Growing Up in Scotland: patterns of maternal employment and barriers to paid work

5 Employment Trajectories And Characteristics Of Mothers Who Left Work After Having A Child

5.1 Introduction

One of the main benefits of using data from a longitudinal study like GUS is that it enables us to look at the circumstances of the same individuals over time. While the initial chapters of this report are concerned with the characteristics and circumstances of mothers of children of a particular age across specific calendar years, this chapter focuses on the circumstances of the same groups of mothers over time.

Firstly, it considers the employment trajectories of mothers in each of the two GUS cohorts from their child’s birth up until the child was aged 5. This kind of analysis can help us understand more about how mothers move in and out of work following childbirth, as well as how this may differ between mothers who had a child in 2004/05 (mothers in BC1) and mothers who had a child six years later (mothers in BC2).

The chapter then focuses on a group of mothers who followed one specific trajectory – those who had a paid job while they were pregnant, gave up work after having a child and had not returned by the time their child turned 5. For these women, childbearing appears to have been of particular significance to their employment trajectory and they are therefore of particular interest to policy makers interested in the barriers facing mothers of young children who want to engage in paid work.

Existing research has shown that mothers who worked during pregnancy are much more likely to be in work after the birth (e.g. Chanfreu et al., 2011). GUS data contains information about whether mothers worked during pregnancy as well as about their employment status at the time the child was aged 10 months, 3 years and 5 years. It is therefore possible to look at whether mothers who worked during pregnancy were also more likely to remain in work until their child reaches school age.

The reasons why women give up work after having a child are likely to be many and varied. However, if the aim is to support mothers to be able to remain in work after childbirth, it is important to understand why mothers leave paid work. For example, are the mothers who give up paid work mainly those who can afford not to have a paid job and who can comfortably choose to stay at home to look after their children? Or are they primarily less advantaged mothers confronted with a number of the barriers often associated with living in more chaotic circumstances, such as low levels of skills and education, low pay and a lack of flexibility in the jobs available to them?

A useful first step towards answering questions like these is to understand more about the mothers who give up paid work after having a child, and how they compare with the mothers who remain in or re-enter work. To do so, this chapter looks at characteristics of mothers in Scotland who left work after having a child and had not returned by the time their child was aged 5. It compares these to the characteristics of mothers who either remained in or returned to work during this period and identifies key differences between the two groups. It also identifies characteristics which are independently associated with mothers giving up work.

5.2 Key variables used in the analysis

5.2.1 Employment trajectories

The employment trajectories looked at in the initial part of this chapter were developed using data collected from mothers in the two GUS birth cohorts when the cohort child was aged 10 months, 3 years and 5 years. Only cases where the child’s mother was the main respondent at all three sweeps of data collection were included in the analysis. Initially, variables were derived which identified a total of eight combinations (or trajectories) of mothers’ employment statuses across the three sweeps ^[24] . For simplicity, these were collapsed into the following four overall trajectories:

• Mothers who were in paid work at all three sweeps, namely at the time the cohort child was aged 10 months, 3 years and 5 years.
• Mothers who were eventually in paid work by the time the child was aged 5, but who were not in work when the child was younger (these mothers were out of work at the time their child was aged 10 months or 3 years, or both).
• Mothers who were in paid work at some point when their child was younger (i.e. when the child was aged 10 months and/or 3 years), but were not in work at the time the child was aged 5.
• Mothers who were not in paid work at any of the three sweeps.

These ‘trajectories’ are based on information collected about mothers’ employment status at specific time points. Mothers who were in work at all three sweeps are assumed to have been in work continuously until the child was aged 5, while those who were out of work at all three sweeps are assumed to have been continuously out of work during this time ^[25] . The trajectories analysis does not take into account any other children the mother may have – the 'child' referred to in the text is the GUS cohort child ^[26] .

5.2.2 Characteristics of mothers who left paid work after having a child and had not returned by the time the child was aged 5

The second part of the chapter uses data from BC2. As above, only cases where the child’s mother was the main respondent at all three sweeps were included in the analysis. The main group of interest in these sections are mothers who gave up work after having a child and had not returned by the time the child was aged 5. Throughout the sections, these mothers are compared with mothers who remained in or returned to work during this period. Definitions of these groups are provided in Table 5‑1.

