Even though the existing body of literature on income inequality between men and women is large, the issue continues to attract the attention of researchers. Recently, a new strand of literature has looked at the gender gaps in income that prevail at the micro level, that is, between a man and a woman who are forming a household. While the overall gender pay gap has barely changed in the EU and the US in the last decade, there is evidence that the share of households in which the woman outearns her partner appears to oscillate around 20–50%, depending on the country and period studied (Figari et al., 2011; Mysíková, 2016). Inequality in men’s and women’s contributions to household income has important consequences for couples’ behaviour. It could affect their decision making, division of unpaid work, allocation of resources, living standards, life satisfaction, and risk of divorce (e.g., Cooke, 2006; Gash & Plagnol, 2021; Sauer et al., 2021).

Our study provides new insights into the linkages between the partners’ relative incomes and their division of housework. Both the theoretical models and the existing empirical studies are inconclusive concerning these associations: i.e., it is unclear whether and, if so, under what circumstances the female partner contributing a larger share of the household income leads to a more equal division of housework, and under what circumstances it may encourage the woman to “act gender”, and thus to take on more of the housework. We contribute to the existing studies on this topic by analysing the role of individual level gender norms in the linkages between income and housework inequality. By focusing on Poland, which has a history of relatively high levels of female employment in full time jobs, and of those women having at the same time large domestic work burden, we also place our study in a different institutional setting than those studied so far. This context is similar to many Central and Eastern European (CEE) countries. It is also distinguished by a rather poor market of childcare and housework services, making it difficult for households to outsource housework (Farvaque, 2015).

For our analysis, we use data from the 2010 and 2014 waves of the “Generations and Gender Survey” for Poland. We find a negative, statistically significant link between the female share of household income and her involvement in housework, even if the woman earns more than her husband, which contradicts the “gender deviance neutralisation hypothesis” formulated in the recent literature. We also show that more egalitarian perceptions of gender norms are not only negatively associated with women’s housework burdens but also moderate the link between a woman’s contribution to the household income and her probability of doing most of the housework. Among traditional couples, the probability of a woman doing all the housework is not related to her contribution to the household income.

The paper consists of five sections. We start by reviewing the relevant literature while focusing on three strands of research: (1) within household income inequality, (2) gender gaps in housework, and (3) the role of gender norms in these fields. In addition to providing a literature review, we also offer insights into the institutional setting of the labour market and women’s position in the labour market in Poland. We then describe the data and the methodological approach we used. We present the results of our models in the Results section. The last section concludes.

Literature Review

Compared to the vast literature on economy-wide gender wage inequality, there are fewer studies on within household income inequality (for a review, see Blau & Kahn, 2017). It is only recently that scholars have started to stress the importance of within-household inequality in pay, pointing to its role in decision making within the household and, in turn, its effects on the allocation of resources, living standards, life satisfaction, and investments of both partners (Bonke & Browning, 2009; Greenstein, 2000; Vogler & Pahl, 2008). Some of this literature has looked at how income inequality between the partners is associated with the couple’s division of unpaid work and whether this division changes once the woman outearns her partner. We contribute to this literature by examining whether higher relative earnings of the female partner reduce the gender gap in unpaid work in the household and by investigating the factors that may influence this link, particularly gender norms.

The share of households in which the female partner earns close to or more than the male partner varies between 20 and 50%, depending on the country and period studied. Bertrand et al. (2015) showed that in the US, the woman outearned the man in 27% of dual earner married couples in 2010. Mysíková (2016) provided her calculations and summarised the previous findings on the female income shares in European countries. She found that the female income shares are the lowest in Southern European countries (below 30% in 2009), are relatively low in Western European countries, are above average in Scandinavia and most CEE countries (35–44%), and are highest in Slovenia and Denmark (> 40%). Klesment and Van Bavel (2017) estimated that the proportion of women who earn more than their husbands varies between 20 and 50% in European countries. Bertrand et al. (2015) documented that the distribution of the female share of household income exhibits a sharp drop to the right of the 0.5 value, i.e., at the point at which the female income exceeds the male income (also discussed by Zinovyeva and Tverdostup (2021) for Finland).

