Child poverty and its consequences

Paul Gregg, Susan Harkness and Stephen Machin

This report looks at children's experience of life in poverty and how it affects their adult lives.

Drawing on the Family Expenditure Survey and the National Child Development Study, the report looks at trends in the number of children living in poverty and the distribution of the income of their families, as well as at changes in the relative economic position of households with and without children over time. The report goes on to consider the extent to which adults economic and social positions are associated with factors observed during childhood. The results suggest that individual and family characteristics, especially those associated with adverse economic and social child development, are linked with subsequent success or failure in the labour market.



March 1999 - Ref 389
Child poverty and its consequences

What is the effect of family poverty on children? This question is examined in a study by Paul Gregg, Susan Harkness and Stephen Machin. Using data from large-scale surveys, it looks at the changing numbers of children in different types of poor families, at how much is allocated to spending on children themselves and at how children's disadvantages affect them after they grow up. The study found:

  • A third of children (over 4.3 million) lived in households with below half average income in 1995/96, up from just 10 per cent in 1968.
  • Most poor children now live in workless households; thirty years ago over two-thirds of poor children had at least one working parent.
  • Two-thirds of children with lone parents are poor, compared with one-quarter of children with two parents. The relative disadvantage of children with one parent has risen since 1979, largely because the chance of the parent having a job has declined.
  • While child poverty (based on a fixed poverty line of half average income in 1995/96) fell sharply in the 1970s, it barely fell at all between 1979 and 1995/96, despite a substantial rise in general living standards. For the poorest fifth of the population, spending on toys, children's clothing, shoes and fresh fruit was no higher in real terms in 1995/96 than in 1968.
  • Young adults who as children suffered financial hardship, were in trouble with the law or played truant have significantly greater than average chances of earning lower wages, being unemployed, spending time in prison (men) or becoming a lone parent (women). These associations exist independently of socio-economic background or experiences in early childhood. They are only partly accounted for by lower educational attainment.
  • Parents who have themselves been disadvantaged in childhood are more likely to have children who do poorly early on at school.

Studies examining large-scale data on the economic experiences of individuals and families have typically used parents or households rather than children as the units of analysis. Children have been seen by economists in particular as having an effect on the welfare of households rather than their own welfare being the focus of attention.

This study used two main surveys, the Family Expenditure Survey and the National Child Development Survey, to explore connections between rising income inequality and labour market inequality, changing family structures and child poverty. One strand of the work looks at trends in child poverty rates and income inequalities since the late 1960s. The second looks at the relationship between childhood disadvantage (an important element of which is child poverty) and subsequent economic and social outcomes among adults. The link between the two strands is the significant long-term impact likely from high child poverty rates, which are likely to extend to undesirable outcomes among adults, including a perpetuation of disadvantage through its being passed on to subsequent generations.

Trends in child poverty
Overall trend
Despite a fall in the number of families with children and declining family size, the number of children living in households with below half average income has risen rapidly over recent decades. In 1995/96 about one-third of all children - 4.3 million - were living in such households, compared with just 1.3 million in 1968. (These and other calculations in the study are based on 'equivalised' household incomes, that is adjusted for the number of members and ages of children.)

While the number of children living in relative poverty has thus more than tripled, this change has to be set alongside a general rise in living standards. More than half of children (55 per cent) in 1968 would have been defined as being in poverty on today's standard (that is, they lived in households with below half average 1995/96 income, adjusted for inflation). But the steep fall in this proportion to 36 per cent by 1979 was not sustained: 33 per cent remained below this income threshold in 1995/96. So poverty on this absolute level did not decline over a 16-year period in which general living standards rose by nearly one-third.

Which children are poor?
The rise in child poverty is closely linked to a growth in the number of children living in families without work. Between 1968 and 1995/96, the proportion of all children with no working parent rose from 2 to 10 per cent of those in two-parent families, and from 30 to 58 per cent of children of lone parents. Workless families accounted for over half (54 per cent) of all poor children in the mid-1990s, compared with under a third (31 per cent) in the late 1960s. Over the same period, the proportion of poor children who lived in lone-parent families rose from 19 per cent to 43 per cent (see Figure 1).

So what was the relative importance of growing family worklessness, compared with the growth in lone parenthood, over this period? One-fifth of the overall increase in child poverty can be attributed specifically to the rise in the proportion of children with lone parents, from 6 per cent to 22 per cent of all children. But a bigger factor has been the rising chance of either one- or two-parent families being out of work. This has been compounded by a growing incidence of poverty within each of the four categories of household in the figure, as defined by working and parental status.

