«Briefing Snakes and Ladders: who climbs the rungs of the earnings ladder Lee Savage September 2011 © Resolution Foundation 2011 E: ...»
Snakes and Ladders:
who climbs the rungs of the
© Resolution Foundation 2011
E: firstname.lastname@example.org T: 020 3372 2960 F: 020 3372 2999
The Resolution Foundation is grateful to all those who provided comments and
insights on earlier versions of this report including Paul Gregg, John Goldthorpe
and Richard Dickens. Reviewers do not bear any responsibility for the content or any errors contained in this report, which remains the author’s own work.
I am grateful to The Centre for Longitudinal Studies, Institute of Education for the use of these data and to the UK Data Archive and Economic and Social Data Service for making them available. However, they bear no responsibility for the analysis or interpretation of these data.
Resolution Foundation Page 2 Contents Introduction
Data and methods
Factors Associated with mobility
Results of the Analysis
Earnings mobility and low to middle income households
Appendix A: Logistics regression results
Resolution Foundation Page 3 Introduction Earnings are the biggest contributor to the living standards of families in Britain, but it is no longer the case that individuals can rely on automatic earnings growth to lift their standard of living. Median wages have been stagnating since 2003 and the share of national income that goes to the wages of low-to-middle earners has fallen from £16 of every £100 in 1977 to £12 in 2010.1 This context makes it all the more important that individuals have the opportunity to earn their way to a better standard of living by progressing up the earnings ladder as they move on in their career. Is it becoming easier for individuals to move up? Who is more likely to move up the earnings ladder and who is at risk of falling behind? The Resolution Foundation’s series of work on social mobility answers these questions.
Intragenerational social mobility, an individual’s ability to work their way up to a better standard of living in their own life time, has received much less attention from researchers and policymakers than intergenerational social mobility, social mobility between generations which looks at the strength of the link between an individual’s income and that of their parents. While the Government’s recent social mobility strategy refers to mobility across the life course, most of the indicators that will track progress focus on improvements for children rather than for adults who are already in work. However, mobility within generations is vital: it gives people the chance to move on in their life through work.
The Resolution Foundation’s work on social mobility focuses on this issue of intergenerational mobility and draws on extensive data from the National Child Development Study (NCDS) and the British Cohort Study (BCS). These studies surveyed every person born in the UK in one week of 1958 and 1970 respectively.2 We compare earnings mobility for these two groups, looking at those born in 1958 that aged from their early 30s to their early 40s during the 1990s, and those born in 1970 that aged from 30 to 38 during the 2000s. Looking at this period captures the ‘peak earnings’ phase of people’s working life and reduces the influence of early career effects that might lead us to overestimate mobility.3 An individual’s thirties are also a critical decade in determining career progression and whether people continue to move on in their working lives beyond the early stages of their career.
In our first paper, Moving on Up?, published in March this year, we found a 22 percent increase in the probability of moving significantly up the earnings ladder in the 2000s compared to the 1990s, giving individuals greater opportunity to materially improve their living standards through work in the 2000s. 4 However, around 40 percent of people in the 1990s and 2000s did not move from their original quintile position in the earnings distribution. These people were more likely to be at either end of the distribution, either stuck at the bottom or protected from moving down from the top. The increase in upwards mobility was concentrated in the middle and upper middle part of the earnings distribution.
This report presents new analysis that identifies the characteristics that influenced whether individuals moved up or down the earnings ladder during the 1990s and the 2000s. It looks at a range of factors including education, gender and region and compares the independent influence of each factor on mobility in each decade.
Overall, we find that gender, education, occupation, working part-time, experience of unemployment and region independently all play highly significant roles in determining individuals’ prospects of upwards and downwards Whittaker, M. and Savage, L., (2011), Missing Out: Why ordinary workers are experiencing growth without gain, Resolution Foundation University of London, Institute of Education, Centre for Longitudinal Studies, National Child Development Study: Sweep 5, 1991 [computer file]. Colchester, Essex: UK Data Archive [distributor], August 2008.
SN: 5567; National Child Development Study:
Sweep 6, 1999-2000 [computer file]. Colchester, Essex: UK Data Archive [distributor], August 2008. SN: 5578; British Cohort Study:
29 year follow-up, 1999-2000 [computer file]. Colchester, Essex: UK Data Archive [distributor], August 2008. SN: 5558; British
Cohort Study: 38 year follow-up, 2008-2009 [computer file]. Colchester, Essex: UK Data Archive [distributor], August 2010. SN:
This is an average figure. Some groups of individuals will find that their earnings peak earlier in life while for others they will peak later in life. Goldthorpe, J. and McKnight, A., (2004), The economic basis of social class, CASEpaper 80, Centre for Analysis of Social Exclusion Savage, L., (2011), Moving on up? Social mobility in the 1990s and 2000s, Resolution Foundation
Gender Men were 40 percent more likely to more up the ladder compared to women in the 2000s – a high figure, but lower than the 51 percent likelihood of upwards mobility enjoyed by men in the 1990s. Women were more likely to move down in both decades.
