All pages

Chapter 4: The Organisation of Households

The division of household labour

Despite a remarkable rise in female labour force participation and a persistent, albeit sluggish, movement towards greater gender equality in society, the changes in women’s economic position did not bring about a real change in the division of housework and care work within the family. While the amount of time women invest in housework has declined in recent decades, the increase in time spent by men on household chores only partially offset this reduction [Ger03] [Col00]. In all industrialised countries, women are still left with the major share of housework and childcare. Even when quantitative differences in the time spouses devote to housework are small, women tend to bear the main responsibility for routine household activities. Routine activities essential to the functioning of the household are typically performed by women, whereas men typically take responsibility for less regular maintenance activities. In conceptualising the division of household labour, then, both the time and responsibility dimensions must be taken into account. Not only is the time devoted to housework by female and male spouses extremely unbalanced, but the performance of household chores remains highly segregated and gender dependent [Orl02]. Only a minority of couples shares responsibility for household tasks [Gol91] [Pre94] [Sti00]. In Hochschild’s [Hoc89] terms, this is the ‘stalled revolution’ which, she argues, results from the lack of institutional arrangements that could ease the life of working parents and balance the demands of work and family.

Recent studies have made some progress in integrating and articulating a number of sociological perspectives on the household division of labour and have included comprehensive data analysis [Bax97] [Bia00] [Col00], but our understanding of this phenomenon and its persistence is still rather rudimentary. At the micro-level, the issue is typically framed in terms of the ‘implicit contract’ between the spouses, which is partially dependent on their personal attributes and labour market activities. On a broader level, the gender division of labour is addressed in terms of the societal factors that influence the organisation of the household. These include the relationship between family, market and state, as well as cultural norms. The latter concern is best addressed by means of comparative analysis that permits the examination of structural factors associated with different patterns of household organisation.

In line with the theme of this course, the current chapter addresses the role of social institutions in influencing the household division of labour. Our aim is to investigate the extent to which different institutional arrangements associated with various ‘regime’ types are related to country variations in the division of household labour and the amount of time allocated to unpaid work by men and women. Furthermore, we are interested in the ways in which the relationship between spouses’ attributes and the characteristics of the household are moderated by the institutional context as captured by the regime typology.

Before addressing the comparative issues, however, we turn to a short review of the central themes that have emerged in the theoretical literature on the micro-level correlates of the gendered division of household labour. It is these relationships that one can then examine in different social contexts in order to evaluate the ways in which systemic characteristics impinge on the organisation of families.

Page 1

Generalised explanations of the division of household labour

Sociological approaches to the household division of labour can be grouped into two broad classes (for a more detailed discussion see [Lew06]. One class of explanations is framed in terms of micro-relations and includes several variants of exchange theory that view the actual practices followed by spouses as representing a negotiated agreement, which may or may not be based on power relations between the spouses (see [Bec81] [Mcd81], respectively). A number of approaches that fit this broad perspective have received detailed exposition and evaluation [Bia00] [Blo60] [Bri94]. Prominent among them is the resource-dependence approach, which focuses on asymmetric power relations that typically places female spouses at a disadvantage.

A second class of explanations is framed in more abstract terms and is represented by the proposition that gender categories are constituted through the actions of individuals and their social relations. Hence, the household division of labour is not merely an agreement struck between two social actors. Rather, the household is the locus of structuration, reflecting gender ideology, socialisation, differential power, while at the same time institutionalising gender distinctions [Ber85] [Hoc89] [Wes87].

Page 2

Resource-dependence

Resource dependence theory focuses on spouses’ relative power, which derives from their access to own resources (often defined as income or education). Women’s access to money of their own gives them alternatives outside of marriage and allows them a ‘voice’ in household matters [Hir70] [Hob90]. Therefore, it grants them greater bargaining power. From this point of view, women's employment is likely to change the relative economic relationship between the spouses, thus narrowing the power gap between them in their contribution to household income [Bri94] [Bit03] [Blo60] [Eve04] [Ros87]. Because of fewer available alternatives to the relationship and when evaluating the equity of the spousal relationship, more dependent women are likely to accept, or may be forced into taking upon themselves, primary responsibility for running the household [Len94].

