{PSOCIAL}

The impact of parents' work and employment on family functioning, parenting style and child development has been of interest to scholars for decades. Research has shown that work-family conflict can have negative effects on both parents and children, including increased stress and irritability, low job satisfaction, and mental health issues (e.g., Vahedi et al., 2018, 2019). Moreover, recent studies have found a link between parents' work-family conflict and children's externalizing problems, such as aggressive demeanour and disobedience (e.g., Dinh et al. 2017).

Drawing on Bronfenbrenner's ecological systems theory, this paper will explore the relationship between parental WFC, irritability, and children's externalizing behaviors. Bronfenbrenner (1979) is one of the first scholars who addressed parental work and employment as important environmental predictors of relationships in the family and child development. Ecological systems theory (Bronfenbrenner, 1979) states that children develop within a complex system of connections at various levels of their environment.

This theory addresses differences in parental styles through several systems: the ontogenic system, which includes psychological and emotional traits as well as socio-demographic characteristics of persons; the microsystem, which describes the immediate surroundings of the individual; the exosystem, which includes wider social settings on which the child does not have any influence, but still they affect the child’s experiences; and the macrosystem, which reflects the impact of society norms and values on the everyday life of the family.

The critical element in the ecological systems model is experience (Bronfennbroner, 1989). According to Bronfennbroner, experience indicates that the environment of human development includes objective and subjective features, which are experienced by the persons living in that environment. Both subjective and objective elements in the model drive the course of human development and neither of them alone is assumed sufficient. Over the life course, human development includes processes of complex reciprocal interaction between a human being and the persons and objects in their environment. To be efficient, this interaction must take place on a regular basis over significant periods of time. Bronfenbrenner calls these forms of interaction proximal processes. These processes include feeding and comforting a baby, playing with a young child, child-child activities, group or solitary play, reading, learning new skills, etc. (ibid). For children, participation in such interactive processes gradually generates motivation, knowledge, the ability to do certain things, and the skill to engage in such activities both with other persons and on their own.

For example, through an increasingly more complex interaction with their parents, children increasingly become agents of their own development. Thus, in the current research, I hypothesize that the emotional state of parents caused by work-family conflict can strongly affect the quality of proximal processes, children’s behavior, and outcomes (Hypothesis 1). More recent family-oriented research has also shown a link between parents’ experience of work-family conflict, detrimental family processes, and child mental health outcomes (e.g., Cooklin et al., 2014, 2016; Dinh et al., 2017). I also expect the effect of work-family conflict on externalizing problems to be stronger for mothers than for fathers (Hypothesis 1.1, gender effect hypothesis), as existing research states that mothers' work-family conflict poses a more persistent influence on internalizing and externalizing problems across childhood and adolescence compared to that of fathers’ (Vahedi et al., 2019).

Other powerful forces at work bear a more objective nature and refer to the context in which the family is functioning. These factors may include welfare state support for families with children or the predominant gender ideology of society. They also refer to the family structure. As Bronfenbrenner points out, even in families with favourable socioeconomic circumstances, children of sole mothers or fathers for whom no other person is acting reliably in a “third-party” role are at greater risk of experiencing externalizing problems and a so called “teenage syndrome” – dropping out of school, substance abuse and anti-social behaviour. Most of these effects are more typical for boys than for girls (Bronfenbrenner et al. ,1996). Thus, one of the hypotheses of the current research states that the effect of work-family conflict on externalizing problems of children is more significant for sole-mother families than for intact families (Hypothesis 1.2, family structure hypothesis), as single mothers report higher levels of work-family conflict than other parents (Minnotte, 2012).

Parents are a key influence on children’s and adolescents’ abilities to regulate their emotions (Sheeber et al., 2000). Theoretical perspectives on observational learning/modelling suggest that parents’ personal emotional tendencies implicitly teach children which emotions and self-management strategies are appropriate. By observing their parents’ reactions to provocative emotional situations, children learn what is expected of them in analogous situations in their own lives. In addition, parents who are prone to experience negative emotions may have difficulties regulating those emotions that, in turn, might be transmitted to their children via observational learning (e.g., Morris et al., 2007). Several groups of researchers have found support for a positive association between parents’ (especially maternal) irritability and parents’ difficulties in managing their own negative emotions and harsh parenting (e.g., Saritas et. al, 2013). Other studies suggest that parents’ irritability is related to aggressive behaviors and harsh parenting towards their children (Shay & Knutson, 2008), which, in turn, could influence the development of their children’s externalizing and internalizing problems (Di Giunta et. al, 2021). Taken together, these results support the view that parent irritability leads to greater parental harsh parenting and adolescent irritability (although the latter could also affect the quality of parenting), both of which increase the likelihood of externalizing and internalizing problems (Di Giunta et. al, 2021).

Thus, I expect parents’ irritability to have an effect on children’s negative outcomes (Hypothesis 2, irritability hypothesis). Research shows that previous levels of parenting quality interact with adolescents’ problem behavior in predicting parenting quality later: higher levels of problem behavior have negative effects on parenting, especially when parenting is already less adequate (Reitz et al. 2006).

Historically Australia has been progressive in terms of its development of both employment rights and women’s citizenship rights (Newsome, 2017). However, until 2010 Australia lagged behind every other Organization for Economic Cooperation and Development nation, except for the US, in introducing a national paid parental leave scheme. Brennan (2009) argues that the lack of a social insurance scheme, and the continued influence of the male breadwinner policy model impeded the introduction of the paid parental leave.