Table 5‑1: Definitions of groups of mothers considered in the analysis

Group of mothers	Definition	Sample sizes (unwtd)
Mothers who left work	Mothers who worked while they were pregnant with the cohort child but were not in work at any of the time points GUS data were collected after the child’s birth (i.e. not in work when the child was aged 10mths, nor when they were aged 3yrs nor 5yrs).	177
Mothers who remained in or returned to work	Mothers who worked while they were pregnant with the cohort child and were in work at at least one time point after the child’s birth (i.e. when the child was aged 10mths and/or 3yrs and/or 5yrs).	2804

Existing research has suggested that a range of household and individual characteristics are likely to be important when considering mothers’ engagement in paid work after childbirth (e.g. Chanfreu et al., 2011; Fagan and Norman, 2012; Smeaton, 2006). In line with this, a number of characteristics and circumstances were considered in the analysis undertaken for this chapter. To ensure consistency, and to enable meaningful predictive analysis, only information on characteristics and circumstances collected at the time the child was aged 10 months was included in the analysis ^[27] .

5.3 Key findings

• Around half of mothers in both cohorts were in work at all three points up until their child was aged 5. Mothers in BC2 were more likely than mothers in BC1 to be in work at all three sweeps (53% of mothers in BC2 compared with 48% in BC1).
• Between one in five and one in four of all mothers were not in work at any of the three time points. The proportion of mothers who were not in work at any point dropped between the two cohorts, from 24% in BC1 to 21% in BC2.
• Compared with all mothers, mothers who worked during pregnancy were more likely to be in work at all three time points (i.e. when their child was aged 10 months, 3 years and 5 years). Around three quarters of mothers who worked during pregnancy fell into this category, compared with around half of all mothers. Mothers who worked during pregnancy were also less likely to be out of work at all three age points. These patterns were evident in both cohorts and are in line with previous research.
• Only a small minority of mothers who worked during pregnancy had left work and not returned by the time the child was aged 5 (10% in BC1 and 7% in BC2). Mothers in BC2 were less likely than mothers in BC1 to have given up work, though differences are small.
• Overall, mothers who gave up work tended to live in less advantaged circumstances than mothers who remained in or returned to work. For example, those who left work were more likely to be younger, to be single mothers, and to be living in the most deprived areas. In contrast, they were less likely to have a degree or to have been working in professional or managerial occupations. And as may be expected, having left paid work, these mothers were also more likely to be in low income households.
• Among mothers living with a partner who was in work, those who gave up work were more likely than those who remained in or returned to work to have a partner who was on a low income but were just as likely to be living with a partner on a high income.
• Being a single mother, having lower levels of educational qualifications, having another child before the cohort child turned 5, and the cohort child having a long-term health condition all independently predicted giving up paid work.

5.4 Mothers’ employment trajectories

Figure 5-A shows the proportion of mothers in each cohort who followed each of the employment trajectories outlined above. It shows figures for all mothers and for mothers who worked while they were pregnant with the cohort child. Looking at all mothers, Figure 5-A illustrates that, in both cohorts, around half of mothers – the largest group – were in work at all three sweeps. A further one in six mothers (17% in BC1; 16% in BC2) were in work at the time their child was aged 5, but had been out of work when their child was 10 months and/or aged 3. Thus, the majority of mothers who were in work at the time the child was aged 5 were also in work when the child was younger.

Mothers in BC2 were more likely than mothers in BC1 to be in work at all three time points (53% in BC2 compared with 48% in BC1).

Around one in five mothers were not in work at any of the three time points. The proportion of mothers who were not in work at any point dropped between the two cohorts, from 24% in BC1 to 21% in BC2. Overall, around one in three mothers were not in work at the time the cohort child was aged 5.

Compared with all mothers, mothers who worked during pregnancy were much more likely to be in work at all three time points and less likely to be out of work at all three points. For example, 70% of mothers in BC2 who worked during pregnancy were in work at all three points compared with 53% of all mothers. Conversely, only 7% of mothers who worked during pregnancy were not at work at any point during the first 5 years of the child’s life, compared with 21% of all mothers (Figure 5-A).

In both cohorts, only a small minority of mothers who worked during pregnancy had left work and not returned by the time the child was aged 5. Mothers in BC2 were less likely than mothers in BC1 to have left work and not returned by the time the child was aged 5: 7% of mothers in BC2 had done so compared with 10% of mothers in BC1. In total, only around one in five mothers who worked during pregnancy were not in work by the time the cohort child was aged 5. This is in line with existing research which has found that working during pregnancy is a strong predictor of mothers’ post-birth labour market participation (e.g. Chanfreu et al., 2011; Fagan and Norman, 2012).

