' ideal divisions (egalitarian or traditional) and the actual division that influences dissolution (Oláh and Gähler, 2014).

Although marriage remains the most popular type of union, cohabitation is becoming more widespread in industrialised countries (Liefbroer and Dourleijn, 2006). However, the literature has shown that marriage and cohabitation remain distinct living arrangements in most countries (Heuveline and Timberlake, 2004). One important difference between the two is the distribution of paid and unpaid work: cohabitors allocate paid and unpaid work more equally between partners (Baxter, 2005; Domínguez-Folgueras, 2012). Another important difference concerns the risk of dissolution, which is higher in cohabiting relationships compared to married unions (Beaujouan and Bhrolcháin, 2011; Jalovaara, 2013). In this paper, we will investigate whether the influence of the division of work on separation differs by union type.

When the division of work is objective, researchers usually describe it as traditional – when women do most of the domestic work – or egalitarian – if the work is equally divided. In this paper we argue that equity is another dimension to be taken into consideration (DeMaris, 2007), and we propose an operationalisation of the division of work that combines traditionalism–egalitarianism with an exploratory measure of equity, by estimating the total number of hours that each partner spends doing any type of work (paid or unpaid). We use 16 waves of the British Household Panel Survey (1992–2008) and also analyse married and cohabiting unions separately in order to investigate whether the division of work is equally important in both types of union.

Background

Divisions of work and union stability

Over the last decades there has been an increase in women's participation in the labour market. Women's new role in the public sphere constitutes an important social change that can affect family life, and an extensive line of research has studied the impact of female employment on union stability, although the findings are not fully conclusive (for a review, see Özcan and Breen, 2012). Many studies have found that the risk of dissolution increases when women work for pay and especially when they work for long hours (Poortman and Kalmijn, 2002; South, 2001). A comparative study of European countries revealed that traditional specialisation arrangements in marriages – where the husband is the main income earner and the wife is responsible for housework – decreased the risk of marital dissolution (Kalmijn, 2007). Causal relations between paid work and divorce are however difficult to establish; working women may have less stable partnerships because the traditional breadwinner model is still the norm for marriages, or because they are able to get divorced if they are not satisfied with their marriage (Sayer et al., 2011; Schoen et al., 2002). Women might indeed enrol in paid work in order to gain the economic independence that would allow them to leave their partners (Özcan and Breen, 2012). Other studies, instead, have hypothesised that women's paid work could be positive for the family as it produces more economic resources that can compensate for other potentially negative effects. As women's labour income increases, couples would be better protected against unexpected events such as illness or unemployment, and this should strengthen the stability of marriages (Oppenheimer, 1997; Sayer and Bianchi, 2000).

Another important consequence of women's increasing paid work is the decline in gendered specialisation in unpaid work (Fahlén, 2016). Although it must be noted that the gendered division of work persists, time use data have consistently shown a trend towards more egalitarian divisions of domestic work in most countries and a positive association between women's paid work and a more egalitarian division of chores (Altintas and Sullivan, 2016). The higher the wives' absolute resources, the more autonomous and independent the wives become: Gupta (2007) finds that when wives' economic resources are high, they can better negotiate parity in the allocation of household labour compared to less autonomous wives. The literature has also considered whether the division of unpaid domestic work influences marital quality and stability, and again in this case the evidence is mixed and results vary across countries. For the USA, Amato et al. (2003) have shown that men's participation in domestic work increased marital happiness for women but had the reverse effect on men, but Wilcox and Nock (2006) did not find this effect to be significant. For the same country, Bellani et al. (2018) find that, if the division of paid and unpaid work underbenefits women, it has a destabilising effect for them, especially if both partners work similar hours; however, the authors did not find the same effect for Germany. In Sweden, Ruppaner et al. (2018) found that men's involvement in domestic work was associated with lower risks of dissolution. For the UK, Chan and Halpin (2002) did not find an effect on dissolution, except for women with high wages, for whom higher loads of unpaid work increased the risk of divorce.

Marriage and cohabitation

Research on the division of work and union stability has often analysed married unions. But marriage is not anymore the only type of intimate, long-term, co-residential relationship: in recent times cohabitation has become widespread and socially accepted as an alternative living arrangement (Liefbroer and Dourleijn, 2006). There is a rich literature investigating the differences between cohabitation and married couples, which has described several areas of divergence, two of which are of central interest for this paper. First, cohabitation is more unstable than marriage (Kalmijn, 2007; Lewis, 2001; Liefbroer and Dourleijn, 2006), and second, cohabitors tend to be more egalitarian than marrieds concerning the amount and type of housework performed by each partner (Batalova and Cohen, 2002; Bianchi et al., 2014; Domínguez-Folgueras, 2012). These differences between married and cohabiting unions can be attributed to the different meaning of these relationships as well as to the associated selection effects.