Put differently, the resource-dependence perspective suggests that the division of household tasks largely reflects the differential power of the spouses within the household, and this in turn is determined by their relative social statuses. This approach involves the assumption that spouses prefer not to engage in housework, and the spouse who can muster a greater amount of resources or can threaten to withdraw resources in spousal negotiations will therefore carry less of the household burden [Bla91] [Bri94] [Mar84]. To the extent that women are less able than their spouses to secure valuable resources for the household (e.g. income and symbolic rewards), they will carry most of the responsibility for household chores. Increasingly, studies have focused on a particular form of resource-power relations which views the division of labour as a function of economic dependence ‘… best defined in terms of one spouse’s reliance upon the other for his or her current income standard’1 (Brines, 1994:657).

Page 3

Time constraints

Time constraints faced by spouses are additional determinants of the organisation of the household. This is clearly the case when it comes to women's employment. Numerous studies have documented the patterns of household work and their determinants (for example, [Gol91] [Bia00] [Bit03] [Bri94] [Bra92] [Col00]). These studies demonstrate that several factors at the household level are associated with more equal participation of spouses in housework. In particular, women’s employment activity has consistently been found to influence the time they devote to housework [Bia00] [Bla91] [Kal90] [Ros87], and, in some cases, also the participation of men in household chores [Bia00] [Dav04].

Women’s labour market activity affects household arrangements in several ways. Firstly, the time available to perform household tasks is reduced; thus, even in households in which women assume the chief responsibility for housework and childcare, the amount of time invested in these activities is lower than in instances where women are not employed. In these cases, the standards of housework are typically modified and less time is devoted to household chores [Bia00]. In some cases, household work may be delegated to other household members (the husband, children) or to market-provided domestic services [Lev06].

Life-cycle factors such as transitions to marriage, childbearing and aging also influence the division of household labour. Having more and younger children increases the demand for housework [Col00] [Cow92], and this typically affects women’s housework more than men’s [Sou94]. Younger women tend to spend less time doing housework and share more with their spouses than older women [Her94], reflecting both cohort and life-cycle changes.

An important mechanism for coping with the time constraints is the outsourcing of domestic tasks, as in the case of frequently eating out or employing a third party to do housework. Studies have repeatedly found that such outsourcing is associated with higher family income [Bit99 ] [Van04] and is positively related to relative power of the female spouse [Coh98]. Although it is commonly believed that outsourcing saves time, only a few studies have actually estimated its effect on the amount of time couples spend on housework. Brines [Bri94] has shown that the relative expenditure on restaurant meals was negatively related to wives’ (but not husbands’) housework time, while Van der Lippe et al. [Van04] have reported that both domestic help and take-away meals were negatively associated with women’s housework and the latter had a weak negative effect on men’s housework as well. Outsourcing, then, appears to be an intervening factor leading to a reduction of time spent on household chores by wives with relatively ample resources. Although outsourcing frees spouses from the performance of certain household chores, there is still the matter of organising and overseeing these activities. This responsibility is typically in the hands of the female spouse, who will contract the help and supervise the actual performance. Hence, while the amount of time the female spouse might spend on housework is reduced, her responsibility is not, and the gendered nature of housework is unlikely to disappear.

Page 4

Exercise 4.1

  1. What does the resource-dependence hypothesis assert with respect to the gender division of household labour?
  2. What does the time-constraints hypothesis assert with respect to the gender division of household labour?
  3. Comparing Germany and Denmark, in what ways would you expect the two countries to differ with respect to the effects of resource-dependence and time-constraints on the gender division of household labour?
  4. Use the multiple regression procedure to examine the resource-dependence hypothesis and the time-constraints hypothesis separately for men and women in Germany and Denmark.

Comment:

In order to examine the effects of resource-dependence and time-constraints on the gender division of household labour, it is necessary to operationally define each of these concepts, as well as to statistically control for other factors that might affect the household division of labour.