Figure 5‑A: Mothers’ employment trajectories, by cohort and by whether mother did paid work while pregnant

Bases: All mothers who took part at all three sweeps / Mothers who worked during pregnancy and took part at all three sweeps. Base sizes (unweighted/weighted): All mothers: BC1=3502/3495; BC2=4088/4078. Mothers who worked during pregnancy: BC1=2712/2553; BC2=3238/3019.

5.5 Characteristics of mothers who left paid work after having a child and had not returned by the time the child was aged 5

5.5.1 Age, education and social class

Mothers who gave up work after having a child and had not returned by the time the child turned 5 tended to be younger than mothers who remained in or returned to work. For example, 11% of mothers who left work were under 20 at the time of their child’s birth, compared with just 3% of mothers who remained in or returned to work. In contrast, 54% of mothers who remained in or returned to work were aged 30 or over compared with just 33% of mothers who left work (Figure 5-B).

Compared with older mothers, younger mothers have had less time to establish themselves in the labour market before having children, something which has previously been shown to be associated with remaining in employment after having a child (see e.g. Chanfreu et al., 2011). From previous analysis of GUS data (e.g. Bradshaw et al., 2014b) we also know that young mothers are more likely to live in disadvantaged circumstances and as such this difference may be a reflection of the additional barriers young mothers face, rather than simply their age. Indeed, additional analysis suggested that being a young mother is not in itself predictive of giving up paid work after childbirth. Further details are provided in section 5.6.

Figure 5‑B: Mother’s age at child’s birth, by whether mother left or remained in/returned to work by time child aged 5

Base: Mothers who were in paid work during pregnancy. Base sizes (unweighted/weighted): Left work=177/216; Remained in or returned to work=3061/2804.

Mothers who gave up work tended to have lower levels of education than those who remained in work (Figure 5-C). In particular, degree level qualifications were much less common among mothers who left work: just 16% of these mothers held a degree level qualification compared with 43% of mothers who stayed in work. In contrast, 9% of mothers who gave up work had no formal qualifications compared with just 3% of those who remained in work.

Figure 5‑C: Mother’s highest level of education when child aged 10 months, by whether mother left or remained in/returned to work by time child aged 5

Base: Mothers who were in paid work during pregnancy. Base sizes (unweighted/weighted): Left work=177/216; Remained in or returned to work=3058/2802.

At least in part, this difference by education may be explained by mothers with degree level qualifications being more likely to be in occupations with career ladders and deferred rewards – aspects which increase the opportunity costs of not returning to work (Smeaton, 2006). Being in different types of jobs than those with lower qualifications – with differing responsibilities and greater levels of autonomy – it is possible that mothers with higher levels of education experience greater levels of satisfaction and enjoyment from the work they do, acting as a stronger motivation to return.

Indeed, as Figure 5-D shows, mothers who left work were considerably less likely to be in professional or managerial occupations ^[28] : just 18% of these mothers were in professional or managerial occupations compared with 46% of mothers who remained in or returned to work. In contrast, mothers who gave up work were much more likely to be in routine or semi-routine occupations – 53%, compared with 25% of mothers who remained in work.

Figure 5‑D: Mother’s occupational classification ( NS-SEC), by whether mother left or remained in/returned to work by time child aged 5

Base: Mothers who were in paid work during pregnancy. Base sizes (unweighted/weighted): Left work=177/216; Remained in or returned to work=3056/2800.

Previous analysis of GUS data suggested that those in routine/semi-routine jobs are less likely to have access to flexible working facilities (Dean et al., 2017). Not having access to flexible working is likely to make it more difficult for mothers to combine work and caring responsibilities. Thus, this may have had an impact on mothers’ decision on whether to remain in work. And indeed, mothers who gave up work were less likely to say they had access to one or more family-friendly working facilities ^{[29] , [30]} : 54% of these mothers said they had access to some form of family-friendly working facilities, compared with 84% of those who remained in or returned to work (Figure 5-E).