Because cohabitation and marriage are two different living arrangements, they might select different individuals, or be associated with different moments in the life cycle: cohabitors tend to be younger, less religious and have fewer children than marrieds. Cohabitation – compared with marriage – has been described also as an incomplete institution in many countries (Heuveline and Timberlake, 2004), where norms are less clearly defined, whereas spousal roles in marriages are clear and more traditional (Brines, 1994). Indeed, as mentioned, cohabitors in different countries seem to favour more egalitarian divisions of work (Batalova and Cohen, 2002; Bianchi et al., 2014; Domínguez-Folgueras, 2012). This leads us to expect a different relationship between the division of work and union stability in married and cohabiting unions. This hypothesis has been studied by Brines and Joyner (1999) for the US, where they found strong evidence that equal-power arrangements lowered the risk of cohabitation dissolution. In a similar study on the Netherlands, Kalmijn et al. (2007) analysed income dynamics, and showed that movements away from equality where associated with higher risks of dissolution in cohabiting unions, whereas specialisation seemed to protect marriages.

Differences in the division of work between marrieds and cohabitors might be related to differences in gender values: marrieds could favour more traditional arrangements and cohabitors could prefer more egalitarian arrangements. This has been an assumption in the literature, although empirical analyses on domestic work do not often include information on gender values. Some researchers have been able to test the effect of consistency between gender attitudes and the division of labour on union stability. Oláh and Gähler (2014) found that the risk of dissolution was higher in couples where there was inconsistency between attitudes – egalitarian or traditional – and the division of work in Sweden. In a similar vein, Schober (2013) has found that inequality in the division of housework reduced stability for women with egalitarian values in the UK.

Equality and equity

In addition to the factual division of work, the perception about the division might also play a role in marital satisfaction and dissolution. It has been shown that, for married women who do the majority of the household work, a high proportion of them are satisfied and consider it fair (Fuwa and Tsutsui, 2010; Nakamura and Akiyoshi, 2015). Nevertheless, wives declare themselves to be more satisfied when the division of housework is fairer and husbands increase their participation in the domestic sphere (Amato et al., 2003). Cohabitors, on the other hand, seem to be more sensitive than marrieds towards the fairness of the division in the US (Baxter et al., 2010). These results illustrate that fairness and equity considerations may also play a role in couple's satisfaction.

According to equity theorists, considerations of distributive justice are essential to any human relationship, and that would also be the case for marriage (DeMaris, 2007). To our knowledge, the subjective dimensions of equity have been analysed in the above mentioned literature on fairness, but are not often integrated in factual measures of work performance, with few exceptions (Bellani et al., 2018; DeMaris, 2007, 2010). So far, the literature has analysed either subjective perceptions about the division of work, in terms of fairness, or the objective division – how much each partner actually does. The factual division is most often described as being ‘traditional’ – when the woman does most of the domestic work – or ‘egalitarian’ – when domestic work is shared. But this categorisation groups together situations that can be very heterogeneous in terms of equity of workloads. For instance, the pure male breadwinner model is traditional, but we can consider it equitable, in the sense that the total workload is shared. In contrast, the situation that Hochschild and Machung (1991) described as the ‘double shift’ – both partners work for pay and the woman does most of the domestic work – would also be traditional, but inequitable, because women assume a higher workload than men.

In this paper we contribute to the literature by integrating dimensions of both equity and equality into our measures of the division of work. We hypothesise that the division of work can have consequences for union stability along those two dimensions: traditionalism and equity, and that both have to be taken into account. In a traditional division of work there is specialisation based on gender: women will do most domestic chores and men do most paid work. In a non-traditional division we find no gendered specialisation, both members of the couple will divide things more equally, or traditional roles can be reversed and men can be in charge of the domestic sphere. Regarding equity, in this paper we will consider that the division of work is equitable if both partners are contributing similar amounts of time to the household total load. The male breadwinner model is equitable if both members of the couple have similar workloads, and the same thing happens with the dual-earner model. However, if both members of the couple are active in the labour market but one of them does most of the ‘second shift’ at home, this arrangement is not equitable, as one of them works much more than the other. This inequitable arrangement could be traditional at the same time if it is the woman who does the second shift.

Regarding the differences between cohabitors and marrieds, the literature has shown that movements away from an egalitarian division increase the risk of dissolution for cohabitation, whereas movements away from a traditional arrangement increase the risk of dissolution for marriages (Kalmijn et al., 2007), and this is what we expect to find (hypothesis 1). The equity of the division will be important for all couples, but we hypothesise that it could be more so for cohabitors than for marrieds (hypothesis 2).