Procedure:

Questions:

  1. Arrange the results of the analysis in a table consisting of four columns (one column for each gender group by country). The relevant coefficients for each variable should be listed in the rows.
  2. For women, what is the effect of resource-dependence on the proportion of housework they perform? How do they differ between Germany and Denmark?
  3. For women, what is the effect of time-constraints on the proportion of housework they perform? How do the results differ for Germany and Denmark?
  4. Discuss the results from 'b' and 'c' for men.
  5. Present your conclusions concerning the similarities and differences between Germany and Denmark and how this might be related to gender regimes.
Nesstar
SPSS syntax
*You need to have a copy of SPSS installed on your computer, and you should download and use the dataset Family, Gender and Work.
*Open SPSS by clicking on the appropriate link.
*Open the ESS data by clicking ‘File’, ‘Open’, and ‘Data’ on the SPSS menu bar before you select the folder and the data set.
*Open a new syntax window by clicking ‘File’, ‘New’, and ‘Syntax’ on the SPSS menu bar.
*You can copy the syntax below and paste it into the syntax window in SPSS.
*Execute the syntax using the 'Run' option on the menu bar.

*Comments on commands start with an asterisk and end with a dot.

*Commands must always end with a dot.

*The following command causes the cases to be weighted by the design weight variable 'dweight'.

WEIGHT BY dweight.

*Create filter variable - only include women from Germany and Denmark living as partner.

USE ALL.
COMPUTE filter_$=gender=0 & (cntry ='DK' | cntry ='DE') & (marital=1 or lvgptn=1).
VARIABLE LABEL filter_$ 'gender=0 & (cntry =DK or cntry =DE) & marital=1 or lvgptn=1 (FILTER)'.
VALUE LABELS filter_$ 0 'Not Selected' 1 'Selected'.
FORMAT filter_$ (f1.0).
FILTER BY filter_$.
EXECUTE.

*SPLIT FILE splits the active dataset into subgroups that can be analysed separately.

SORT CASES BY cntry.
SPLIT FILE LAYERED BY cntry.
 
REGRESSION
/MISSING LISTWISE
/STATISTICS COEFF OUTS R ANOVA
/CRITERIA=PIN(.05) POUT(.10)
/NOORIGIN
/DEPENDENT hwkpwd1
/METHOD=ENTER pphincr wkhct age children.

*Create filter variable - only include men from Germany and Denmark living as partner.

USE ALL.
COMPUTE filter_$=gender=1 & (cntry ='DK' | cntry ='DE') & (marital=1 or lvgptn=1).
VARIABLE LABEL filter_$ 'gender=1 & (cntry =DK | cntry =DE) & (marital=1 or lvgptn=1) (FILTER)'.
VALUE LABELS filter_$ 0 'Not Selected' 1 'Selected'.
FORMAT filter_$ (f1.0).
FILTER BY filter_$.
EXECUTE.
 
REGRESSION
/MISSING LISTWISE
/STATISTICS COEFF OUTS R ANOVA
/CRITERIA=PIN(.05) POUT(.10)
/NOORIGIN
/DEPENDENT hwkpwd1
/METHOD=ENTER pphincr wkhct age children.

*Turn off the split file and weight, and select all cases.

SPLIT FILE OFF.
WEIGHT OFF.
USE ALL.
Sample solution

Problem

  1. The resource-dependence hypothesis asserts that the division of household labour largely reflects the relative power of the spouses. More specifically, it suggests that the more balanced the statuses of the spouses (e.g. they have similar earnings), the more equal the division of the household labour. Alternatively, the more dependent the female spouse is on her husband's income, the greater will be her share of housework.
  2. The time-constraints hypothesis asserts that the division of household labour is a function (in part) of time availability. Specifically, the more time a woman spends in the labour market, the less time she will devote to housework. This generally means a more equal division of housework between the spouses.
  3. Denmark is characterised by the individual-independence regime. One might expect, therefore, that the division of household labour will not be much affected by the relative earnings of the spouses. Hence, we would expect a stronger negative effect of a female spouse’s share of total household earnings on her share of housework in Germany than in Denmark. As to the time constraints, we would expect a weaker effect in Germany, since the state-dependence regime still upholds a more traditional family model (even when women join the labour force).