Figure 5‑E: Proportion of mothers who had access to family-friendly working facilities in job held during pregnancy, by whether mother left or remained in/returned to work by time child aged 5

Base: Mothers who were working as an employee during pregnancy and did not change jobs in first 10 months after child’s birth. Base sizes (unweighted/weighted): Left work=154/188; Remained in or returned to work=2786/2528.

5.5.2 Household income and managing financially

Mothers who gave up work and had not returned by the time their child was aged 5 tended to live on lower incomes at the time the child was 10 months than mothers who remained in work or had returned to work during this period (Figure 5-F). For example, at the time their child was aged 10 months, 40% of the mothers who left work were living in the poorest fifth of households – compared with just 13% of the mothers who stayed in or returned to work.

The lower level of household income seen among mothers who left work is likely to be at least partially explained by the fact that the household income measure is closely related to a mother’s employment status – mothers who are in paid work are likely to have higher incomes than those not working. Nonetheless, household income is also likely to be strongly associated with things like mothers’ social class and hourly earnings. We can therefore use it as a proxy measure of the economic context in which mothers make decisions about whether to return to work. For example, if we assume that mothers in lower income households are more likely to be in lower paid jobs, for these mothers, remaining in or returning to a low-paid job may not be worthwhile if childcare costs exceed the amount they would earn. Also, as noted above, previous analysis of GUS data showed that parents on lower incomes are less likely than those on higher incomes to have access to family-friendly working facilities (Dean et al., 2017). Conversely, assuming that mothers in higher income households are more likely to be in well-paid jobs with access to family-friendly working facilities, these mothers may find it easier to combine paid work with childcare responsibilities. Furthermore, paying for childcare is likely to be less of an issue for mothers in higher income households than for mothers in low income households – although previous GUS analysis has shown that families in more disadvantaged circumstances are more likely to use informal (i.e. free) childcare (Bradshaw and Wasoff, 2009).

Figure 5‑F: Equivalised household income when child aged 10 months, by whether mother left or remained in/returned to work by time child aged 5

Base: Mothers who were in paid work during pregnancy. Base sizes (unweighted/weighted): Left work=154/188; Remained in or returned to work=2786/2528.

Mothers’ subjective perceptions of how the family is managing financially may also play a role in decisions about whether or not they remain in or return to work after having a child. Figure 5-G shows the distribution of how mothers said their household was getting on financially at the time the child was aged 10 months.

The difference in reporting ‘not managing very well’ between mothers who returned to work and those who did not was not statistically significant. However, mothers who left work were less likely than mothers who remained in or returned to say that they were managing ‘well’ or ‘quite well’. 37% of mothers who gave up work said this compared with 49% of mothers who stayed in or returned to work. This reflects the same general pattern seen in the household income data that mothers in poorer financial circumstances are more likely to give up work.

Figure 5‑G: How household is managing financially when child aged 10 months, by whether mother left or remained in/returned to work by time child aged 5

Base: Mothers who were in paid work during pregnancy. Base sizes (unweighted/weighted): Left work=177/216; Remained in or returned to work=3057/2800.

5.5.3 Family characteristics

Compared with mothers who stayed in or returned to work, mothers who gave up work were much more likely to be lone parents: 37% of mothers who gave up work were lone parents at the time the child was aged 10 months, compared with just 14% of those who returned to work (Figure 5-H). This is perhaps unsurprising given that single parents do not have another adult carer in the household to care for the child and only one adult to generate income. Thus, organising and paying for childcare is likely to be a more significant issue for single mothers compared with partnered mothers. To some extent, this argument is supported by the findings in chapter 4 which suggested that at the time their child was aged 10 months, single mothers who were looking for work were more likely than partnered mothers to report childcare issues as one of the main reasons they had not found work. Notably, no differences were evident among mothers of 3 and 5 year old children. While overall the proportion of mothers who referenced childcare as a barrier increased as the child aged, it is possible that single mothers are more sensitive to the cost of childcare. Thus, the likely reduction in childcare costs as the child reached pre-school age may have had more of an influence on single mothers than on partnered mothers ^{[31], [32]}.

Figure 5‑H: Whether mother is living with a partner when child aged 10 months, by whether mother left or remained in/returned to work by time child aged 5

Base: Mothers who were in paid work during pregnancy. Base sizes (unweighted/weighted): Left work=177/216; Remained in or returned to work=3061/2804.

