21 October 2010
From Bowlby to Balls:
Changes in the compatibility of motherhood and employment
Professor Heather Joshi
It was rather a risky operation giving a title to this talk six months ago because I did not know whether Mr Balls was still going to be in charge of the British family, but at least he alliterated nicely with John Bowlby, whom some of you may have heard of.
I wanted to talk about one feature of social change in the last 60 years, which is that motherhood and employment have become increasingly, if not perfectly, compatible, from a situation where they were supposed to be mutually exclusive activities.
In the 1950s, John Bowlby, the psychiatrist, had views about the importance of the attachment of the infant to the mother, which underwrote a general feeling that mothers should be at home with their children and therefore that motherhood and employment were not compatible. However, as you are probably all aware, 60 years on, in the 21st Century, we find that governments have been quite keen on encouraging mothers to have a paid role as well as a parental one. Even beyond the government in which Ed Balls was the Secretary of State for Children, Schools & Families, the Coalition Government also has to live in an environment where motherhood and employment are no longer mutually exclusive.
I thought that you would be interested today if I traced a few features of this transformation that has occurred in this and other countries since the Second World War. I have a group of slides describing the rise of the working mother, although I do not really like that term because mothers at home do plenty of work, even though it is not paid. I have another section about the impact of motherhood on earnings. I will then turn to the question of whether equality for women was having an impact on fertility and was a threat to the survival of the species and the reproduction of the British family. This query was widely posed 20 years ago. I will then briefly show some information explaining why the birth rate has not dwindled to zero. This has not happened because mothers are not the only people who look after children. I will then turn to another controversial question which asks whether the increased participation of mothers in paid work has damaged the development of their children. I have a few summaries of research to show you about that question, and then I will summarise with a few remarks about what the implication of all this is for policy in the 21st Century.
I am going to outline what economists sometimes call stylised facts about the increasing participation of mothers in paid work in the second half of the Century, by taking three longitudinal snapshots of successive generations of women born across the 20th Century. These are culled from the information collected in the first three British Birth Cohort Studies, which were set up initially to collect information about health, but have, along the way, produced a lot of information about socioeconomic change as well, so I will just focus on those today.
Before we look at the picture about the increased paid work by mothers of young children, I thought I would fill in, graphically, a little bit about the changing context for successive generations of British women. The horizontal axis shows time and is divided into 12 year chunks which is a long-term division of different generations. Roughly speaking, two of those units across the horizontal axis would represent a generation.
1946, 1958 and 1970 are the years in which the National Birth Cohort Studies began, following everybody born in a week in those three years. I have projected backwards to show the experience of people born in the pre-War era, looking at the evidence collected in those Birth Cohort Studies on the parents of children born in the post-War era. The top line on the whole graph suggests that until the cohort born in 1970, virtually all women were married.
The dark blue line below it shows that virtually all of them became mothers. However, that drops down to about 80% as projected for the 1970 cohort. The rest of the lines on that graph are a whistle-stop tour of the educational history of the British population – men and women.
The lines plunging down from the top left hand corner to the bottom are the proportion of men and women in each cohort who entered adulthood with no qualifications. This is complemented by the proportions, shown on the lines in the bottom, which entered adulthood with tertiary qualifications – this is increasing.
In the earlier years, women were over-represented – were more likely not to have any qualifications at all, and in the earlier years, men were more likely to get higher qualifications. However, the pink line and the red line are crossing on the far right hand side of the diagram. This is showing that women are overtaking men in the proportion entering or getting higher qualifications and there is still a female lead at the end of the 20th century. So we have a picture that, for the women who mostly became mothers over the second half of the 20th Century, there was a steady increase in the proportion of them who were well-qualified and a steady drop in the proportion of them who had no qualifications whatsoever.