Australian social policy has been premised on the male breadwinner/female caregiver dichotomy (O’Connor et al., 1999). The ideal of the male breadwinner was built into foundational institutions of the Australian welfare state such as the “family wage”. A ban on married women’s employment in the Commonwealth public service was lifted only in 1966 and eligibility for most income support payments continues to be assessed on family rather than individual income (Mahon et. al, 2016). Debates about childcare and parental leave policy reflect continuing tensions between supporters of the dual-earner family and institutions and practices that pull back towards more traditional patterns.

Overall, over the past 40 years, there have been significant changes in the composition of Australia’s active workforce. There have been large increases in the employment participation of women, who comprised 48% of the total employed in 2022, compared to 36% in 1979 (Australian Institute of Family Studies, 2023). Additionally, the proportion of women who were employed full-time increased from 26% to 34% over the same period (ibid). This trend can be attributed to changing societal attitudes towards women's roles in the workforce, increased access to education and training, and changes in government policies and workplace practices aimed at promoting gender equality. However, it is important to note that there are still significant gender disparities in employment outcomes in Australia. Although the employment rate of mothers with children under the age of 5 increased from 30 to 66% between 1981 and 2021, those with younger children are more likely to work part-time than full-time (LSAC, 2018).

At the beginning of the 21st century the male breadwinner model was still relatively strong, and women’s citizenship was defined by their role as mothers (Pocock, 2005). Nowadays dual-earner families are predominant. In intact families studied in LSAC the employment rate of fathers was quite high and stable during the waves 5-8, at the level of 94-95%. The level of mothers’ employment was relatively high as well, but the significant part of them worked part-time, and their share did not decrease with child’s age. One of the explanations for this could be that women are more likely to go part-time than men because in the absence of state support it enables them to combine work and family and caring responsibilities. The growing share of men who work part-time (from 11.5% in 2012 to 13.8% in 2018) could be the consequence of changes in the labour market (Australian Government, 2020). Due to increased competitive pressures and changes in technology, employment forms have become more flexible, which resulted in the emergence of many new part-time jobs. It may also be due to the fact that men began to take on more housework or caring responsibilities than they did decades ago.

The share of male breadwinner families with children in 2021 did not exceed 15% (Australian Bureau of Statistics, 2021). Still, male breadwinner model is proving to be quite resilient in Australia (Baxter and Hewitt, 2013). The disproportionate female share of part-time employment has helped entrench male breadwinners. Even though women have greatly increased their involvement in paid work over several decades, mothers still tend to be secondary earners.

In that category, the rate is about 52%. This may imply that Australia still has a strong male-breadwinner institutional framework that continues to hinder gender equality in paid and unpaid work (Baxter & Hewitt, 2013).

Literature on gender division of unpaid domestic duties shows that employed mothers in Australia typically deal with a disproportionate share of household and childcare responsibilities in comparison to employed fathers, they are more likely to bear the majority of the caregiving tasks, and their experience of work-family conflict may be particularly important for family functioning and children’s outcomes (Westrupp et al., 2016). In this respect, it is interesting to compare the effect of mothers’ and fathers’ WFC on their children’s externalizing problems in different types of families.

The dependent variable in the models is child’s externalizing problems. Those are mother-reported and calculated as an average of the answers to the statements: “My child is constantly fidgeting”, “My child often fights/bullies other youth”, “My child is easily distracted”. The answers range from 1 (not true) to 3 (certainly true). The reason behind using only mothers’ estimation of children’s externalizing problems is that fathers’ answers had a lot of missing values. I compared the mean estimation of children’s externalizing problems among mothers and fathers, and the difference was very small.

The independent variables are work-family conflict and parents’ irritability, and their interaction. In LSAC parents rated the degree to which employment-related constraints impacted their family life and parenting (e.g., “Because of my work responsibilities my family time is less enjoyable and more pressured”). Items were rated on a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree), and a work-family conflict score is calculated by averaging all the items, which was recoded into a score from 1 (low WFC) to 3 (high WFC) for the regression analysis.

Parents’ irritability was reported on the frequency of their hostile, angry or rejecting behaviors toward the child using the question “How often are you angry when you punish this child?” with a five-point scale ranging from 1 (never) to 5 (all the time) and was recoded to the score 1 (low irritability) to 2 (high irritability) for the regression analysis.

I control for: children’s characteristics (gender, victimization experience); parents’ personal characteristics (education); work characteristics (full-time/part-time status); family characteristics (household income, presence of siblings).

he method used is a modification of random effects that allows for diff

The method used is a modification of random effects that allows for differentiation of within and between effects (REWB) (Allison, 2009; Bell et. al, 2019). The reason for choosing this method is the following. In longitudinal data, such as LSAC, individuals or social units are measured on a number of occasions, which means occasions (at level 1) are nested within the individual or entity (at level 2). Thus, we can have “within” effects that occur at level 1, and “between” effects that occur at level 2 (Howard, 2015; Bell et. al, 2019), and these two different effects should not be assumed to be the same. Specifically, the impact of constant, between-person sources of variation must be differentiated from the impact of time-specific, within-person sources of variation. As Hoffman and Stawski (2009) put it, “persons should be modelled as contexts”.

Common random effects (RE), though they can model the influence of time-constant covariates, like child’s gender or parents’ level of education, on the variable of interest, assume that between and within effects are equal, which may not be the case. Fixed effects model (FE) provides an estimate of the within effect but can say nothing about relationships with independent variables that do not change over time—only about deviations from the mean over time. FE specification reveals almost nothing about the level-2 entities in the model (Bell et. al, 2019). REWB model gives exactly the same results for the within effect (coefficient and standard error) as the FE model but retains the between effect which can be informative and cannot be obtained from a FE model (ibid). Thus, REWB is of more encompassing nature than either RE or FE, with heterogeneity modelled at both the individual (level 2) and occasion (level 1) level.

The dependent variable (i.e., child’s externalizing problems) is regressed onto WFC and parents’ irritability.