Context of the study

In this paper we will analyse data from the UK. As in most Western countries, British couples postpone marriage and increasingly opt for cohabitation before getting married or as an alternative to marriage (Wilson and Smallwood, 2007). Indeed, according to the Office for National Statistics (ONS) the number of cohabiting couples doubled between 1996 and 2012 in the UK: from 3 million to 5.9 million persons. In the UK, there is not yet a ‘common law marriage’ that provides similar legal status to marriage and cohabitation. This might account for some of the different levels of union dissolution in married and cohabiting couples. Whereas the risk of marriage dissolution after 5 years is around 8%, and close to 21% after 10 years, about one in three cohabiting relationships end by the 5th year, and about 40% by the 10th year (ONS, 2012, 2013).

Furthermore, cohabitors and marrieds have different socioeconomic profiles. Married individuals tend to be older, more educated and more traditional than cohabitors, who tend to be younger, less educated and more often classified as working class (Beaujouan and Bhrolcháin, 2011; Kiernan, 2004). Smart and Stevens (2000) show that British women consider cohabitation as a way to avoid either single motherhood or marrying men they are not certain they want to marry. Although cohabitors in the UK see their living arrangement either as a prelude to or substitute for marriage (Berrington, 2001), cohabitation is increasingly seen as an alternative to marriage among the working class and those enrolled in education (Kiernan, 2004).

Methods

Data

This study uses panel data from 16 waves of the British Household Panel Survey (BHPS), between 1992 and 2008. The BHPS is an annual household-based panel that includes data on a broad range of social issues such as household composition, income, health, socio-economic values, education, housework, and labour market behaviour. Data on housework were first collected in 1992, so this is the first wave included in our analysis. Originally, the BHPS covered 5538 households and 9912 individuals drawn from 250 areas of Great Britain. The individuals included in the first wave were considered original sample members and attempts were made to re-interview them annually even if they moved to another household. As for the quality of the data, the BHPS has high levels of response rates (e.g. over 85% for individual interviews in almost every wave) and low levels of item non-response. The mean level for item non-response at the individual level in the first 13 waves is about 1.7%, and about 3.2% at the household level (Lynn, 2006).

In this article, we follow all the heterosexual couples that started a co-residential union within the period observed (1992–2008) and we observe them once per year. Couples that started their relationship before the panel were not included in the sample because we have no idea about their work distribution patterns and they are therefore not comparable with couples that we can observe since the start of the co-residential union. This choice has important consequences for the analysis because we are not analysing couples that have been in long-term co-residential relationships, which could follow a specific pattern.

Variables

The dependent variable is separation, understood as the moment when the couple stops living together. The variable takes the value 0 if the partners were living together in any given year, and value 1 if they separated during that year. Cohabiting individuals who subsequently marry are right censored in the cohabitation sample on the year of the marriage, and hence counted as married thereafter, but we include a control to identify them in the models for marriage. After cleaning the data 220 married couples and 425 cohabiting couples were excluded because they had missing information for all the independent variables of interest. The final sample to be analysed includes 1518 marriages and 1781 cohabiting couples. During the observation period we recorded 438 separations of cohabitors and 156 separations among marriages.

The main independent variable in the analysis is the division of paid and unpaid work between partners, lagged one year. In order to construct that variable we take into account the answers provided to three questions in the survey. To quantify unpaid work we used the question: ‘About how many hours do you spend on housework in an average week, such as time spent cooking, cleaning and doing the laundry?’ To account for paid work we used the question: ‘Thinking about your (main) job, how many hours, excluding overtime and meal breaks, are you expected to work in a normal week?’, and combined it with a question on overtime: ‘How many hours overtime do you usually work in a normal week?’ These questions were answered individually by each member of the couple and from the responses we created two separate variables for unpaid and paid work, for women and men. This allows us also to calculate the gap between members of the couple by subtracting the amounts of time declared by both partners. We have also used the distribution of the resulting gaps in order to determine whether the gaps were significant enough. We considered that one partner did more unpaid or paid work than the other when the difference between the time spent on either type of work was more than half a standard deviation away from the average gap between partners (we have tried also one standard deviation but results did not change). In this sense, we measure whether the woman does significantly more paid or unpaid work than the man, or vice versa. Paid and unpaid work contribute differently to the wellbeing of the family, and they also entail different social rewards in terms of resources and status, but both are necessary and demand a significant time investment. Our variable takes both types of work as comparable in the latter sense.

We then combine the resulting gaps in paid and unpaid work to construct our dependent variable, which has four categories. In the first one we find women who spend more time on domestic work than their partners, while men spend more time on paid work than women. We label these couples ‘traditional equitable’ because there is a traditional division of work but the workloads are balanced. In the second category we have couples where both are working similar amounts of time in the labour market, but women are doing significantly more domestic work than men. We also include in here women who spend more time than their partners in paid work and then do a similar amount of unpaid work. We label this category ‘traditional inequitable’ because women are doing a second shift. In the third category, both partners dedicate similar amounts of time to paid and unpaid work. We have also included in this category couples in which the man does more housework than his partner, while the woman does more paid work than he does. We label this category ‘non-traditional equitable’ and use it as the reference category in the regression analysis. Finally, in the fourth category we have couples who reverse the traditional ‘second shift’ model, in which both partners spend similar amounts of time in paid work and where the man does more housework than the woman. We also include in this category men who spend more time in paid work than their partners and do at least similar amounts of unpaid work. We label this category as ‘non-traditional inequitable’.