Now, here is the picture with which I am choosing to present to you the history of the rise in mothers’ employment. Instead of using a normal, familiar measure from the Census, which is how many mothers at any one time are in paid work, I am measuring the amount of time a woman stayed out of paid work after her first child. I am plotting the median, which is the time it would take for half of the mothers to rejoin the labour force. The general direction of the graph is downwards. There was a pretty long gap, on average, for the women who were born in the first half of the Century, having their children in the 1940s and 1950s. The length of gap for half of them was over 12 years and some never went back at all. Whereas, by the time of the women born in the 1950s and 1970 on the right hand side of the diagram, the gap between the first birth and the first re-entry into paid work has dropped pretty close to zero on the black line, which is the line for all women. Around the black line, there is a blue line and a red line.
The blue line traces the returning to work behaviour of women who have degrees, and the red line is the returning to work behaviour of women who have no qualifications.
What I think is quite interesting is that, until the middle of that picture, the blue and red lines are pretty close together, that all mothers had very similar attachment or disattachment to the labour force and they were all staying out of paid work until their children were well into primary school age, irrespective of their educational attainments. In contrast, in the late 20th Century, highly qualified women were much more attached to paid work. I have plotted the last two ones having a zero break, which does not mean to say they do not take maternity leave, but means they are back in the labour force within maternity leave, whereas the diminishing unqualified women stay out of paid work for much longer. This generates a new kind of polarisation between mothers, which did not exist two or three decades earlier, when the patterns were more or less similar in all families.
This next picture is from the 21st Century and taken from the age three sweep of the Millennium Cohort, which is the fourth National Birth Cohort Study that I am using. This is a conventional snapshot. It shows the number of mothers who have paid work when their child is three. The figure is about half, but the other things I want to draw your attention to is that most of that work is part-time – the slashed bits of the bars – and that the distribution is, in accordance with the previous picture, socially patterned. The bar on the left represents the graduate women, 70% of whom have paid work when their child is three. The bar next to the overall average on the right represents the women with no qualifications of whom only 20% are in paid employment. I am pointing to a new source of social inequality, amongst all the others that exist in British society, which did not exist in the 1950s and 1960s.
I am now going to take a longer lifetime perspective on the role of paid work in a woman’s lifetime. A woman’s working life is likely to be interrupted a bit, depending on how well-educated she is and when she has a child, and she is also likely to work part-time when she does enter the labour force as a mother.
I did some work several years ago with the late Hugh Davies, and we simulated lifetime earnings for women who had different levels of education and different numbers of children, so that we allowed for the impact on lifetime earnings of taking time out; of working shorter hours; and of being paid less when working part-time. The results are very sensitive to how well-educated the woman is assumed to be and when she is assumed to be born. The women with no children will have a higher lifetime income than the women who have two or four, going down the bar and the different colours within those bars are reflecting the level of qualification of these simulated women.
In contrast, for the simulated cohort on the right hand side, born around 1970, the lifetime earnings are much more sensitive to the level of education. The lifetime earnings of mothers are not very sensitive to the number of children they have if they are highly qualified, whereas if they are not qualified, the lifetime earnings of mothers are very much less than the lifetime earnings of childless women.
A simulated male lifetime can be added into this thought experiment. If a person is male rather than female then he will be paid more because he is a man, but also the men in the data we were looking at did not have their incomes affected by how many children they had. Therefore, the relative lifetime earnings of men and women were catching up for the cohorts born in 1970, but they had not caught up completely.
Although there is not full gender equity, there has been an improvement in the relative pay of women to men due to equal pay legislation - the Equal Pay Act; and improvements in opportunities for women.
There was a very strong school of thought which argued that if this trend continued, women would stop having children. However, the more recent international evidence shows that the countries in which people are eschewing motherhood are not the countries in which mothers go out to work. These are the countries in which it is very difficult for mothers to go out to work in Southern Europe. The countries with the highest fertility rates have been observed in this nice place called Scandinavia, where the institutions are family-friendly and high mothers’ employment goes hand-in-hand with relatively high fertility rates.