Answers

  1. Table with regression results
  2. The effect of the proportion of household income provided by the female spouse (this is the complement of dependence) on the time she devoted to housework (as a proportion of the total time spent by spouses) is b = -0.266 in Germany, and b = -0.221 in Denmark. In both countries, the coefficients are significant and they are very similar in magnitude.
  3. The effect of time constraints on the time she devoted to housework (as a proportion of the total time spent by spouses) is very small and not statistically significant in either Germany or Denmark.
  4. The effect of the proportion of household income provided by the male spouse (this is another way of representing the female spouse’s dependence) on the time he devoted to housework (as a proportion of the total time spent by spouses) is b = -0.291 in Germany, and somewhat weaker in Denmark (b = -0.169). This may represent a weaker relationship between power and the division of household labour in the latter country, as one might expect. Interestingly, in Germany, the time constraints on men are not related to housework, but they are negative and borderline significant in Denmark (the more time the male spouse devotes to work in the market, the smaller his contribution to housework). That is, the male’s contribution is more sensitive to time constraints in Denmark than in Germany.
Page 5

Family roles and gender structuration

In contrast to the exchange/dependency-perspective and its derivatives that are generally conceptualised as (gender-neutral) micro-decisions, the second class of arguments can be defined as the gender structuration proposition. Better known as ‘doing gender’, this proposition suggests that social differences between women and men are constructed by means of ‘… a complex of socially guided perceptual, interactional and micropolitical activities …’ (West & Zimmerman, 1987: 126). Once constructed, they are used to reinforce gender differences. Hence, the performance of daily activities in the household and the emergent division of labour reproduces gender as a social category and reinforces male and female roles, identities and attitudes. The division of household labour, then, is not only about productivity but also about the ‘production’ of gender and gender relations through ‘everyday performance’ [Bat02] [Gre00]. Indeed, a study by Bittman and colleagues [Bit03] found that Australian women who are less dependent upon their husbands actually devoted relatively longer hours to housework compared with women who were more economically dependent on their spouses in order to compensate for their gender-role deviance (Bittman et. al 2003: 207). The authors interpreted these findings as empirical support for the ‘gender role’ hypothesis which posits that the division of household labour is not merely a negotiated social arrangement that balances the market and family inputs of the spouses. Instead, the household division of labour is constitutive in that it structures (or reinforces) gender categories at the same time that household goods and services are being produced.

With changing patterns of employment and family formation, the traditional model has given way to a variety of alternatives, and some researchers have predicted growing individualisation whereby the male and female roles become ‘de-complementary’ [Bur94] and the family shifts from ‘a community of need’ to ‘an elective relationship’ [Bec02]. In other words, as women’s participation in paid employment becomes universal, and as households become more dependent on dual earners, gender roles become less distinct. We believe that the blurring of gender roles in the economic sphere also contributes to the blurring of the traditional division of responsibilities within the household.

To the extent, then, that the household remains a locus of gender production, we might expect that the gender division of household labour would not be fully accounted for by resource-dependence relations or by time constraints. Culturally ingrained gender distinctions are most likely to materialise in the division of responsibility for household chores rather than in the time spent on housework. Furthermore, such gender distinctions should be related to more general gender attitudes held by the spouses. Indeed, Blair and Lichter [Bla91] found that when spouses held more egalitarian attitudes, the household was characterised by lower family work segregation, and Greenstein [Gre00] reported that husbands’ contribution to domestic labour was affected by the interaction of both spouses’ gender ideologies.

Page 6

Exercises 4.2 - 4.3

Exercise 4.2

To be able to solve this exercise, you need to have software (SPSS) installed on your computer, and you should download and use the dataset Family, Gender and Work.

  1. Who are likely to express more egalitarian gender-role attitudes, women or men? Explain.
  2. Thinking of Greece, Belgium and Sweden, and based on the typology of gender regimes, in what country would you expect to find less egalitarian gender-role attitudes?
  3. In which of the three countries would you expect to find the smallest difference in attitudes between women and men?
  4. Test your hypotheses.

Comment:

For this exercise, it will be necessary to create an index consisting of several gender-role items. It will be necessary to test the scaling properties of the index and to construct a new variable from the items comprising the index. The new variable will then be used to answer the questions above.

Procedure:

Question:

  1. Discuss the results of the analysis with reference to questions 1 to 4 listed above.
SPSS syntax
*You need to have a copy of SPSS installed on your computer, and you should download and use the dataset Family, Gender and Work.
*Open SPSS by clicking on the appropriate link.
*Open the ESS data by clicking ‘File’, ‘Open’, and ‘Data’ on the SPSS menu bar before you select the folder and the data set.
*Open a new syntax window by clicking ‘File’, ‘New’, and ‘Syntax’ on the SPSS menu bar.
*You can copy the syntax below and paste it into the syntax window in SPSS.
*Execute the syntax using the 'Run' option on the menu bar.