From a theoretical point of view, the link between the female contribution to the household income and the division of unpaid work between the partners should start with the analysis of the man’s and the woman’s allocation of time to paid work (which, together with wage obtained in exchange for work, determines the partners’ relative incomes) and other activities, which may include housework. Labour economists have provided three general frameworks for analysing male and female labour supply. First, male and female labour supply decisions may be individual, with separate utility functions and budget constraints, as posited in the classical model. Second, the partners’ labour supply decisions may be analysed jointly, with the incomes and hours of work pooled over the two individuals. However, this so called “neoclassical” approach treats households as single optimising units and ignores the within household allocations. By contrast, the third approach of labour supply modelling breaks from the unitary perspective and takes the collective perspective. This approach considers the household as a set of individuals with their own preferences who decide how to share their overall resources and maximise their utility functions given their budget constraints (Chiappori, 1988, 1992; Fortin & Lacroix, 1997). The model is simplified because it assumes that nonmarket work is equal to leisure, which may not be the case in families, particularly families with children (Apps & Rees, 1997). The important insight that these models provide is that the labour supply of the man and the woman who are forming the family/household are interdependent and that when the couple has children, these relationships and the degree of interdependence in the family become even stronger. While these theoretical frameworks help understand women’s lower labour supply relative to men’s, their findings regarding the division of domestic work among dual earners have been inconclusive.

This paper focuses on the link between the relative contributions of the male partner and the female partner to the household income and the division of unpaid work in the household. On the one hand, we would expect to find a negative association between a woman’s share of the household income and her involvement in housework. This can be related to both her greater bargaining power associated with her larger contribution to the household income and her higher absolute level of income, as higher earnings allow for outsourcing housework. This reasoning draws on economists’ theory of specialisation in marriage, which argues that a partner who earns more and thus has a higher market value also has a higher opportunity cost in the form of unpaid labour performed at home (Becker, 1991). A partner who earns more will thus contribute to home production less. Recently, it has, however, been shown empirically that comparative advantage plays a minimal role in the gender division of labour within couples (Siminski & Yetsenga, 2022).

On the other hand, cultural factors can add complexity to these considerations and suggest a nonlinear relationship between the relative income of partners and their involvement in housework. In line with the "doing gender" perspective (West & Zimmerman, 1987), a woman who earns more than her partner may take on more of the housework to compensate for having violated the “male breadwinner” norm—the social expectation that a man earns more in a couple. Violating the social "male breadwinner" norm could, in turn, have additional consequences, such as increasing the risk of marriage dissolution (Bertrand et al., 2015; Lippmann et al., 2020). The so called “gender deviance neutralisation hypothesis” thus appears at a certain point of the relative income distribution: once woman outearns men, she takes up a larger share of housework to normalise the nonnormative income distribution that appears in the household (Bittman et al., 2003; Lyonette & Crompton, 2015; Syrda, 2022).

The empirical evidence on the association between income and housework inequality in couples has been mixed. While some studies have reported that an increase in a woman’s relative income or her educational level are both associated with her male partner doing a larger share of the housework (Fetterolf & Rudman, 2014; Lewin-Epstein et al., 2006), other studies found no such relation (Nitsche & Grunow, 2016). Several authors have shown that, on average, a woman who earns more than her partner spends more time on housework, which confirms the gender deviance neutralisation hypothesis that women compensate for breaching gender norms (Bertrand et al., 2015; Bittman et al., 2003; Brines, 1994; Syrda, 2022). Yet, some studies find no evidence for this hypothesis (e.g., Lyonette & Crompton, 2015; Simister, 2013).

Importantly, existing studies often use different reference groups (i.e., some studies analyse whether men whose partners earn more than they do carry out more housework than other men, but other studies focus on whether men do more housework than their female partners) and various measures of income. Indeed, it is unclear whether gender inequality in housework is related more strongly to a woman’s absolute income or her income relative to that of her male partner (Hook, 2017). Some authors focused only on relative incomes (Bertrand et al., 2015; Brines, 1994; Nitsche & Grunow, 2016), but others only considered absolute measures (Gupta, 2006; Gupta & Ash, 2008).