Following Oláh and Gähler (2014) and Schober (2013), we also control for gender attitudes. To do so, we considered the responses to one of the statements in the BHPS survey related to the traditionalism of men and women. Most studies (e.g. Chan and Halpin, 2002; Schober, 2013) have used six of the BHPS statements to account for the gender role attitudes of individuals. However, the only statement which addresses the way in which housework and paid work should be distributed is the statement ‘The husband should earn and the wife should stay at home’, and we considered it as the most relevant for the objective of this paper. We have also run the models using an index, but results did not change. Statements on traditionalism were only asked every other year, so we treated the variable as unchanged between subsequent observations. We considered individuals had ‘traditional’ attitudes towards the division of work if they agreed or strongly agreed with the statement, and ‘non-traditional’ attitudes if they strongly disagreed, disagreed, or didn't agree nor disagree. To be sure, agreement with an affirmation is not the most accurate proxy for gender attitudes, but it is the only variable available in the survey and has been used in previous research (Schober, 2013; Oláh and Gähler, 2014). We have run interactions between gender attitudes and the division of work, to test whether it was the actual division or its consistency with values that mattered, but results were not significant, so we have decided not to include the interaction in the models presented here.

Finally, we take into account other control variables that have been shown to have an impact on union disruption in the literature. Some variables are expected to reduce the likelihood of separation: having a mortgage (yes/no), having college education (yes/no), duration of the co-residential union (in years), age of partners at the start of the relationship (in years), and having a child under four years of age living in the household (yes/no). Having children is likely to increase the amount of domestic work to be done, but also imposes other constraints on parents and has been shown to lead to more traditional arrangements (Grunow and Evertsson, 2016; Zabel and Heintz-Martin, 2013).

Other variables have been shown to increase the risk of dissolution: difference in age between the partners (in years), experience of previous cohabitation (for marriages), and number of children. Regarding income, Kalmijn et al. (2007) have shown that higher levels of household income protect marital stability, whereas women's higher income shares can have a negative effect on marital stability for marrieds, as it deviates from the traditional model. We therefore include an absolute measure for the household total income and also a relative categorical variable for women's contribution to the household income, in three categories (below 0.4, between 0.4 and 0.6, above 0.6).

It must be noted that our paper focuses on routine domestic work and we do not include childcare in our analysis. The explanation for this choice is both theoretical and pragmatic. There are important differences between childcare and routine domestic work: childcare is not always an activity to be avoided or distributed between the members of the couples as it might occur with routine housework. At least part of the childcare is considered as something to be shared and enjoyed together (Hallberg and Klevmarken, 2003). Also, because we are interested in comparing marriage and cohabitation, marrieds are more likely to have children than cohabitors (60% of marrieds versus 40% of cohabitors had children living with them). Finally, child care is only relevant for couples with children, and especially so for couples with small children (Zabel and Heinz-Martin, 2013); we have run models only for this subsample in order to check the effects of childcare, but the number of separations was very small (60 cases in all) and results were not significant. In any case, the division of childcare, its articulation with domestic work, and the possible effects on separation is an interesting extension to be addressed in further research.

Method

We use event history modelling (Yamaguchi, 1991) to study how the distribution of work within the couple affects the risk of dissolution between 1992 and 2008. We use a discrete-time logistic model because observations are annual, and we run separate analyses for cohabiting and married individuals because the literature has shown different effects of several variables in both unions. Observations start when the partners start living together and are censored at the end of the panel, when they separate, or if they drop out. Cohabitors who marry are censored at marriage. The dependent variable is dichotomous (0 = continue together, 1 = dissolution), making the duration of episodes follow a log-logistic distribution. Therefore, once the data are organised into a couple-year form, the results are estimated from the following equation

\begin{matrix} log [\frac{P_{it}}{1 - P_{it}}] = α + β_{1} (distribution of work)_{it} \\ + β_{2} t_{it} + β_{k} X_{i (t)} + β_{l} X_{i (t - 1)} + ɛ_{it} \end{matrix}

where

P_{it}

is the conditional probability of partnership _i experiencing a separation in year _t (from 1 to 16), given that they have not yet dissolved or been censored prior to wave _t. The distribution of work specifies the way in which the partners allocate the time spent on paid and unpaid work. t is relationship duration, _k refers to time-constant socio-demographic variables measured when the couple enters the survey (e.g. age at the beginning of the relationship), and _l are the time-varying predictors (e.g. household total income), measured at _t−1, the wave prior to the potential separation.