Have children suffered more than average?
The research shows that not only have child poverty rates increased, but that poverty has risen faster among families with children than without them. In 1995/96, 28 per cent of households containing children were poor compared with 15 per cent of those without children. This poverty rate had risen by 21 percentage points since 1968 for those with children, as compared with a 10 percentage point rise for childless households. Lone parents have overtaken pensioners as the group with the lowest average income, and the income gap between pensioners and couples with children is closing. Moreover, while the overall distribution of household incomes has become more dispersed both for those with and without children, for the former there is a big concentration of very poor families relying on benefits, whereas for childless households the distribution is more even.

Expenditure on children
An alternative to measuring child poverty by income is to look at family spending patterns. This has the advantage of being a more direct indicator of whether children's needs are met, although it can be harder to measure it accurately. The available data show expenditure to be slightly more equally distributed than income across children's families, but the overall widening of inequality is confirmed. Very similar amounts are spent on children, on average, whether they live in one- or two-parent families, once employment status is taken into account. The biggest difference is seen not according to the number of parents but according to whether there is income from work. Moreover, the gap in spending on children whose parents are in or out of work has increased. Since 1968, average real spending in two-parent families without work has risen only slightly, and for non-working lone parents it has fallen.

But how much of this household spending is devoted to children themselves? The inequalities in overall expenditure are repeated in spending on toys, children's clothing, shoes and fresh fruit and vegetables. Moreover, there has been no real rise in spending on these items for the poorest fifth of the population over the past three decades. These findings suggest that increasing inequality in expenditure is having a direct impact on the living standards of children, and excluding the poorest from the rising material well-being of the majority.

Child development and adult economic and social outcomes

The second strand of the study showed that a number of economic and social disadvantages faced during childhood are strongly associated with the subsequent economic success or failure of young adults. Based on the interviews of cohorts participating in the National Child Development Study, the research examined characteristics at various points in the lives of people born in 1958. It looked at whether as children individuals suffered from 'family disadvantage' in the form of poverty, lone parenthood, having an unemployed father or having been taken into local authority care. It also looked at whether their behaviour by age 16 could be characterised as 'delinquent' in terms of having been involved with the police or having played truant from school. The effects of these childhood experiences were separated out from the impact of 'initial conditions' in people's lives, as defined for example by the educational levels of their parents and the cognitive abilities that they displayed at the age of seven.

The study found that those with disadvantaged or 'delinquent' backgrounds fare badly in terms of earnings and employment chances as young adults, even at the age of 33. Men are also more likely to have had a spell in prison and women are more likely to be lone parents, by the age of 23, if they have negative childhood experiences.

To what extent do these poor outcomes in adulthood arise from the fact that disadvantaged and 'delinquent' children get on average vastly inferior educational results? The study found that education was important as a 'transmission mechanism', but that it typically accounted for under half of the differences identified. The disadvantages identified had a significant impact above and beyond their effect on education.

Of the family-based measures of childhood disadvantage, poverty was found to be by far the most important force linking childhood development with subsequent social and economic outcomes. Being brought up in a lone-parent family, for example, does not seem to matter in the absence of family poverty.

Finally, the study demonstrated an intergenerational link in the cycle of family disadvantage. It looked at the tested cognitive ability, at an early age, of the children of the 1958 cohort. Where the parents had themselves grown up in socially disadvantaged situations, the average cognitive ability of children was lower. This suggests a potentially important cross-generational link that may well spill over to affect the subsequent economic fortunes of the children of disadvantaged individuals.


This study shows how the economic position of families strongly affects the present and future welfare of children. Whether a family is able to meet the material needs of its children depends more on whether it has income from work than directly on whether it has two parents. However, the much greater amount of time that lone parents spend on average out of work means that a higher proportion of them are unable to meet their children's needs as they grow up. Such economic disadvantages can lead to both economic and social difficulties in adulthood, and feed through to the next generation. So today's high level of child poverty is likely to have continuing negative effects as the present generation grows up. Conversely, any measures that successfully address child poverty, especially by giving more households access to jobs, are likely to have wide-ranging effects in the years ahead, that go beyond the improvement of the immediate welfare of poor children.

About the study

The study is based on an econometric study of large-scale micro-economic data sources.

The full report, Child development and family income by Paul Gregg, Susan Harkness and Stephen Machin, is published for the Foundation by YPS (ISBN 1 902633 06 7, price £10.95 plus £2 postage).

More detailed technical papers are available on request from the Centre for Economic Performance at the LSE. These are: P.Gregg, S. Harkness and S. Machin (1998) 'Poor kids: Trends in child poverty 1968-96', mimeo; and P.Gregg and S. Machin (1998) 'Childhood disadvantage and success or failure in the youth labour market', Discussion Paper 397.