Some demographers have even gone so far as to say that the only way to combat low fertility is to improve gender equity, and that the only way to improve gender equity is to improve the access of parents to paid work and workers to family-friendly measures.
The reason why the increased participation of women in paid work and of their pay did not wipe out reproduction was that women were not the only people looking after the children. The large provisions of childcare and the role of other family members in looking after children must also be accounted for. The traditional division of labour, which was reinforced by the Bowlby ideology, has gradually diversified and it is no longer mandatory for men not to do any housework, just as it’s no longer mandatory for women not to do any paid work.
This is just some evidence of the range of childcare arrangements that were being made by the mothers of the Millennium Cohort that I showed you earlier on. The area at the top of these bars shows formal childcare - nurseries or group settings in general. The mauve bars represent childminder arrangements, which are not that common, but are one of the options. The bottom bars are labelled “partner or self”, so it is mostly the fathers looking after the child while the mother is mostly doing a part-time job, and the yellow areas are the grandmothers, or the grandparents, playing, at age three, a very important role in family arrangements. As these children got a bit older, the newly available universal nursery schools meant that less was needed from family members, or paid helpers. There is some difference in the graph between families in which the mother’s job is full-time and where it is part-time, and also some difference where the mother has a degree and where she is less qualified. The graduate mothers and the full-timers are more likely to make use of formal childcare.
I do not have much evidence about the growing role of fathers in helping to raise the children of working mothers. This picture shows not so much what they do, but their aspirations. Responses were given most recently when the children from the Millennium Cohort were age seven. Parents were asked if they had enough time with their child. Those who did not have paid jobs, on the left, said that they have just about enough time. On the far right, there are the employed fathers, who are the group who were least likely to say that they spend enough – they themselves are saying that they do not spend enough time with their child. Not spending enough time with the child is given by the brownish lines on the far right hand side, and the green bar is the number of fathers saying that they have enough time with their child. This shows that the institutions in society are leaving us with fathers who wish they had more time with their child than they actually do. So, if I did have time budget data on how much time children spend with their fathers, which I do not, one would have to temper it with the sort of evidence that fathers do aspire to spend more time with their children.
I am coming to the next contentious issue about whether the participation of mothers outside the home has had a deleterious effect on the children who are born. These children do seem to still be getting born, but this issue asks whether they are failing to flourish because their mothers are going out to work. To some extent, there are many people who believe that children must be harmed, and there is no evidence that I could present which would conclusively prove otherwise. However, I have been involved in crunching an awful lot of data which relates scores of children’s cognitive and behavioural development. Using British and US data, I can then compare children whose mothers went out to work when they were very young with children whose mothers stayed at home.
The British data does not come from the Millennium cohort. It comes from the children of the 1970 cohort, where we have evidence on children of a range of ages, from four to 16. Their development is rated from various tests. The top one is about the maths test, the pink one down there are various tests of literacy. We are controlling for as many other factors which might contaminate the relationship as we can. The horizontal axis shows British children aged four to five on vocabulary, British children aged five to six on reading, and the next two panels are about American children from a comparable survey. Within each test, we have divided up whether the mother was working full-time or part-time after she returned from maternity leave, but before the child’s first birthday.
The blob is an estimate of the average difference on each score, between the children with the working mother and the children with the not-working mother. The vertical lines around them are the margin of error. If the margin of error crosses the horizontal line of zero then the estimate is not significantly different from zero. Only one vertical line does not cross zero, and that has a green blob. That represents reading tests for American children, but in none of the tests that we could apply to the British children was there anything statistically different between the children whose mothers had gone back to work in their first year and those who had stayed away for longer. The mothers had virtually all been on maternity leave, so they were not returning to work as young as the American mothers would have done, where there was less maternity benefit and less maternity leave.
So the summary of all this is that motherhood and paid work are no longer incompatible. There is diversity rather than uniformity in modern patterns of parenthood. Childrearing is no longer just the mother’s monopoly, and children, although they are still more the business of mothers, women have business which extends beyond their children.
©Heather Joshi, 2010