Results

Table 1 presents the main descriptive statistics for the independent and control variables in the analysis (all observations). This descriptive picture shows some differences between cohabitors and marrieds. In what regards our dependent variables, we observed more instability in cohabiting unions: 8.6% separated during the observation period, versus only 1.8 of marriages. Married couples tend to be more educated (we find more couples where at least one or both members have college education) and are much more likely to have a mortgage (more than 71% of marrieds versus almost 55% of cohabitors). Marrieds entered the union at a later age (30.84 years) than cohabitors (28.68 years) and have been in the union for longer (10.20 years versus 7.5 for cohabitors). As expected, we find children present more often in married couples' households. In this sense, the data are coherent with the literature on the socio-demographic characteristics of couples in the UK that we reviewed in the background section.

Table 1.

Descriptive statistics for married and cohabiting unions (time n-observations).

	Cohabitation (N = 5062) 428 events		Marriage (N = 8421) 156 events
Variables	Mean	SD	Mean	SD
Distribution of unpaid and paid work
Traditional equitable	.2769	.4475	.1884	.3911
Traditional inequitable	.2558	.4363	.3741	.4839
Non-traditional equitable	.1953	.3964	.2037	.4027
Non-traditional inequitable	.2060	.4045	.1773	.3819
Missing	.0660	.2483	.0565	.2309
Controls
Income
Ln (household total annual income)	10.08	.7605	10.31	.6450
Female partner's income share
Below 0.4	.2469	.4313	.2604	.4389
Between 0.4 and 0.6	.7049	.4561	.6959	.4601
Above 0.6	.0392	.1940	.0310	.1734
Missing information	.0090	.0945	.0127	.1121
Education
None of them has tertiary education	.4594	.4984	.3837	.4863
She has tertiary education	.1612	.3677	.1626	.3691
He has tertiary education	.1382	.3452	.1526	.3596
Both have tertiary education	.1688	.3746	.2441	.4296
Missing	.0722	.2589	.0570	.2318
Gender role attitudes
He has egalitarian attitudes	.8443	.2815	.8546	.3525
He has traditional attitudes	.0950	.2932	.1107	.3138
Missing	.0607	.2388	.0347	.1830
She has egalitarian attitudes	.9133	.3626	.9123	.2829
She has traditional attitudes	.0754	.2641	.0877	.2829
Missing	.0113	.1060	.0000	.0000
Union duration in years	7.5	7.5	10.20	10.20
Age at union formation in years	26.68	26.68	30.84	30.84
Partners' age difference	2.053	5.917	2.004	5.746
Number of children in household	.7456	1.013	1.046	1.086
Child under four years present in household
No	.9283	.2580	.9062	.2915
Yes	.0717	.2580	.0938	.2915
Mortgage (ref. partners do not have a mortgage)
No	.4475	.4973	.2821	.4500
Yes	.5486	.4977	.7179	.4500
Missing	.0039	.0625	.0000	.0000

Note: All variables are time varying, except age at union formation and age difference between partners.

However, it is interesting to note that cohabitors and marrieds are very similar in other dimensions. Income and women's contributions to household income are similar in both types of couple – in over 70% of observations women contributed less than 0.4 and less than 4% contributed above 60%. Age differences between the partners average two years in both unions. Contrary to what we expected, marrieds and cohabitors are also very similar in what concerns gender attitudes: 85.46% of married men and 84.43% of male cohabitors were egalitarian. Women were more egalitarian than men and living arrangements did not seem to make a difference: 91.23% of married women were egalitarian, 91.33% of female cohabitors too. These differences in gender values were not statistically significant.

Regarding our main independent variable, the division of work is traditional for about 50% of observations in both types of couples (53.27% for cohabiting couples and 56.25% for marrieds). However, we observe more often balanced arrangements in cohabiting unions (47.22% of observations) than in marriages (39.21%). The most frequent division in marriages was traditional inequitable, whereas in cohabitations it was traditional equitable.

Table 2 presents the results from our discrete-time event history models. The coefficients for the distribution of work take as a reference the ‘non-traditional equitable’ arrangement, which would be the egalitarian model where both partners work for pay and share the housework. According to our results, cohabitors are more sensitive to the distribution of work, as a traditional inequitable arrangement increases the risk of dissolution when compared to the non-traditional equitable arrangement. In other words, cohabiting couples where the woman does a double shift experience a 33.68% higher risk of dissolution – compared to couples where the division is traditional but equitable. This result offers partial support to hypothesis 2. However, we found no significant effects for marrieds. If we use the traditional equitable model as a reference category (results not shown) we do not find a significant effect either. This is not consistent with our hypothesis and points at married couples being less sensitive to changes in the division of domestic work.

