In response to arguments that the ‘social mobility problem’ has been overstated and that social mobility as a policy aim is futile, Jo Blanden reviews research that her and colleagues have conducted into intergenerational mobility in the UK. She argues that there is good evidence that relative intergenerational income mobility declined over time in the UK and that governments can indeed affect the mechanisms which can promote social mobility. To do so they must take a close look at the overall shape of society.
‘Social mobility’ has become a policy buzz-phrase and has been used as a motivation for government initiatives from both Labour and the current coalition government. There are many of these, and they include provision of free nursery places for poorer 2-year olds, the Pupil Premium and the phased increase of compulsory schooling up to age 18. Earlier in the month Alan Milburn’s Social Mobility and Child Poverty Commission published its first State of the Nation Report, the objective of which is to hold government to account for the likely impact of its actions for Social Mobility and Child Poverty. Social mobility is big policy business.
Philip Collins, Tony Blair’s former speech writer and now a journalist at the Times, has been talking lately about the futility of social mobility as a policy aim (at the Liberal Democrat Conference, in Prospect and on This Week with Andrew Neil. He argues that many politicians and commentators do not fully understand what they mean when they discuss social mobility; that the evidence that Britain has a ‘mobility problem’ has been overstated; and questions if education policies can make any difference. In fact, he says “I can’t think of a single education reform in the 20th Century that had a marked impact on relative social mobility at all. Not one”.
This blog explains how my work with colleagues has informed the consensus view that Britain has a mobility problem. I argue that social mobility is an important concept, despite being a somewhat misused term.
Our research and the policy consensus
In 2001 Paul Gregg, Stephen Machin, Alissa Goodman and I first published findings that showed that the earnings in their 30s of men born in 1970 are more closely linked with their parental income at age 16 than was the case for the same measure for an earlier cohort of men born in 1958.
This research was timely. The New Labour government was concerned with education and the plight of the poor, but less worried about inequality towards the upper end of the distribution. A social mobility objective therefore sat more easily than one based on overall equality. The fact that mobility got worse for those growing up in the 1970s and 1980s also chimed with New Labour’s 1997 theme song ‘Things can only get better’.
The research led to a quite extraordinary media reaction which has continued over many years (References to our work pop up in the most unexpected places. For example, see the third paragraph in this Lucy Mangan column on the late-1980s sitcom Roseanne.). Unfortunately the language used to describe our main findings has often been imprecise. It is certainly not true that mobility has ‘ground to a halt’ or ‘fallen to its lowest level’ (examples of the use of this language are taken from John Goldthorpe’s Journal of Social Policy Paper and include John Humphrys, the Mail online and the Guardian).
What are we measuring?
The income mobility result can be illustrated in the following concrete way. Take two neighbouring families with 16 year old sons; one with an income twice as high as the other and assume they have the average amount of mobility in the population. If the boys were born in 1958 the son of the richer family would have earned about 20% more than his poorer friend when both are 33, for boys born in 1970 the gap between peers would be about 30% at 30.
Notice that when explaining the findings I compare two neighbours in both generations. The measure of mobility we use is relative. It is in stark contrast to the measure of absolute class mobility which asks if the son has a different social class to his father. This type of mobility is more likely to occur in periods where the number of white collar jobs available is expanding. Absolute social class mobility in the UK grew in the middle decades of the 20th century, but plateaued in the second half (See Goldthorpe and Mills here). Sociologists also measure relative class mobility, which compares the chances of moving up the hierarchy for sons of working class fathers with sons of professional fathers. Relative social class mobility, even when measured on the same data as the income findings, reveals very little change over time (See work by Goldthorpe and Jackson here).
Economists v Sociologists: whose interpretation is correct?
In an interview published on this blog, John Goldthorpe casts doubt on the validity of the income mobility results. He states that “The data…from the British birth cohort studies of 1958 and 1970, are ‘missing’ for a large proportion of cohort members; they are collected in quite different ways for the two cohorts; and, most seriously, they provide only ‘one-shot’ measures of family income.” Naturally, we have made strenuous efforts to check the robustness of our findings and discover why they differ from those based on social class. We demonstrate that they do not result from missing values in income and are very unlikely to be generated by a lack of comparability or error in the measures of family income.
In fact, intergenerational income mobility can change even if class mobility does not if the intergenerational persistence of income changes over time within social class groups. For example, someone with a father in the bottom 10% of managers and professionals by income became less likely to move to the top 10% of this group themselves. Just looking at differences across seven social class groups (as Goldthorpe’s work does) misses important aspects of changes in inequality unrelated to social class that have occurred over the past thirty years in Britain.
Our findings on income mobility have been confirmed by another study using a different dataset and methodology which finds that income mobility falls between cohorts born in 1961 and 1972. By this time next year there will also be results from a major study by Patrick Sturgis and Franz Buscha which uses data based on the UK censuses from 1971 to 2011 to compare trends in mobility for both social class and income measures. The findings should provide another piece of evidence in understanding the pattern of mobility in the UK.
Can social mobility be achieved through government policy?
There is therefore good evidence that relative intergenerational income mobility declined over time in the UK, but it is a substantial leap from this finding to the adoption of social mobility as a policy goal. Indeed, it would make sense for policy to focus on the drivers of mobility rather than promoting mobility in general. Our analysis of the drivers of the fall in income mobility indicates that mobility fell as the relationship between family background and educational attainment grew. In order to promote mobility, this link needs to be reduced. My current research with Lindsey Macmillan indicates that the standards agenda in education policy from the mid-1990s onwards coincides with a catch-up for those from poorer backgrounds at basic attainment levels. However, there is less evidence that poorer children are making progress at getting the very best GCSE results or access to the top universities.
I would agree with Philip Collins and John Goldthorpe that some of the intractability of these problems is related to the large material inequalities faced by children and their families in the UK. It is no coincidence that income mobility declined in the 1980s as income and earnings inequality in the UK rose and that the most mobile nations also tend to be more redistributive. Governments can affect the mechanisms which can promote mobility, but to do so they must take a close look at the overall shape of society.
Note: This article gives the views of the author, and not the position of the British Politics and Policy blog, nor of the London School of Economics. Please read our comments policy before posting.
About the Author
Jo Blanden is Senior Lecturer in Economics at the University of Surrey and Visiting Fellow at CEP and CASE, LSE.