Interestingly, mothers who lived with a partner who was not in work at the time the child was aged 10 months were just as likely to leave work as those living with a partner who was in work (Table 8-2 in Appendix B). Analysis restricted to mothers whose partners were in work at the time the cohort child was aged 10 months showed that partners of mothers who gave up work tended to earn less than partners of mothers who remained in or returned to work. For example, among mothers who lived with a partner who was in paid work, 27% of those who left work lived with a partner who earned less than £12,000 per year, compared with just 16% of mothers who remained in or returned to work (Figure 5-I). Interestingly, mothers who left work were just as likely as those who remained in or returned to work to be living with a partner in the highest income group: among mothers who lived with a partner who was in paid work, 18% of those who left work lived with a partner earning more than £31,200 per year compared with 15% of mothers who remained in or returned to work ^[33] .

Figure 5‑I: Partner’s annual take-home pay when child aged 10 months, by whether mother left or remained in/returned to work by time child aged 5

Base: Mothers who were in paid work during pregnancy and were living with a partner who was in paid work at the time the child was aged 10 months and where information about the partner’s income was provided. Base sizes (unweighted/weighted): Left work=92/95; Remained in or returned to work=2045/1756.

The analysis also showed differences according to the partner’s socio-economic status. Among mothers whose partner was in work, those who gave up work were less likely than those who stayed in or returned to work to be living with a partner in a professional/managerial occupation (31% compared with 46%). In contrast, partnered mothers who left work were more likely than those who remained or returned to be living with a partner in a routine or semi-routine occupation (33% compared with 20%, Figure 5-J).

These figures suggest that a partner’s socio-economic circumstances are likely to play a role in whether a mother leaves work or not after having a child. However, they also suggest that this role varies significantly for different mothers. Thus, multiple (and at times contradictory) patterns seem to be at play. For example, for some mothers who are living with a partner who is not in paid work, engaging in paid work may be a financial necessity. At the same time, living with a partner who is not in paid work may simply be another indicator that a mother is living in circumstances which make labour market participation more difficult for both her and her partner (for example, having low levels of skills and education). Conversely, mothers living with a partner in well-paid work have a higher level of flexibility to choose not to engage in paid work, if they prefer, compared with mothers living with a partner who is not working or is on low pay. At the same time, mothers living with a partner in a professional/managerial job may be more likely to be in professional/managerial jobs themselves and therefore be less likely to give up work.

Figure 5‑J: Partner’s occupational classification when child aged 10 months, by whether mother left or remained in/returned to work by time child aged 5

Base: Mothers who were in paid work during pregnancy and were living with a partner who was in paid work at the time the child was aged 10 months. Base sizes (unweighted/weighted): Left work=121/122; Remained in or returned to work=2652/2282.

In addition to considering the presence of a partner in the household it is also worth looking at the presence of any siblings. The presence of older children does not seem to be associated with a mother’s likelihood of giving up or staying in work. At the time the cohort child was aged 10 months, there were no differences in the number of children living in the households of mothers in the ‘left work’ group and mothers in the ‘remained or returned to work’ group. Given the additional caring responsibilities, we might have expected mothers with a higher number of children to be more likely to give up work.

It is worth remembering, however, that at 10 months the cohort child was (in the vast majority of cases) the youngest child in the household and was therefore likely to require more intensive care than any older siblings. Indeed, there were indications of a relationship between a mother having another – younger – child and giving up work. 48% of mothers who gave up work had a younger child living with them at the time the cohort child was aged 5, compared with 41% of mothers who remained in or returned to work. This difference was only borderline significant (p=.073). This suggests that while the presence of younger siblings may be part of the picture, it is unlikely to be the main factor driving mothers to give up work.

Figure 5‑K: Whether younger child in household when cohort child aged 5, by whether mother left or remained in/returned to work by time child aged 5

Base: Mothers who were in paid work during pregnancy. Base sizes (unweighted/weighted): Left work=177/216; Remained in or returned to work=3061/2804.

5.5.4 Area characteristics

Mothers who left work tended to live in areas with higher levels of deprivation than mothers who remained in or returned to work (Figure 5-L). For example, at the time the child was 10 months old, 34% of mothers who gave up work were living in the most deprived areas, compared with just 19% of mothers who stayed in or returned to work. Conversely, just 10% of mothers who left work lived in the least deprived areas compared with 21% of mothers who did not.