Table 2.

Discrete-time event history logit models for union dissolution.

	Cohabitation	Marriage
Distribution of work (ref. non-traditional equitable)
Traditional equitable	−.1232 (.1790)	.1096 (.2683)
Traditional inequitable	.3368* (.1661)	.2582 (.2758)
Non-traditional inequitable	.2170 (.1580)	−.1007 (.2896)
Missing	−1.032* (.4122)
Ln (household total income)	.1508^† (.0802)	−.3181** (.1159)
Women's income share (ref. 0.4–0.6)
Below 0.4	.2896^† (.1521)	−.2290 (.2305)
Above 0.6	−.2763 (.2985)	.6243^† (.3762)
Missing	2.587*** (.6826)
Education (ref. none has tertiary education)
He has tertiary education	−.1753 (.1807)	−.5489* (.2597)
She has tertiary education	.2940* .1477)	.0025 (.2249)
Both have tertiary education	−1218 (.1727)	−.8707** (.2885)
Gender role attitudes (ref. non-traditional attitudes)
He has traditional attitudes	.3186^† (.1700)	.4632^† (.2396)
She has traditional attitudes	.1967 (.1843)	−.3330 (.2949)
His attitudes missing	1.750*** (.2214)
Mean age at the start of union	−.0345*** (.0081)	−.0380** (.0110)
Partners' age difference	.0174* (.0086)	−.0178 (.0168)
Number of children	.3327*** (.0837)	.3973*** (.1045)
Child under four (ref. No)	−.4557 (.3044)	−.8467* (.3907)
Mortgage (ref. No)	−.2684* (.1186)	.0195 (.1807)
Duration of relationship in years	−.7283*** (.0568)	−.3571*** (.0902)
Duration of relationship squared	.0257*** (.0039)	−.0007 (.0054)
Premarital cohabitation		0.4503*** (0.1314)
Constant	−.6358 (.7999)	2.689* (1.211)
Number of events Number of couples Number of observations	438 1781 5062	156 1518 8421

Note: Robust standard errors in parentheses.

^†p < .1; *p < .05; **p < .01; ***p < .001.

Regarding the other covariates that should affect the risk of dissolution, as expected, effects for marriage and cohabitation diverge. For married couples, longer durations of the union, higher total income, higher age at union formation, and having a child under four years of age are associated with lower risks of dissolution. If the man or both partners have college education, they also have a lower risk of dissolution, compared to couples where neither of the partners has college education. On the other hand, premarital cohabitation and the number of children increase the risk of dissolution. Having traditional gender roles does not have a significant effect on the risk of separation, although the effect is positive and close to significance for men.

For cohabitation, age at union formation and duration show the same trend as for marriage, although the effect of duration is corrected by the results of the squared variable. Having a mortgage is also associated with lower risks of dissolution for these couples. However, income seems to be less relevant for cohabitors, as the variable does not reach statistical significance. Having more children in the household increases the risk of dissolution for cohabitors, as it did for marrieds. Age differences and women's college education also make a difference for cohabitors, increasing the risk of dissolution. Having traditional gender roles does not have a significant effect, but we find the same trend as in marriage: the effect is positive and close to significance for men. It is also significantly higher for couples in which men did not answer to the questions on gender values.

Discussion

Nowadays families are more diverse and unstable than they were in the past, and they have to adapt to men's and women's new roles in paid and unpaid work. In this paper we have analysed the effects of the division of work on couple stability, taking into account that marriage and cohabitation might follow different dynamics. We expected marriages to be protected by more traditional arrangements and cohabitation to be more stable under egalitarian arrangements, as the literature has suggested. Our main contribution is the integration of equity aspects of the division by considering the division of the total workload – paid and unpaid work – between the partners. Equity has been analysed before using subjective perceptions on fairness, and our idea in this paper was to tackle it from the point of view of balanced workloads, as well as testing the traditional/egalitarian dimension. We have analysed these hypotheses using data from 16 waves of the BHPS (1992–2008), including information from both partners.

Results support our hypotheses only partially. For cohabitations, we find that couples who move from an egalitarian and equitable model – sharing paid and unpaid work equally – to a ‘traditional inequitable’ arrangement – where women do a second shift – have a higher risk of union dissolution when compared to egalitarian equitable couples. The support for our hypothesis is partial because we find no significant effect for the other traditional arrangement nor for the other inequitable one, although the sign of the coefficients would point at inequity, more than inequality, increasing the risk of dissolution. Our results would suggests that dissolution is associated with movements away from this specific arrangement – both egalitarian and equitable – and not from any type of egalitarian arrangement as found by Kalmijn et al. (2007). This finding is in line with the effect that Bellani et al. (2018) found for all couples in the USA, where women's double shift had a destabilising effect. For marriages, we did not find support for our hypothesis, as there was no association of traditionalism or inequity and risks of dissolution. According to our findings, marriage stability in the UK would not be significantly associated with changes in the division of work.