The existing empirical research has also pointed to factors other than those related to income that can shape the division of housework within the household. It has, for example, been shown that women’s and men’s education is an important factor in the division of domestic work (Hersch & Stratton, 1994). However, the question arises as to whether and, if so, how the role of educational attainment has been changing given the increasing shares of women with tertiary education (in the population of women) and of better educated women who marry men with a lower level of education than theirs (Hou & Myles, 2008; Nitsche et al., 2018). Gender norms have also attracted economists’ attention in recent years, following the theoretical foundations established by Akerlof and Kranton (2010) and Kranton (2016). The traditional perception of gender roles correlates, e.g., with more unequal division of housework between partners (Coltrane, 2000; Cunningham, 2007), higher gender pay inequality, and lower employment rates of women (Flèche et al., 2020; Fortin, 2005; Steiber & Haas, 2012). Gender norms are also important when it comes to the relationship between the relative income of partners and women’s involvement in housework because they can mediate and moderate the strength of this relationship (Fetterolf & Rudman, 2014; Syrda, 2022). This is primarily because the support for traditional gender roles appears to be more prevalent among women who earn less but at the same time do more housework than their male partners.

Gender Norms and Gender Inequality in Poland

Most of the existing evidence on inequality in income and housework between partners is based on data from Western countries, including the US, Australia, and Western Europe. However, it is important to study various institutional contexts of family policies and labour market institutions, as these play a role in understanding different dimensions of gender inequality (Baxter & Tai, 2016; Hook, 2006; Kil et al., 2016; Steiber & Haas, 2009). In this paper, we provide evidence for Poland. By doing so, we contribute to the existing studies by focusing on an institutional setting that is very different from Western Europe. Like other CEE countries, Poland has a history of relatively high female labour force participation and women having a relatively large domestic work burden. Since the socialist era, women in Poland have generally been perceived as the main providers of child care while simultaneously being expected to work for pay and contribute to the household budget (Pascall & Manning, 2000; Treas & Widmer, 2000). This perception is reflected in public policies in Poland, which encourage mothers to take relatively long parental leave while providing relatively little public child care for the youngest children (Robila, 2012; Szelewa & Polakowski, 2008). Thus, it is expected by both society and the institutional setting in Poland that women take care of their family and household while also working for pay (Martín-García & Solera, 2022). As the option to work part-time is limited in Poland, and there is little working time flexibility (Magda, 2020), many women end up with a double burden of care/household and full-time paid work responsibilities. At the same time, gender inequality is substantial regarding the division of household chores and childcare. On average, women spend over two hours more daily than men on household and family care (even if they also work full-time), which is, for example, twice as much as in Germany (Eurostat, HETUS surveyFootnote 1).

In the 2014 Eurobarometer survey on gender equality the Polish society appears as more conservative than other EU countries, particularly those from Western Europe (European Commission, 2015). Almost 70% of Poles agreed with the statement, “All in all family life suffers when the mother has a full-time job”, compared to 60% in EU28, 64% in West Germany, and 45% in East Germany. The support for the statement "A father must put his career ahead of looking after his young child" is also higher in Poland (40%) than in EU28 (29%) or West and East Germany (25%). Almost 20% of respondents from Poland agreed that men should have more rights to a job than women when jobs are scarce, compared to an EU28 average of 14% and 7% in Germany. Men declare their support for women’s equality in the labour market less frequently than women, partly explaining the mismatch between declarations concerning desired and pursued family models.

Gender wage inequality holds strong across all European countries. However, like many other CEE countries, Poland stands out with a large discrepancy between the raw and adjusted gender pay gap. Average wages paid (hourly) to men and women are quite similar, with a difference of 6–7%. However, once we adjust these for individual and workplace characteristics, the gender pay gap increases above the EU average to 17%–23% (Goraus et al., 2017; Leythienne & Ronkowski, 2018). The fact that women are better educated than men in most CEE countries partially explains why we observe high unexplained gender pay gaps in this region.