This confirms the suggestion made by other research (e.g. Chanfreu et al., 2011) that mothers living in less advantaged circumstances are more likely than those in more advantaged circumstances to leave work after having a child. It is possible that the jobs available in the most deprived areas are of a nature which makes it more difficult to combine paid work with having a young child. For example, jobs in these areas may be more likely to be lower-skilled and therefore also to be low-paid and offer less flexibility than the jobs available to mothers living in less deprived areas. However, multivariable analysis which controlled for the influences of other factors found no independent relationship between the level of area deprivation and the likelihood of mothers giving up work. This suggests that the differences by area deprivation seen in Figure 5-L are explained by other differences between mothers in the two groups (further details in section 5.6).

Figure 5‑L: Level of area deprivation ( SIMD quintiles) when child aged 10 months, by whether mother left or remained in/returned to work by time child aged 5

Base: Mothers who were in paid work during pregnancy. Base sizes (unweighted/weighted): Left work=177/216; Remained in or returned to work=3061/804.

Whether mothers left work did not appear to be associated with whether they lived in urban or rural locations (Table 8‑3 in Appendix B).

5.5.5 Maternal and child health

Mothers who gave up work were neither more nor less likely to suffer from a long-term health condition than mothers who stayed in or returned to work. At the time the cohort child was 10 months old, 13% of mothers who left work reported having a long-term condition, compared with 11% of mothers who remained in or returned to work (Figure 5-M). While this may seem counter-intuitive, it is worth remembering that the analysis was restricted to those who were in work during pregnancy. As such, that we do not see any differences is likely to simply reflect the fact that mothers with the most severe health conditions were less likely to be in work during pregnancy (and therefore not included in the analysis).

The child’s health appears to have more of an impact. Mothers who gave up work were slightly more likely than those who remained or returned to report that their child had a long-term condition at the age of 10 months. 16% of mothers who gave up work said this, compared with 11% of mothers who remained in or returned to work (Figure 5-M). It is worth noting that the measure covers a range of conditions, not all of which are likely to require much additional care. However, some of the conditions undoubtedly do and even though the difference seen here is relatively small and is only borderline significant (p=.073), having a child with severe health problems, especially those which require constant care, is likely to have an impact on a mother’s likelihood to give up work.

Figure 5‑M: Proportion of mothers/children with long-term health condition when child aged 10 months, by whether mother left or remained in/returned to work by time child aged 5

Base: Mothers who were in paid work during pregnancy. Base sizes (unweighted/weighted): Left work=172/211; Remained in or returned to work=3044/2785.

5.6 Main factors associated with leaving work

A number of the characteristics and circumstances considered above are likely to be strongly interrelated. This means that, on the basis of the bivariate analysis outlined above, we cannot be sure which characteristics and circumstances are themselves associated with mothers giving up work, and which are related primarily due to their relationship with another characteristic. To address this, multivariable logistic regression models were fitted to identify characteristics and circumstances which were independently associated with a mother leaving work after having a child and not having returned by the time the child was aged 5. The results are discussed below.

5.6.1 All mothers

The bivariate analysis suggested that, compared with mothers who remained in or returned to work, mothers who gave up work were likely to be younger and to be living in more deprived areas. They were also more likely to be single mothers, and less likely to have a degree or to have been working in professional or managerial occupations. Finally, whether a mother had another child within the first five years of the cohort child, and whether the cohort child had any long-term health problems or disabilities also seemed to play a role.

These characteristics and circumstances were entered into a multivariable logistic regression model which predicted a mother leaving work after having the cohort child and not having returned by the time the child turned 5 ^[34] .

The results indicate that being a single mother, having low levels of educational qualifications, having another (younger) child, and the cohort child having a long-term health condition were all independently and positively associated with a mother giving up paid work. For example, the odds of leaving work after having a child and not having returned by the time the child turned 5 were almost four times higher for mothers who had no educational qualifications than for mothers who had a degree. The odds of single mothers leaving work were twice as high as those for partnered mothers.

As a whole, a mother’s occupational classification was not significant in the model. Nonetheless, there were some indications that mothers in routine/semi-routine occupations were more likely than mothers in professional/managerial occupations to give up work.

Neither the level of area deprivation nor the mother’s age were independently associated with a mother giving up work. Thus, the associations found in the bivariate analysis are likely explained by the fact that mothers who have lower levels of education – a characteristic which was independently associated with mothers being more likely to give up work – are also likely to be younger in age, and to be living in the more deprived areas.