These results confirm the idea that cohabitation and marriage are different types of union in what regards the division of work. In line with the literature, we find that cohabitors and marrieds favour different domestic divisions of work, but equity (in the sense of having similar workloads) seems to be an important dimension for cohabitors. However, for marriage we do not find that the division of work influences dissolution, so we cannot say that marriage is protected by more traditional arrangements: in our results marriage is simply more stable than cohabitation, irrespective of the division of work. This would be in line with the literature that points at commitment as a key difference between both unions, but it could also be related to the meaning of the relationship: if cohabitation is seen as a trial period, it is more easily dissolved. Selection could also play a role here, although we have controlled for the main socio-demographic factors identified in the literature. We tried to test the role of gender attitudes in our analysis, but our results were not conclusive. We think that social desirability might have played a role in this lack of effect, as most of the respondents showed egalitarian attitudes, and thus the level of variation was small. The survey data was not specialised in gender roles but asked only several questions on this topic amid a large questionnaire, and self-presentation concerns have been shown to influence respondents (Krumpal, 2013).

Our exploratory analysis shows that it can be interesting for researchers to integrate equity to factual analyses of the division of work, although the measure that we have used here has important limitations. We have approached the idea of equity with a measure that is related to the distribution of the total workloads, but the workloads reported by individuals are not completely objective. In addition to this, individuals might formulate equity in a different and more complex way, for instance taking into account child care, leisure time, commuting, or others. An important aspect of the division of work that we cannot take into consideration with these data is the mental load of organising housework and planning for tasks. In addition to this, our measure of time spent in work – paid or unpaid – is not the most adequate, as it is based on individuals estimation and we know that time-use data would be better suited to analyse time. The measures used in this paper are therefore subject to improvement with other datasets.

In this paper, we have tried to contribute to the literature by introducing an exploratory measure of objective equity and equality in the research on the division of domestic work and its effects on couple stability. We think that our results open new lines of research on the domestic division of work and its consequences. Taking child care into consideration would be an interesting next step for research on this issue, especially for families with small children, when the time demands are higher.

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Iñigo Fernández received a pre-doctoral fellowship from the Department of Education of the Basque Government, Spain.

References

Altintas

Sullivan

(2016) Fifty years of change updated: Cross-national gender convergence in housework. Demographic Research 35(16): 455–470.

Amato

Johnson

Booth

, et al. (2003) Continuity and change in marital quality between 1980 and 2000. Journal of Marriage and Family 65: 1–22.

Batalova

Cohen

(2002) Premarital cohabitation and housework: Couples in cross-national perspective. Journal of Marriage and Family 64(3): 743–755.

Baxter

(2005) To marry or not to marry: Marital status and the household division of labor. Journal of Family Issues 26: 300–321.

Baxter

Haynes

Hewitt

(2010) Pathways into marriage: Cohabitation and the domestic division of labor. Journal of Family Issues 31: 1507–1529.

Beaujouan

Bhrolcháin

(2011) Cohabitation and marriage in Britain since the 1970s. Population Trends 145: 1–25.

Bellani

Esping Andersen

Pessin

(2018) When equity matterns for couple stability: Comparing German and US couples. Journal of Social and Personal Relationships 35(9): 1273–1298.

Berrington

(2001) Entry into parenthood and the outcome of cohabiting partnerships in Britain. Journal of Marriage and Family 63: 80–96.

Bianchi

Lesnard

Nazio

, et al. (2014) Gender and time allocation of cohabiting and married women and men in France, Italy and the United States. Demographic Research 31: 183–216.

10.

Brines

(1994) Economic dependency, gender and the division of labor at home. American Journal of Sociology 100: 652–688.

11.

Brines

Joyner

(1999) The ties that bind: Principles of cohesion in cohabitation and marriage. American Sociological Review 64: 333–355.

12.

Chan

Halpin

(2002) Union dissolution in the United Kingdom. International Journal of Sociology 32: 76–93.

13.

DeMaris

(2007) The role of inequity in marital disruption. Journal of Social and Personal Relationships 24: 149–166.

14.

DeMaris

(2010) The 20-year trajectory of marital quality in enduring marriages: Does equity matter?. Journal of Social and Personal Relations 27(4): 449–471.

15.

Domínguez-Folgueras

(2012) Is cohabitation more egalitarian? The division of household labor in five European countries. Journal of Family Issues 34: 1623–1646.

16.

Esping-Andersen

Billari

(2015) Re-theorizing family demographics. Population and Development Review 41(1): 1–31.

17.

Fahlén

(2016) Equality at home – A question of career? Housework, norms, and policies in a European comparative perspective. Demographic Research 35: 1411–1440.

18.