Part-time employment is not widespread in Poland (and in the CEE in general), not even among women. Only 10% of Polish women work part-time, compared to over 30% in EU28 on average. The average number of hours worked weekly by women is only 3 h less than men, whereas the EU15 average is 6 h, and it is almost 8 h in Germany. The difference in the gender gap in working hours between the CEE and Western European countries is attributed mainly to lower hourly wages in CEE: one income from a full-time working man is often not enough for the family to ensure their financial needs. Poland’s median hourly earnings in 2014 were 38% of the EU28 average.Footnote 2 The need for second earners explains high female employment rates (prime age) that are seen in virtually all CEE countries, despite inflexible labour markets and the lack of government policies aimed at families and work and family life reconciliation (Matysiak & Węziak-Białowolska, 2016).

We believe this context makes our study particularly interesting, as we consider a setting in which both partners work primarily full time and long hours, women are faced with a double burden of paid work and domestic duties, and there is a limited possibility to outsource these services at the market. In countries studied, the "one and a half earner" model prevailed, with women as secondary workers working part-time.

Data and Methods

Our study uses data from two waves (2010 and 2014Footnote 3) of the Generations and Gender Survey (GGS) for Poland (United Nations, 2005). The GGS collects rich data on, among other information, fertility, partnership histories, household structure, intergenerational relations, and values and attitudes (including individual perceptions of gender norms). It is a panel survey in which one respondent (chosen randomly among adult household members) provides information about the entire household. The GGS is one of few surveys that collect information about the division of unpaid and paid work in the household, the wages of both partners, and gender norms.

Our sample consists of households in which both partners (aged 20–59) were working and were receiving non-zero income. We include households in which the partners were employees (outside of agriculture) but exclude households in which both partners were self-employed, as such couples often claim equal incomes. The final sample is comprised of 4070 couples. Even though the respondent who provided answers to the questions about the household was chosen randomly, there were more female respondents than male respondents in both the total sample and our sample (see Table 1 below). This may be because men were more likely than women to refuse to participate in the survey or because men were less likely to be at home during the interview,Footnote 4 as the interviews were conducted in person.

Table 1 Descriptive statistics
Full size table

We start the analysis by examining the degree of within-household income inequality and present the distribution of this measure across the households. We then investigate the link between the partners’ division of housework and the income inequality between them, which is given by the following equation:

$${Y}_{i}={\alpha }_{0}+{\alpha }_{1}{\left(her\, income\, share\right)}_{i}+\alpha {X}_{i}+ {\varepsilon }_{i}$$

Our dependent variable is derived from three questions asking the respondents who does the specific chore (i.e., cooking, doing dishes, or cleaning) in their household. It is defined as a dummy variable equal to one if the woman was doing most of all three domestic chores and is equal to zero otherwise. We focus on cooking, doing dishes, and cleaning because these household chores are carried out daily. In contrast, other chores, such as making repairs or shopping, are performed less regularly.

We estimate probit modelsFootnote 5 on pooled data from two waves. Our key independent variable is \(her\, income\, share\), which captures the derived within couple income inequality (it is measured as a percentage share of the woman’s income in the couple’s total income). In the regressions, we sequentially control for variables that may influence the relationship between the woman’s share of income and her involvement in housework. We include the following variables that are indicated in Table 1: the sex of the respondent, the partners’ ages, his and her level of education, the presence of children of different ages, the number of years of living together, a marriage indicator, the total number of adults in the household, and living in a rural or an urban area (individual and household characteristics), his and her working hours, and household’s total absolute income.Footnote 6

Additionally, to verify the gender deviance neutralisation hypothesis, we estimate a model allowing the relationship between income inequality and the division of housework to be non-linear. In this model specification, her income share is replaced with three dummy variables that describe the distribution of the within couple income inequality. They are defined as follows: (i) a dummy variable equal to one if, in a given household, her income share is below the median of her income share distribution (which corresponds to her income share being less than 0.43), (ii) a dummy variable equal to one if in a given household her share of income is above the median but less than the 90th percentile of her income share distribution (which corresponds to her income share being between 0.43 and 0.56), and (iii) a dummy variable equal to one if in a given household her income share is above the 90th percentile of her income share distribution (which corresponds to her income share is more than 0.56). The cutoff points (50th and 90th percentiles) are chosen following Hook (2017).