As discussed in section 5.5.2, the bivariate analysis suggested that mothers living in the poorest households were more likely to give up work after having a child. Unsurprisingly (given the reasons set out in section 5.5.2), living in the poorest households remained strongly and positively associated with a mother leaving work, even when controlling for the influence of other factors. Notably, however, the presence of a partner became less significant once household income was added to the multivariable model ^[35] . This could suggest that the partner’s income plays an important role – something which is considered further in section 5.6.2.

The bivariate analysis also indicated that mothers who did not have access to any family-friendly working facilities were more likely to give up work after having a child (cf. section 5.5.1). To test whether this relationship remained significant when controlling for the influence of other factors such as the type of job and the mother’s level of education, whether the mother had access to family-friendly working facilities was added to the multivariable model (after household income had been removed). Because of the way the information about family-friendly working facilities was collected ^[36] , this analysis was restricted to mothers who did not change jobs during the first ten months after the cohort child’s birth, and to mothers who worked as employees while they were pregnant (i.e. mothers who were self-employed are not included). The results ^[37] show an independent relationship between not having access to any family-friendly working facilities and giving up work after having a child, with odds of leaving work for mothers who had no access to family-friendly working facilities being more than one and a half times as high as for mothers who did have access (OR=1.7). When interpreting these results it is important to bear in mind the possibility that mothers who were in work at the time of the interview may be more likely to be aware of any family-friendly working facilities offered to them than mothers who had not worked since having a child. Nonetheless, the results seem to indicate that access to family-friendly working facilities is a factor which should not be overlooked when considering circumstances influencing mothers’ engagement in paid work after having a child.

5.6.2 Partnered mothers and single mothers

As previously suggested, their partner’s income is likely to have at least some bearing on mothers’ decisions about whether or not to remain in or return to work after childbirth. To test this, a separate multivariable model was fitted for mothers who were living with a partner. The model included all the individual and household variables included in the initial model, except whether a partner was present in the household. In addition, this model included partner-specific variables such as the partner’s socio-economic class and their annual take-home pay ^[38] .

Some of the results were similar to those found for all mothers. For example, a mother having lower levels of education and having another (younger) child were both strongly and positively associated with giving up work, while the mother’s age was not. The association between the cohort child having a long-term health condition and leaving work found for all mothers was not evident among partnered mothers. Also in contrast to the findings for all mothers, there were some indications that partnered mothers living in the most deprived areas were more likely than those in the least deprived areas to give up work (area deprivation as a whole was borderline significant in the model, p=.070).

The analysis for all mothers found a tentative association between mothers’ socio-economic status and whether or not she gave up work, with mothers in routine/semi-routine occupations appearing to be more likely to give up paid work than mothers in managerial or professional occupations. Among partnered mothers, working in lower supervisory or technical occupations or in routine/semi-routine occupations was associated with giving up work and not having returned by the time their child was aged 5. For example, the odds of partnered mothers working in lower supervisory and technical occupations giving up work were more than two and a half times bigger than for partnered mothers working in professional/managerial occupations.

As noted above, when looking at all mothers who had not changed jobs in the first 10 months after having the cohort child, we found that not having access to flexible working – or at least not being aware of it – was associated with giving up work. Given the suggestion made by previous research (e.g. Dean et al., 2017) that those in routine/semi-routine occupations are less likely to have access to flexible working, another multivariable model was fitted for partnered mothers who had not changed jobs in the first ten months after the cohort child was born and who were not self-employed (results not shown). Interestingly, this showed no significant relationship between access to flexible working and a mother’s propensity to give up work. This seems to suggest that, although it may play a role, a lack of access to flexible working does not (fully) explain why mothers in routine and semi-routine occupations were more likely to give up work after having a child.

Neither their partner’s socio-economic class nor their level of income was independently associated with whether or not a mother gave up work after childbirth on an overall level ^[39] . However, there were some indications that mothers with a partner on a ‘middle’ income were more likely than mothers with a partner on high incomes to give up work after having a child ^[40] .

To supplement the analysis looking at partnered mothers, a further multivariable model was fitted for single mothers ^[41] . Among single mothers, after controlling for other factors, none of the characteristics and circumstances found to be associated with leaving paid work for all mothers were found to be significant. However, it is difficult to say whether this is due to genuine differences or whether it is simply due to the low base sizes in the analysis of single mothers ^[42] .