The Effect of Changes in Maternal Employment and Family Composition on Children's Behaviour - May 2002

2.1 Family Composition Change and Children's Behavioural Well-being

A large amount of research has investigated the effect that family structure and changes in family composition have upon the well-being of children (Clarke et al., 1998; Jekielek et al., 1998; McLanahan, 1997; Menaghan et al., 1997; and Thomson, 1994). A number of different family structures are considered in this research, including the effect of being in a lone-parent family (with either biological parent) and living in a step-family with one of the biological parents and their partner/spouse. Comparisons are then usually made to children in intact or stable two parent family situations.

Findings from these studies point to changes in the structure of the family as being associated with behaviour problems in children, particularly for young boys (Morrison et al., 1994; Thompson, 1994; Peterson and Zil, 1986). Problems which occur include impulsive/hyperactive behaviour and school behaviour troubles. Alterations to family composition seem to affect girls less, although girls are likely to experience problems when a parent remarries (Peterson and Zil, 1986). Nonetheless, for particular family composition types there may be longer lasting effects. For example, growing up with a divorced or never-married mother, for both boys and girls, is posited to lead to an overall reduction in behavioural and educational well-being (McLanahan, 1997). Further evidence suggests that marital conflict prior to a family disruption is not that important a predictor of behaviour problems. Rather, it is contended, the effect of the marital disruption or change itself is what affects children's behaviour (Morrison et al., 1994).

Research also points to changes in the composition and structure of children's families by divorce, separation or re-marriage, as having a quite profound effect on the economic circumstances of the family (Picot et al., 1999). Divorce or separation of low-income families in particular exposes their children to very high risks of entering poverty, while remarriage dramatically reduces the risk of a child being poor (Picot et al., 1999). Further research indicates that declines in the economic circumstances of children following divorce are linked to behaviour problems (Morrison and Cherlin, 1992). While parental resources (such as engaging and interacting with their children in activities) can help to offset the effect of some of these changes (Clarke et al., 1998), it is also the case that the emotional and economic difficulties surrounding family composition changes can affect the ability of parents to provide support and aid to their children in the short term (Menaghan, Mott and Cooksey, 1997: 4).

2.2 Maternal Employment and Children's Behavioural Well-being

Female participation rates in the Canadian labour market have been growing since the 1960s. Many families have come to rely on dual incomes for economic resources and still more for economic necessity. Much recent research has examined the role that participation in the paid work force by women with children may have on their children's development and well-being (Harvey, 1999; Cook and Willms, 1998; Lefebvre and Merrigan, 1998; Menaghan, Mott and Cooksey, 1997; Cooksey, Menaghan and Jekielek, 1997; Greenstein, 1995; Parcel and Menaghan, 1994; Vandell and Ramanan, 1992; Bayder and Brooks-Gunn, 1991; Belsky and Eggebeen, 1991; Desai et al., 1989).

The timing of a mother's return to employment after the birth of a child is an area of much debate. Vandell and Ramanan (1992), Parcel and Menaghan (1994), and Greenstein (1995) all find that early maternal employment has no adverse effect on children's Peabody Picture Vocabulary Test-Revised scores (PPVT-R) or behaviour problems. However, other researchers (Bayder and Brooks-Gunn, 1991; Belsky and Eggebeen, 1991; Desai et al., 1989) contend that early maternal employment is associated with negative effects on the child's PPVT-R scores and can affect the child's behavioural development. The conclusions of these studies do not provide any simple relationship between maternal employment and children's behavioural well-being — the findings of one study frequently are opposite to that of another, even when based on the same data from the National Longitudinal Survey of Youth (NLSY) in the United States.

Other research work focuses on the effects of maternal employment on older children. Cooksey, Menaghan and Jekielek (1997) investigate the effect on children's behaviour in middle-childhood (those aged from 6 to 11 years) of three aspects of maternal employment conditions: employment status, work stability and occupational complexity. Employment status measures whether a mother was employed or not, work stability measures the number of hours usually worked per week, and occupational complexity measures the content of the work including aptitude, direction, control and planning. They find that children whose mothers have greater work stability and more complex occupational tasks generally experience fewer behavioural problems.