As we are particularly interested in examining whether gender norms can help explain the division of housework among couples, in the subsequent analysis, we analyse individuals’ perceptions of gender equality at work and home. We examine the direct link between gender norms and the within household division of domestic chores and how this association moderates the relationship between the within household income inequality and the woman’s involvement in housework. To measure gender norms, we use the standard questions on work and family life (Fortin, 2005):

  1. (1)

    When jobs are scarce, men should have priority (variable work).

  2. (2)

    A preschool child is likely to suffer if his/her mother works (variable child).

We recode the answers on a five-point Likert-type scale (1-“Strongly Agree” to 5-“Strongly Disagree”) into three categories: agree / strongly agree, neither agree nor disagree (indifferent) and disagree / strongly disagree. To answer the questions on the moderating role of gender norms and the revealed link between the woman’s share of the household income and the gender division of domestic chores, we re estimate the models and interact the indicators of gender norms with the measure of the woman’s share of the household income.

Table 1 presents descriptive statistics of the variables included in the analysis. It presents the data separately for all households regardless of who was the respondent, and depending on whether the respondent was male or female.

Results

The Distribution of Income within Households

The distribution of the within couple income inequality, as measured by the share of the couple’s total income that was earned by the woman, is presented in Fig. 1a. Figure 1b additionally plots the cumulative distribution function of the obtained estimates. The results show that the median share of income earned by the woman was 42.9% (mean of 42.7%). The share of households in which the woman was contributing less than 50% to the total income earned by the couple was 69%. More than 13% of women were earning a share of the household income that was equal to that of the male partner, and in almost 18% of households, the woman was out earning the man. These numbers are in line with the estimates for CEE countries reported by Mysíková (2016).

Fig. 1
figure 1

a Share of income earned by women, distribution b Share of income earned by women, cumulative distribution

Full size image

What were the characteristics of low- and high-inequality households? Couples in which the woman was younger than the man tended to be more unequal (i.e., the woman contributed less than the man to the total household income). By contrast, income inequality was the lowest in couples in which the woman was older than the man. There was also a clear educational pattern: the female share of the total household income was highest among couples in which she had tertiary education, and he did not, and was lowest among couples in which he was tertiary educated but she was not. There was no statistically significant difference in the female share of household income between couples in which both partners had university level education and couples in which neither partner had tertiary education. Finally, the woman’s contribution to the household income was slightly lower in households with children (Table 2).

Table 2 Female share of household income by couple characteristics (mean)
Full size table

The Distribution of Housework within Households

In most surveyed households, the woman mainly did all three analysed tasks (Table 3). Cooking was found to be the most gendered chore. Only or mainly the woman was doing the cooking in 61% of households, while the corresponding values for doing dishes and cleaning were 48% and 53%, respectively. Only 30% of the households reported that the cooking responsibilities were equally divided, compared to 39% for doing dishes and cleaning. The housework was being done mainly or only by the man in only a small percentage of the households (4% to 8%, depending on the task). Outsourcing unpaid work (“someone else does it”) was also rare, with less than 5% of the households reporting outsourcing regardless of the task analysed. In our analysis, we looked at the probability of the woman being heavily involved in the housework, meaning that she was doing all three chores, cooking, dishes, and cleaning, and found that this was true for 30% of the households.

Table 3 The distribution of housework by gender of the respondent
Full size table

For each household, either the male or the female partner answered the questions about the division of housework. When comparing the answers, it is important to note that they referred to different households. The answers were skewed towards the gender of the respondent: both the men and the women who were responding to the questions and providing information about the whole household were more likely to report that they, and not their partner, were doing the housework. However, the differences between answers provided by the male and the female respondents were not big enough to change the general picture described above.