Frisco

Williams

(2003) Perceived housework equity, marital happiness and divorce in dual-earner households. Journal of Family Issues 24: 51–73.

19.

Fuwa

Tsutsui

(2010) A cross-national comparison of the perceived fairness of the division of household labor. Japanese Journal of Family Sociology 22: 52–63.

20.

Grunow

Evertsson

(2016) Couples' Transitions to Parenthood. Analysing Gender and Work in Europe, London: Edward Elgar Publishing.

21.

Gupta

(2007) Autonomy, dependence or display? The relationship between married women's earnings and housework. Journal of Marriage and Family 69: 399–417.

22.

Hallberg

Klevmarken

(2003) Time for children, a study of parents' time allocation. Journal of Population Economics 16(2): 205–226.

23.

Heuveline

Timberlake

(2004) The role of cohabitation in family formation. The United States in comparative perspective. Journal of Marriage and Family 66: 1214–1230.

24.

Hochschild

Machung

(1991) The Second Shift: Working Parents and the Revolution at Home, New York: Viking.

25.

Jalovaara

(2013) Socioeconomic resources and the dissolution of cohabitations and marriages. European Journal of Population 29: 167–193.

26.

Kalmijn

(2007) Explaining cross-national differences in marriage, cohabitation and divorce in Europe 1990–2000. Population Studies 61: 243–263.

27.

Kalmijn

Loeve

Manting

(2007) Income dynamics in couples and the dissolution of marriage and cohabitation. Demography 44: 159–179.

28.

Kiernan

(2004) Unmarried cohabitation and parenthood in Britain and Europe. Journal of Law and Policy 26: 33–55.

29.

Krumpal

(2013) Determinants of social desirability bias in sensitive surveys: A literature review. Quality & Quantity 47(4): 2025–2047.

30.

Lewis

(2001) Debates and issues regarding marriage and cohabitation in the British and American literature. International Journal of Law, Policy and the Family 15: 159–184.

31.

Liefbroer

Dourleijn

(2006) Unmarried cohabitation and union stability: Testing the role of diffusion using data from 16 countries. Demography 43: 203–221.

32.

Lynn

(2006) Quality Profile: British Household Panel Survey. Version 2.0: Waves 1 to 13: 1991–2003, UK: Institute for Social and Economic Research, University of Essex.

33.

Nakamura

Akiyoshi

(2015) What determines the perception of fairness regarding household division of labor between spouses?. PLoS ONE 10(7): e0132608.

34.

Office for National Statistics (2012) Short Report: Cohabitation in the United Kingdom. Office for National Statistics, UK.

35.

Office for National Statistics (2013) Divorce in England and Wales 2011. Statistical Bulletin 1–16. Office for National Statistics, UK.

36.

Oláh

Gähler

(2014) Gender perceptions, division of paid and unpaid work, and partnership dissolution in Sweden. Social Forces 93: 571–594.

37.

Oppenheimer

(1997) Women's employment and the gain to marriage: The specialization and trading model. Annual Review of Sociology 23: 431–453.

38.

Özcan

Breen

(2012) Marital instability and female labor supply. Annual Review of Sociology 38: 463–481.

39.

Poortman

Kalmijn

(2002) Women's labour market position and divorce in the Netherlands: Evaluating economic interpretations of the work effect. European Journal of Population 18: 175–202.

40.

Ruppaner

Bradén

Turunen

(2018) Does unequal housework lead to divorce? Evidence from Sweden. Sociology 52(1): 75–94.

41.

Sayer

Bianchi

(2000) Women's economic independence and the probability of divorce. A review and reexamination. Journal of Family Issues 21: 906–943.

42.

Sayer

England

Allison

, et al. (2011) She left, he left: How employment and satisfaction affect women's and men's decisions to leave marriages. American Journal of Sociology 11: 1982–2018.

43.

Schober

(2013) Gender equality and outsourcing of domestic work, childbearing, and relationship stability among British couples. Journal of Family Issues 34: 25–52.

44.

Schoen

Astone

Kim

, et al. (2002) Women's employment, marital happiness, and divorce. Social Forces 81: 643–662.

45.

Smart

Stevens

(2000) Cohabitation, Breakdown, London: Family policy studies centre.

46.

South

(2001) Time dependent effects of wives' employment on marital dissolution. American Sociological Review 66: 226–245.

47.

Wilcox

Nock

(2006) What's love got to do with it? Equality, equity, commitment, and women's marital quality. Social Forces 84: 1321–1345.

48.

Wilson

Smallwood

(2007) Understanding recent trends in marriage. Population Trends 128: 24–32.

49.

Yamaguchi

(1991) Event History Analysis, Newbury Park: Sage.

50.

Zabel

Heinz-Martin

(2013) Does children's age impact the division of housework? A comparison of France, eastern and western Germany. European Societies 15(3): 663–685.