*Comments on commands start with an asterisk and end with a dot.

*Commands must always end with a dot.

*The following command causes the cases to be weighted by the design weight variable 'dweight'.

WEIGHT BY dweight.

*Reverse coding for G7.

RECODE
mnrsphm (1=5) (2=4) (3=3) (4=2) (5=1) into sameresp.
VARIABLE LABELS sameresp 'Men should take as much responsibility as women for home and children'.
VALUE LABELS sameresp 1 'Disagree strongly' 2 'Disagree' 3 'Neither agree nor disagree' 4 'Agree' 5 'Agree strongly'.
FORMATS sameresp (f2.0).
EXECUTE.

*Create filter variable - only include respondents from Greece, Belgium and Sweden living with partners.

USE ALL.
COMPUTE filter_$=(cntry ='BE' | cntry ='GR' | cntry ='SE') & (marital=1 or lvgptn=1) .
VARIABLE LABEL filter_$ 'cntry =BE or GR or SE and (marital=1 or lvgptn=1)(FILTER)'.
VALUE LABELS filter_$ 0 'Not Selected' 1 'Selected'.
FORMAT filter_$ (f1.0).
FILTER BY filter_$.
EXECUTE.
 
RELIABILITY
/VARIABLES=wmcpwrk mnrgtjb sameresp
/FORMAT=NOLABELS
/SCALE(ALPHA)=ALL/MODEL=ALPHA.
 
COMPUTE gndridl = mean (wmcpwrk,mnrgtjb,sameresp).
VARIABLE LABEL gndridl 'Gender ideology (index, mean of wmcpwrk,mnrgtjb,sameresp)'.
EXECUTE.
 
T-TEST
GROUPS=gender(1 0)
/MISSING=ANALYSIS
/VARIABLES=gndridl
/CRITERIA=CIN(.95).

*SPLIT FILE splits the active dataset into subgroups that can be analysed separately.

SORT CASES BY cntry.
SPLIT FILE LAYERED BY cntry.
 
T-TEST
GROUPS=gender(1 0)
/MISSING=ANALYSIS
/VARIABLES=gndridl
/CRITERIA=CIN (.95).

*Turn off the split file and weight, and select all cases.

SPLIT FILE OFF.
WEIGHT OFF.
USE ALL.
Sample Solution

Problem

  1. We would expect women to express more egalitarian views on average than men as part of a more general expectation of reaching parity with men in all spheres of life.
  2. Given that Greece best fits the family-dependence regime, we would expect the gender-role attitudes in this country to be less egalitarian.
  3. The smallest difference is expected in Sweden. Sweden is characterised by the individual-independence regime, which is premised on the ideology of minimising gender inequalities.

Answer

  1. The gender-role items form a scale with moderate reliability (α = 0.59). Overall, we find that males and females do not differ much in their attitudes (mean = 3.50 for men and mean = 3.59 for women). However, the difference is statistically significant (t = -2.9), so we reject the null hypothesis that there is no difference in the mean scores for men and women. As expected, gender-role attitudes are least egalitarian in Greece and most egalitarian in Sweden (high scores represent more egalitarian attitudes). As to gender differences in the three countries, the gap is largest in Greece and narrowest in Belgium (no significant difference between men and women). The latter finding was not expected, as we anticipated that men and women would hold more similar views in Sweden than in other countries.

Exercise 4.3

  1. Formulate a hypothesis regarding the effect of respondents’ gender-role attitudes on the household division of labour.
  2. Test your hypothesis separately for women and men in each of the three countries discussed in Ex. 4.2.
  3. To do so, you should add the gender-role index you created (Ex. 4.2) to the regression analysis you performed in Ex. 4.1
SPSS syntax
*You need to have a copy of SPSS installed on your computer, and you should download and use the dataset Family, Gender and Work.
*Open SPSS by clicking on the appropriate link.
*Open the ESS data by clicking ‘File’, ‘Open’, and ‘Data’ on the SPSS menu bar before you select the folder and the data set.
*Open a new syntax window by clicking ‘File’, ‘New’, and ‘Syntax’ on the SPSS menu bar.
*You can copy the syntax below and paste it into the syntax window in SPSS.
*Execute the syntax using the 'Run' option on the menu bar.