Figures 2 and 3 show the division of housework by her share of the household income divided into three groups that reflect the distribution of the within couple income inequality (< = 50th percentile, between 50 and 90th percentiles, and >  = 90th percentile). Both when all of the chores were analysed separately and when they were combined, the woman’s involvement in housework decreased when her share of income increased. This decrease in the woman’s housework load was largely offset by the increase in the proportion of households in which both partners were doing the chores (Fig. 2). The proportion of households in which he was doing most of the chores also increased, but mainly among households in which the respondent was male. The proportion of households that reported outsourcing the chores did not change as the female share of household income increased.

Fig. 2
figure 2

The division of housework by the gender of the respondent and her share of household income divided into three groups (< = 50th percentile, between 50 and 90th percentiles, and >  = 90th percentile): separate chores

Full size image
Fig. 3
figure 3

The division of housework by the gender of the respondent and her share of household income divided into three groups (< = 50th percentile, between 50 and 90th percentiles, and >  = 90th percentile): heavy female involvement

Full size image

The Link between Housework and Income Inequality

To investigate the link between the female share of household income and her involvement in unpaid work in the household, we regressed the probability that only she was doing all three housework tasks (cooking, doing dishes, and cleaning) against the female share of the household income and a number of explanatory variables. We ran four specifications: (1) one that controlled for the survey wave as well as individual and household characteristics (partners’ ages and education, the presence of children and the age of the youngest child, a dummy for being married, the number of years living together, a dummy for living in a rural area, and the total number of adults in the household), (2) one that added controls for job-related variables (her and his number of hours worked), (3) one that added a dummy for households with a total income below the median to capture the potentially nonlinear association between absolute income and housework inequality, and (4) one in which her income share was replaced with three dummy variables as described in data and methods section. The models were estimated on a total sample and for male and female respondents separately.

We found a strong, negative relationship between the within− couple income inequality and women’s involvement in housework: women who contributed more to the household income were more likely to share the housework more equally (Table 4). Adding individual, household, and job related control variables [columns (2), (6), and (10)] led to a decrease in the size of the effects: the marginal effects of her share of household income decreased by almost 20% when the respondent was female, and by 45% when the respondent was male. Columns (3), (7), and (11) added a control for households with a total income that was below the median income. The marginal effects of her share of the household income increased for both the men and the women and regained statistical significance for the men. The results indicate that a one percentage point increase in her share of household income led to a 0.22% (0.34%) decrease in the probability that she was doing all the chores when the respondent was male (female).

Table 4 Determinants of the probability that mostly she does the chores, marginal effects from probit regressions
Full size table

We also found no support for the gender deviance neutralisation hypothesis. As presented in columns (4), (8), and (12) of Table 4, the probability that the woman does all three chores does not differ for the households in which the woman outearns men (her income share > 0.56) and in which partners’ contribution to households income is more equal (her income share is between 0.43 and 0.56). This is also shown in Fig. 4b, which presents predicted probabilities from the model presented in columns (4), (8) and (12) in Table 4. Based on these results, we conclude that the higher her share of income, the less likely she is to do most of the housework, even if she earns more than her partner.

We recognise that the estimated relationship between her share of the household income and the household’s division of housework may not represent a causal relationship due to the potential problem of reverse causality. On the one hand, it is possible that women's lower earnings lead to them doing more unpaid work. On the other hand, it is equally possible that she is earning less than her partner because she is performing more housework and thus has less time for market work. Nevertheless, our analysis is exploratory, and the results might motivate further efforts to determine the causal relationship between within household income inequality and the division of housework.

Fig. 4
figure 4

a Predicted probabilities from the models in which within− couple income inequality is measured by a continuous measure reflecting her income share (models (3), (7), (11) Table 4). b Predicted probabilities from the models in which within− couple income inequality is measured by three dummy variables describing the distribution of her income share (models (4), (8), (12) Table 4)

Full size image

Gender Norms and Gender Gaps in Housework

A negative link between the female share of household income and the female share of the housework burden can arise due to factors that impact a woman’s wages (and their share in the household budget) and the unpaid work she does in the household. In light of the existing literature, gender norms are among the key factors that may contribute to this negative association. Women with traditional views and attitudes might place a higher value on family and spend more time on housework than women with liberal attitudes, regardless of their relative pay. In parallel, women less attached to traditional values might demand more gender equality in allocating housework at every level of their contributions to the household budget.