Cooksey, Menaghan and Jekielek (1997: 641) contend that the overall effects of maternal employment/ non-employment are dependent "on the quality and quantity of [parent's] work." Where the paid work of mothers is considered of higher complexity and quality, mothers are more likely to interact with their children warmly and be less concerned with behaviour conformity. When their employment is less desirable or stressful, they are not able to be as responsive to children or provide as stimulating environments. Adverse work and family circumstances, they posit, negatively affect the behavioural well-being of these children. As well, they observe that maternal resources, such as level of education, cognitive level, self-esteem and age, are important for behavioural development in children. These resources not only shape and influence the work conditions in a family, but also have significant direct effects on children's behavioural problems.

A review of much of the literature on maternal employment and children's outcomes by Harvey (1999) endeavours to reconcile many of the conflicting results. The differences in findings for many of the studies are due, Harvey maintains, to differing methodologies even when using the same data sources. In fact, Harvey's own analysis, using recent NLSY data, finds that correlations among the early parental employment variables and child outcome variables show mothers' working early in the child's life is generally associated with more positive child outcomes. More intense maternal employment (working more hours), however, is associated with less positive child outcomes.

After the application of control variables, Harvey finds few simple effects of early parental employment. Among mothers who were employed during the first 3 years of a child's life, the only significant effects were found for those mothers regarding the timing of their return to employment and whether they experienced any breaks in employment during this three year period. Returning to work later and greater continuity of employment in the three year period by the mothers, were both associated with somewhat higher compliance in 3 to 4-year-old children, although this effect was small. Among mothers who were employed during the first 3 years of their child's life, working more hours was associated with significantly lower receptive language abilities — again, however, the effect was small. There are no significant main effects of early maternal employment status. Harvey's results do though, provide partial support for the hypothesis that early parental employment has a positive effect on children's development by increasing family resources, which is found to positively affect children's behaviour problems.

Recent Canadian research on maternal employment, using data from the NLSCY yields results that are similar to those found in the USA. Lefebvre and Merrigan (1998) find that maternal employment and non-employment are not associated with children's cognitive development (using the PPVT for 4 and 5-year-olds). However, maternal full-time work does affect the probability of negative behaviour outcomes for children aged 4 to 11 years. Other effects on children's behaviour such as female family head or step-family characteristics, are found to be relatively more important than maternal employment.

As well, Cook and Willms (1998), using the NLSCY, find that maternal employment status is a factor affecting the amount of time parents have to spend with their children (what they refer to as "engagement"). They find that higher levels of engagement with children are associated with lower levels of negative behaviour disorders and more pro-social behaviours in children. Cook and Willms note, however, that the age of the child is more closely related to the level of parental engagement than maternal employment status. Cook and Willms call for policies that reinforce the role of the family as care provider, including a high level of engagement, while maintaining the family's income level.

In general the studies reviewed have found that a relationship exists between mother's employment and labour market variables, a family's economic circumstances and children's behavioural outcomes. Mothers' employment is associated with an increased likelihood of older children experiencing behavioural problems, but the evidence is less convincing for younger children. The pattern, stability and complexity of maternal employment are found to be important factors affecting the likelihood of behaviour problems in children. As well, children in families under economic pressure find that they face additional problems.

The research work presented here will analyse these general conclusions using data from the NLSCY. As well, information from the longitudinal research work of Menaghan, Mott and Cooksey (1997) will also be considered. These authors contend that maternal characteristics such as level of education, are relatively stable over time and have their strongest effects on children's initial level of behaviour problems, and stabilise the behaviour of children over time. This will be investigated in the multivariate analysis. It is also claimed that social stressors such as changes in employment and family composition, affect the quality of parenting provided to children — through decreasing the amount of time available for parenting and also increasing the stress and decreasing the emotional well-being of mothers. The analysis in this study will investigate, therefore, whether children in family situations where there is an increase in social stressors, through frequent employment transitions or changes in family composition, experience greater behavioural problems. Menaghan, Mott and Cooksey (1997) also claim that mother's employment is beneficial for children's behaviour problems, but that unstable maternal employment (as a social stressor) will diminish these benefits. The analysis here will examine whether relatively stable maternal employment is beneficial for children's behavioural development and if highly unstable employment is relatively worse. Similarly, changes in family composition can be seen to have benefits for family and parental resources and spending time with children, thus diminishing behavioural problems. However, family composition changes can also alter home environments negatively with subsequent problems for children's behaviour. Again, cross-sectional and multivariate analyses in this study will examine whether or not more stable family composition leads to fewer behavioural problems for children.