*Comments on commands start with an asterisk and end with a dot.

*Commands must always end with a dot.

*The following command causes the cases to be weighted by the design weight variable 'dweight'.

WEIGHT BY dweight.

*Create filter variable - only include respondents from Greece, Belgium and Sweden living with partners. *Female.

USE ALL.
COMPUTE filter_$=gender = 0 & (cntry ='BE' | cntry ='GR' | cntry ='SE') & (marital=1 or lvgptn=1).
VARIABLE LABEL filter_$ 'gender = 0 & (cntry =BE or GR or SE) & (marital=1 or lvgptn=1) (FILTER)'.
VALUE LABELS filter_$ 0 'Not Selected' 1 'Selected'.
FORMAT filter_$ (f1.0).
FILTER BY filter_$.
EXECUTE.

*SPLIT FILE splits the active dataset into subgroups that can be analysed separately.

SORT CASES BY cntry.
SPLIT FILE LAYERED BY cntry.
 
REGRESSION
/MISSING LISTWISE
/STATISTICS COEFF OUTS R ANOVA
/CRITERIA=PIN(.05) POUT(.10)
/NOORIGIN
/DEPENDENT hwkpwd1
/METHOD=ENTER gndridl pphincr wkhct age children.

*Turn off the split file and select all cases.

SPLIT FILE OFF.
USE ALL.

*Create filter variable - only include respondents from Greece, Belgium and Sweden living with partners. *Male.

USE ALL.
COMPUTE filter_$=gender = 1 & (cntry ='BE' | cntry ='GR' | cntry ='SE') & (marital=1 or lvgptn=1).
VARIABLE LABEL filter_$ 'gender = 1 & (cntry =BE or GR or SE) & (marital=1 or lvgptn=1) (FILTER)'.
VALUE LABELS filter_$ 0 'Not Selected' 1 'Selected'.
FORMAT filter_$ (f1.0).
FILTER BY filter_$.
EXECUTE.

*SPLIT FILE splits the active dataset into subgroups that can be analyzed separately.

SORT CASES BY cntry.
SPLIT FILE LAYERED BY cntry.
 
REGRESSION
/MISSING LISTWISE
/STATISTICS COEFF OUTS R ANOVA
/CRITERIA=PIN(.05) POUT(.10)
/NOORIGIN
/DEPENDENT hwkpwd1
/METHOD=ENTER gndridl pphincr wkhct age children.

*Turn off the split file and weight, and select all cases.

SPLIT FILE OFF.
WEIGHT OFF.
USE ALL.
Sample Solution

Problem

  1. Hypothesis: the more egalitarian the gender-role attitudes of the respondent, the more equality in the household division of labour.

Answers

Turning first to women, the coefficient estimating the effect of gender-role attitudes (gender ideology) on the household division of labour is only statistically significant for Greece (b= -0.221, t = -3.823). The more egalitarian a woman’s ideology, the smaller is her relative contribution to housework. This generally means that housework is divided more equally.

The results for men indicate that, both in Greece and in Sweden, the more egalitarian the views held by the male spouse, the larger his relative contribution to housework (b = 0.2 and b = 0.338 in the two countries, respectively). In Belgium, the coefficients are not statistically significant, neither in the female nor in the male models.

Page 7

Societal context of family arrangements

Comparative studies that have investigated differences between countries in the division of housework have emphasised the role of institutional arrangements and cultural factors in affecting intra-household processes [Kal90] [Cal91] [Bax97] [Eve04] [Ger03] [Bat02] [Dav04] [Lew06]. These studies varied with regard to the list of (industrialised) countries they included and the data sources they utilised. Their main findings can be summarised as follows: (a) In all countries, gender is an important determinant of the household division of labour, as women perform the lion’s share of housework. (b) Countries differ in the amount of time invested in housework by both women and men, and in the extent of gender segregation of household chores. In some countries, equality is greater because women allocate less time to housework (while men’s contribution is similarly low). In other countries, men participate more in unpaid work than in other countries. (c) Different factors may affect the division of labour and the contribution of each spouse to housework in different countries. For example, cultural factors, especially those related to gender ideology, may mediate the effects of time constraints or resource dependence on the division of housework [Bit03].