To explore whether gender norms matter for the division of housework, we extended the analysis by adding explanatory variables that measured individual perceptions of women in the labour market and at home, i.e., the two variables described in the data section as "work" and "child". It is important to note that the norms refer to individual attitudes, which may have differed between the male and the female respondents. We paid attention to this potential gender difference in our regressions, which were estimated separately for the subsamples of male and female respondents.

The marginal effects obtained from probit models that accounted for gender norms are presented in Table 5. For comparative purposes, the table also includes the marginal effects from the models that did not include measures of gender norms (columns (1) and (4) in Table 5, which correspond to columns (7) and (11) in Table 4). We found that the gender norms were related to the woman's engagement in housework. When the male respondents provided the information, the female partner’s involvement in housework, as measured by the probability that she was doing all three domestic chores, was lower if the man had more liberal views. This was indicated by negative coefficients obtained for those (male) respondents who disagreed with the statement that men should be given priority access to jobs, and that the child suffers when the mother works. The result was stronger and significant at the 10% level for the second measure, while the strength of the relationship was lower and statistically insignificant for the first measure.

Table 5 Determinants of the probability that mostly she does the chores, marginal effects from probit regressions that include measures of gender norms
Full size table

The results obtained for the female respondents confirmed these findings, as the woman’s involvement in housework was lower in the households in which the woman had more liberal gender attitudes. This was especially apparent for the norms measured by the question of whether men should be given priority access to jobs. The negative coefficient of 0.063 indicated that women who disagreed with such a statement were by 6.3 percentage points less likely to do all three domestic chores primarily by themselves.

In addition, we investigated whether the gender norms moderated the observed link between the female share of household income and her share of housework. To investigate whether this was the case, we re ran the analysis while adding an interaction term between the female share of household income and the respondent’s views on gender norms. Because the gender norms were defined as categorical variables, we calculated marginal effects for our main variable of interest (female share of income) at each value of the two gender norms variables. The results are presented in Table 6 for men and women separately. Given the previous findings showing that different measures of gender norms were valid for male and for female respondents (i.e., for women, the more relevant measure of gender norms was the variable reflecting priority access to jobs; while for men, it was a variable reflecting women’s involvement in child care), we only present the interactions of the more relevant gender norms measure.

Table 6 Marginal effects of the association of her share of the household income with the probability that mostly she does the chores, at different values of variables indicating gender norms
Full size table

We found that the negative relationship between the female share of household income and the woman’s involvement in housework (the probability that mainly she was cooking, doing dishes, or cleaning) was absent among the more traditional female respondents, i.e., among those who agreed with the statement that men should have priority access to jobs when they are scarce. At the same time, the relationship remained strong for women who were indifferent about or disagreed with the statement that men should have priority access to jobs. The findings were similar for men, although they were statistically significant only among those indifferent about the effects of a mother working on her children. In general, the results suggest that among less traditional individuals, the greater female share of total household income—which potentially reflected the woman’s improved position in the household and her greater bargaining power—had a stronger negative link to the probability that the woman was the sole provider of housework.

Conclusions

We studied the link between within household gender inequality in income and housework, extending the existing literature by performing an analysis for Poland, a country where female involvement in both housework and paid work is high, and by uncovering the role of gender norms. We used data from the two waves of the Generations and Gender Survey for Poland, which provided detailed information on the characteristics of the households, including their demographic features, incomes, values, attitudes, and various intra household decisions. A few important findings emerged.

