Features: October 31st, 2006

The Persistence of Poverty Across Generations

By Jo Blanden and Steve Gibbons

This study examines the magnitude of the link between child poverty and poverty later in adult life using members of two cohorts from two national datasets, one group in their teens in the 1970s and the other in their teens in the 1980s.

The study looks at the following questions:

  • How great an impact does living in a poor family as a teenager have on the chances of living in poverty in the early thirties?
  • How much has this impact changed between the two cohorts that were teenagers in the 1970s and the 1980s?
  • How far do other characteristics at age 16 and in adulthood explain these links?

How far do the effects of early disadvantage continue to be felt as individuals reach middle age (42 for those who were teenagers in the 1970s)?

The size of the link between poverty across generations

The study finds evidence of a significant persistence of poverty from teenhood to the early thirties. This persistence is measured by comparing the chances (or ‘odds’) of being poor if one’s parents are poor with the chances of being poor if they are not (the ‘odds ratio’) Of those who were teenagers in the 1970s:

  • For those whose families were poor when they were 16, 19 per cent of those with poor parents are poor and 81 per cent are not. Individuals are four times more likely to be non-poor than poor in their early thirties.
  • For those with parents who are not poor, 90 per cent are not poor in later life while 10 per cent are poor. In this case, individuals are nine times more likely to be non-poor than poor if their parents were non-poor.

Calculations based on the odds ratio find that, for those who were teenagers in the 1970s, the chances of being poor as an adult double if they were poor as a teenager. Similar calculations for the earlier cohort show that those who were teenagers in the 1980s are nearly four times as likely to be poor in adulthood. Therefore, comparing the persistence of poverty across the cohorts indicates that the strength of this persistence has approximately doubled.

For teenagers growing up in the 1970s, teenage poverty doubled the odds of being poor adults. Being poor as a teenager in the 1970s also doubled the odds of being poor in early middle age (age 42) by 2000. For this group, teenage poverty is therefore as strongly related to middle-age poverty as it was to poverty in earlier adulthood.

This is perhaps surprising: we might expect the influence of teenage poverty to fade as the years go by. One explanation could be that teenage poverty influences poverty in early adulthood, and this then links through to poverty in later life. However, accounting for poverty at age 33 has very little impact on the odds ratios for poverty at age 16. The link between poverty in teenhood and adulthood continues through to middle age, regardless of whether or not a person is recorded as poor in their thirties. It is also clear that the association between poverty at different points in adulthood is much stronger than that between childhood poverty and adult poverty.

Understanding why poverty persists

It is extremely difficult to pin down the factors that cause the persistence of poverty. Income poverty goes hand in hand with numerous other forms of deprivation, some of which are consequences of the lack of resources in the household and others of which lead to poverty in themselves. Many of these aspects of deprivation may be a result of other underlying factors that are very hard to measure and which persist through individuals’ lives. For all these reasons, it is extremely difficult to really understand the causal processes that lie at the route of the persistence of poverty through the lifecycle.

In order to gain some understanding of how poverty is transmitted across generations the researchers examined the link between teenage poverty and adult poverty when the other characteristics of the child’s family are held constant. This enables us to find out whether it is disadvantage in general rather than income poverty that is harming children’s life-chances. It also enables the analysis of which aspects of disadvantage are particularly harmful.

The results of this exercise make it clear that:

  • Poor teenagers in the 1970s grew up to be poor because of more general family background disadvantages, in particular, parental non-employment and low education. Poverty itself had little or no direct effect over and above these teenage family factors.
  • For teenagers in the 1980s, poverty had a direct effect on the chances of ending up in poverty, even allowing for differences in these same aspects of family background. Certainly, family background differences account for much of the persistence from child poverty to adulthood, but the odds of a poor teenager being a poor adult were much larger than for a non-poor teenager.

This provides some grounds for suggesting that redistribution could have had a beneficial impact for those growing up in the later cohort.

A similar analysis tells us which adult characteristics help to explain the persistence of poverty between teenhood and adulthood. Unsurprisingly, being out of work, having a partner out of work or having little accumulated work history are the factors most closely associated with poverty – both for adults in middle age and in their thirties – though low education plays an important role too. Our understanding of the persistence of poverty can be improved by analysing which of these characteristics are most closely linked with disadvantage and poverty in the teenage years.

The study finds that earlier disadvantage is associated with all of these later outcomes. One of the reasons for the stronger persistence among those who were teenagers in the 1980s is that teenage poverty became more closely linked to the likelihood of a person being out of work in their early thirties. The main factors linked to being out of work in adulthood are low education, lone parenthood and ill health. However, educational attainment does not explain the rise in persistence: the risk of poor teenagers in the 1980s ending up without qualifications was not much greater than for poor teenagers in the 1970s. a girl in the 1970s, a poor teenage girl in the 1980s was at higher risk for lone parenthood, and at higher risk for incapacity through illness in her thirties. These facts can explain part, though not all, of the rise in the intergenerational persistence over this period – but only for women.


This study presents two main new findings on the extent of the persistence of poverty.

  • First, the persistence of poverty from the teens into the early thirties has risen over time, with teenage poverty having a greater impact on later outcomes for teenagers in the 1980s compared with teenagers in the 1970s. This finding adds to the wider evidence that family background has had a growing impact on later outcomes between these cohorts.
  • Second, the link between poverty in teenhood and adulthood continues to have a bearing through to middle age for those who were teenagers in the 1970s (born in 1958). This is the case regardless of whether or not the person was poor in their thirties. In other words, an adult who was a poor teenager continues to be at higher risk of poverty by middle age even if they were out of poverty in their thirties.

The findings on why poverty persists are less clear-cut, and reveal multi-dimensional causes. The results suggest that initiatives to improve skills and employment opportunities are probably the only sensible way to tackle the problem of persistent poverty and that there is no quick fix available through other more specific interventions. Despite the lack of specific policy prescriptions that can be drawn, it is clear that children in poverty are more likely to grow up to be poor, a result that highlights the importance of the policy agenda to reduce child poverty and disadvantage but not through income transfers alone.

About the project

The data used are from the National Child Development Study (all children born in a week in 1958) and the British Cohort Study (all children born in a week in 1970).

The core data used are on income and other characteristics at age 16 for both cohorts, as well as information on later income and characteristics at age 33 for the first cohort and age 30 for the second cohort. The study also uses this information on income at age 42 for the older group.

Jo Blanden is a lecturer in economics at the University of Surrey, UK and Steve Gibbons is a lecturer in economic geography in the Department of Geography and the Environment at the London School of Economics, UK. Both authors are research associates in the LSE’s Centre for Economic Performance.

For further information

The full report, The persistence of poverty across generations: A view from two British cohorts by Jo Blanden and Steve Gibbons, is published for the Foundation by The Policy Press.