Several arguments were put forward in order to explain differences between countries in the patterns of time allocation to unpaid work. In one of the first comparative studies on the issue of unpaid work, Kalleberg and Rosenfeld [Kal90] stated that ‘Cross-national research is necessary to examine how variation in institutional structures, policies, and cultural values affects the division of labour between men and women in the family and in the labour market’ (p. 332). One of the most commonly mentioned institutional factors is the countries’ welfare regime or, more specifically, the level of family-supportive policies [Kal90] [Ger03] [Bax97] [Eve04] [Fuw04].

Theoretical models of the household division of labour suggest that labour market policies that support women’s employment and improve their market position should promote equality within the household. This expectation is based on the premise that such policies give married women access to independent economic resources, which, in turn, grant them power to negotiate household relations. Several researchers have largely attributed country variation in the division of household labour to the level of support for working mothers (for example, [Bax97] [Cal91] [Ger03]. In general, family policies (such as maternity leave plans and subsidised childcare arrangements) are expected to improve women’s position by allowing them to combine family duties with market involvement. However, they may at the same time preserve the gendered division of labour precisely by making it possible for women to coordinate family and market activities.

It is important to note that men, who are still perceived as, and expected to be, the main providers, are often constrained by their employment contracts and have limited flexibility in contributing to housework. Family-supportive policies seldom target men’s involvement in family work or their role as parents (except for a few paternity leave programmes), and it is thus not clear whether, by affecting women’s economic activity, policies also influence the behaviour of women and men within the family. However, the general view is that policies, and especially those which directly affect women’s employment such as the availability of childcare facilities, maternity leave and child allowances, affect women’s time allocation to paid and unpaid work and consequently structure the relationship between work and family.

Page 8

Work-family regimes and household division of labour

Following this general argument, we hypothesise that the household division of labour can be expected to be most egalitarian under an individual-independence regime where women are encouraged to participate in paid employment and to receive more support from the state for doing so. As we noted earlier, in the individual-independence regime, women are not only more likely to gain access to independent resources that increase their power within the household, they are also more likely to relegate much of the housework (especially activities related to child rearing) to state institutions. Other gender-equality measures, especially those implemented in the labour market, also increase women’s power in relation to that of their spouses, and they thus serve to increase equality within the household.

Since women who participate in full-time paid work hold positions in the labour market that are more similar to men’s than other women do, it stands to reason that they also establish a more egalitarian division of labour at home [Sti00] [Fuw04]. We therefore expect a more egalitarian division of labour in the market-dependence regime than in the family and state-dependence regimes. This should be most evident among highly educated women (who have better opportunities in the labour market) and those who work continuously on a full-time basis. In other words, having a dual-earner ideology without state support creates a bifurcation among women. Those who can maintain their employment also achieve higher equality at home, while women who interrupt their employment due to childcare costs and lack of support, experience greater inequality within the family.

Although the household division of labour under a market-dependence regime may be more egalitarian on average than in other regimes, as noted above, we also expect that the effect of spouses’ relative resources on their household arrangements will be strongest in this regime, because market considerations play such an important role in it. This means that we would expect to find a stronger relationship between the extent of women’s dependence on their spouses and the division of labour in the household in the market-dependence regime than in other regime types. Similarly, time availability will be an important consideration in assigning household tasks in this regime, since the basic family ideology promotes gender equality through market involvement.

In the family-dependence regime, the division of household labour is expected to be more segregated. This is mainly because the family-dependence regime is premised on a traditional gender ideology, and the state provides no support for working mothers. The traditional ideology, combined with market constraints (few part-time jobs, gender inequality in the market) leads women to assume the chief responsibility for family work. Finally, in the state-dependence regime, while the division of labour is still traditional and gender-segregated, the fact that the state provides some direct support to women provides both actual and symbolic resources women can use to negotiate the household division of labour. In this context, reduced time availability associated with employment and the gender beliefs held by the spouses are expected to affect intra-household arrangements.