First, we found a negative association between the female share of household income and the woman’s involvement in housework. We also found no support for the gender deviance neutralisation hypothesis, as women that earned more than their partners were not increasing their involvement in housework. This is in line with findings from UK and Sweden (e.g., Evertsson & Nermo, 2004; Lyonette & Crompton, 2015; Simister, 2013) but contradicts the results found for the United States and Australia (e.g., Bittman et al., 2003; Evertsson & Nermo, 2004). Second, we found that the respondents’ attitudes towards the positions of men and women at home and in the labour market mattered for the division of housework, regardless of the partners’ respective contributions to the household income. In the households where the respondents expressed more liberal views on women’s labour market and care activities, women were less likely to do most of the housework. The perceptions of gender norms were not only directly associated with the division of housework, but also these perceptions moderated the link between the woman’s contribution to the household income and her probability of doing most of the housework. Among less traditional individuals, the greater female share of total household income had a stronger negative association with women’s involvement in housework. This relationship may have reflected the woman’s increased position in the household and her greater bargaining power. However, among traditional couples, the negative relationship between the female share of income and the female involvement in housework does not hold, suggesting that gender norms are the decisive factor in the division of housework.

From a research perspective, we add to existing studies by showing the need to study gender norms as factors behind the division of unpaid and paid work. Our results also indicate the importance of studying survey data on gender norms and engagement in housework separately for men and women, as the respondents’ answers tended to be biased towards their sex. Finally, we study the relationship between within household gender inequality in income and housework in a different institutional and gender context than studies so far. Poland represents a fairly traditional society, with high female involvement in domestic duties and paid work and institutions that are unsupportive of work-family balance and reconciliation.

Gaining in depth knowledge about how men and women forming a household share housework, what factors explain the division of domestic duties between partners and how these change over time is essential from a practice and policy perspective. More equal sharing of domestic work is perceived as a necessary step towards more gender equality in society and advancement in other dimensions, paid market work in particular. At the same time, as indicated by our results, women’s higher earnings relative to men appear together with men’s greater participation in housework, freeing even more time for women that may be devoted to market work, education, leisure or other activities. This has important policy implications: narrowing gender pay gaps should translate into more equality not only at work but also at home. We also see other practical implications of our analysis. Given that women become less likely to do most of the housework if they earn more, they thus have an even higher incentive to improve their labour market position, as higher absolute earnings should translate into higher income share in the household.

Our study has its limitations, which can be addressed by future research. First and foremost, our results, similar to most studies on the topic, offer no causal interpretations. We can only speak of correlations between the female share of household income and the division of housework without stating the direction of this relationship. Thus, it is unclear whether lower levels of income inequality between the partners contributed to a more equal division of housework or whether more equal shares of housework allowed women to take up a job, likely more demanding but with a higher financial reward. Future research should help to understand the causal relationship between cultural norms, gender attitudes towards domestic and labour market work and gendered labour market outcomes.

Second, the data we used provided information on the declared division of housework but no exact number of hours spent performing housekeeping duties. This limits the comparability of our findings with other studies that used the number of hours spent on housework as their dependent variable in the analysis (e.g., Bittman et al., 2003; Hook, 2017; Syrda, 2022). Furthermore, as other works mentioned above, this paper focused on a limited number of domestic duties: cooking, washing, and doing dishes. There are, however, other activities that are performed daily and which are often neglected in the existing data collections (e.g., dropping off children to schools or after− school classes and back) or which are performed less often (such as repairs or shopping).

Finally, we do not know to what extent our results might be considered more general, particularly for currently male single− earner couples. Women from these couples may decide to enter/ return to the labour market, but it is unclear whether their partners/ husbands will be willing to take up their share of housework. Further studies on gender inequality in paid and unpaid work—and their interlinkages—are needed, especially in the light of the changes in the family models, post-Covid increases in work flexibility, and the envisaged policy changes that aim at increasing men’s involvement in unpaid work and weakening the stigmatisation associated with gender roles and the gendered division of labour (e.g., the 2019 EU work-life balance directive that improves fathers’ access to parental leaves and parents’ working time flexibility; Chieregato, 2020). These changes are likely to influence the within household division of market work and housework and will thus require further studies on how the linkages between income and housework change over time.