While the proposition that macro-level attributes should eventually be reflected at the micro-household level has been gaining support, the social mechanisms involved are not clearly understood. Two opposing mechanisms may be at work in this case. Family welfare policies may reduce the dependency of women on their spouses by altering the balance of resources, leading to a more egalitarian organisation of the household. At the same time, policies that encourage women’s labour market participation by integrating family and work activities and reducing women’s time constraints may actually contribute to greater inequality in the organisation of the household. This is because reduced-hours employment, generous paid leave schemes, and public childcare arrangements give women greater flexibility to combine employment with household tasks and relieve men of the responsibility for care work (see Hook’s [Hoo06] findings). It can be argued that state support for working mothers will help to increase gender equality within the household when it is accompanied by a perception of a gender-neutral division of labour, but not when the dominant ideology espouses a gendered division of labour. Thus, in the state-dependence regime, state support is expected to encourage a more segregated division of housework than in the individual-independence regime. Moreover, when dual-earner couples are compared across regimes, higher inequality within households is expected in the state-dependence regime than in the family-dependence regime, since in the latter women who participate in paid employment are most likely to be a highly selected group, either because of their market opportunities or family needs. They are also more likely to work full-time as a result of limited opportunities for part-time work.

Page 9

Exercise 4.4

To be able to solve this exercise, you need to have software (SPSS) installed on your computer, and you should download and use the dataset Family, Gender and Work.

  1. Sweden and Greece represent two different gender regime types. In which country would you expect couples to spend more time on housework?
  2. In which country would you expect men to perform a larger share of the housework?

Procedure:

Question:

  1. Relate the results to the hypotheses stated above.
SPSS syntax
*You need to have a copy of SPSS installed on your computer, and you should download and use the dataset Family, Gender and Work.
*Open SPSS by clicking on the appropriate link.
*Open the ESS data by clicking ‘File’, ‘Open’, and ‘Data’ on the SPSS menu bar before you select the folder and the data set.
*Open a new syntax window by clicking ‘File’, ‘New’, and ‘Syntax’ on the SPSS menu bar.
*You can copy the syntax below and paste it into the syntax window in SPSS.
*Execute the syntax using the 'Run' option on the menu bar.

*Comments on commands start with an asterisk and end with a dot.

*Commands must always end with a dot.

*The following command causes the cases to be weighted by the design weight variable 'dweight'.

WEIGHT BY dweight.

*Create filter variable - only include respondents from Greece and Sweden living with partners.

USE ALL.
COMPUTE filter_$=(cntry ='GR' | cntry ='SE' ) & (marital=1 or lvgptn=1).
VARIABLE LABEL filter_$ '(cntry =GR or SE) & (marital=1 or lvgptn=1) (FILTER)'.
VALUE LABELS filter_$ 0 'Not Selected' 1 'Selected'.
FORMAT filter_$ (f1.0).
FILTER BY filter_$.
EXECUTE.
 
T-TEST
GROUPS=cntry('GR' 'SE')
/MISSING=ANALYSIS
/VARIABLES= hwktwd1
/CRITERIA=CIN(.95) .

*Create filter variable - only include male respondents from Greece and Sweden living with partners.

USE ALL.
COMPUTE filter_$=gender=1 & (cntry ='GR' | cntry ='SE' ) & (marital=1 or lvgptn=1).
VARIABLE LABEL filter_$ 'gender=1 & (cntry =GR or SE) & (marital=1 or lvgptn=1) (FILTER)'.
VALUE LABELS filter_$ 0 'Not Selected' 1 'Selected'.
FORMAT filter_$ (f1.0).
FILTER BY filter_$.
EXECUTE.
 
T-TEST
GROUPS=cntry('GR' 'SE')
/MISSING=ANALYSIS
/VARIABLES= hwkpwd1 hwkpwe1
/CRITERIA=CIN(.95) .

*Turn off the weight and select all cases.

WEIGHT OFF.
USE ALL.
Sample solution

Problem

  1. In general, we would expect couples to spend less time overall on housework in Sweden as the incorporation of women into the labour force results in less time spent at home (and research has found that the reduction in time women spend on housework is not compensated by an increase in time invested by male spouses).
  2. We would expect men to perform a larger share of housework in Sweden due to the egalitarian ideology associated with its family regime.

Answer

  1. We find that the men’s share of housework is greater in Sweden than in Greece, whether on a typical weekday or a weekend. The differences are statistically significant. The findings are in line with our hypothesis.
Go to next chapter >>
Page 10