Education Holding a degree is associated with significantly increased prospects of moving up the earnings ladder. Those with less than degree-level education were at least 37 percent less likely to move up in both the 1990s and 2000s.
However, the risk of downward mobility for those who did not hold a degree increased substantially in the 2000s as the proportion of workers with a degree increased.
Occupation Some occupational categories were associated with greater chances of upwards mobility. For example, professional occupations such as teachers and lawyers had a 55 percent increased chance of upwards mobility in the 2000s compared to managers. Others were associated with an increased risk of moving downwards, such as glaziers, electricians and builders.
Unemployment Those who fell into unemployment for any length of time during their thirties – even if they subsequently returned to work – were much less likely to move up the earnings ladder and were at least 79 percent more likely to move down. The penalty for a period of unemployment was greater in the 2000s than in the 1990s.
Part-time work Part-time work also incurred a penalty, with individuals who worked part-time for the entire decade being 87 percent more likely to move down the earnings ladder in the 2000s. The part-time penalty was greater in the 2000s than in the 1990s, although the penalty for people who shifted from full-time to part-time work during the decade was less severe in the 2000s than in the 1990s.
Region Regional differences in mobility became more apparent in the 2000s, with people in London being significantly more likely to move up the earnings ladder than people in other regions. Those in the South West were 83 percent more likely to move down than those in London while people in the North East had a 73 percent greater risk of downward mobility, controlling for individual skill level within each decade. This change in the significance of regional background between the two decades largely reflects changes between the two decades. Between the 1990s and 2000s, London has become a focus for job growth in the types of industries that offer the greatest opportunities for earnings progression.
Resolution Foundation Page 5 Trends in earnings mobility over the life course In March 2011 the Resolution Foundation published the first report in its series on social mobility, Moving On Up?
It compared levels of mobility for our two cohorts, one that aged from 33 to 42 in the 1990s and the other that aged from 30 to 38 in the 2000s. A number of key findings emerged. First, we showed that there was a substantial increase in upwards and downwards earnings mobility in the 2000s compared to the 1990s, although this was starting from a low base. In particular, there was a 22 percent increase in the probability of moving significantly up the earnings distribution (defined as an upward movement of three or more earnings deciles) in the 2000s compared to the 1990s. In broad terms, these findings concur with research by Dickens and McKnight. Using the Lifetime Labour Market Database, their research showed that mobility increased between 1998 and 2005 but fell between 1980 and 1998.
Second, our analysis showed that upwards and downwards mobility were not evenly distributed, being concentrated in the upper middle section of the earnings distribution. The highest earners in the 2000s continued to be largely sheltered from downward mobility and those at the bottom were far less likely to move up a substantial distance than those in the middle. Around 50 percent of people in the top and bottom quintiles remained in the same position in the earnings distribution throughout their thirties and into their early forties. The comparatively high level of immobility in the bottom quintile suggests that it is difficult for an individual to make a significant step up in earnings if they start from a very low base.
Some of the explanation for the increase in earnings mobility in the 2000s comes from changes in the wider economic context in Britain between the two decades. Inequality grew during the 1990s, stretching the distance between all the rungs on the earnings ladder7. This may have reduced mobility in the 1990s as individuals had further to travel to move from one decile to another.
Overall, the analysis in Moving On Up? showed some improvement in upwards mobility in the 2000s compared to the 1990s, albeit from a low initial base. However, there still remains the question of who experiences mobility and why. This paper looks beneath the headline rates of mobility presented in Moving On Up? to assess for the first time the individual characteristics of those who experienced upwards and downwards mobility across both decades and analyses the differences that are observed.
We assess mobility at a key point in the working lives of individuals as they move from their early thirties to their late thirties/early forties. The rationale for this is that firstly, by the time an individual is in their early thirties early career effects that might cause increased mobility, such as graduates taking low paid, entry level positions before rapidly progressing up the ladder, should be less influential. Secondly, wage rises tend to be associated with life course evolution and research shows that on average, earnings tend to rise until a person is in their early to mid 40s.9 There are variations within this general finding and some groups, such as those in higher managerial and professional occupations, continue to see their earnings rise into their late 50s, while others see theirs plateau from their early to mid 30s, for example individuals in routine and semi-routine occupations.10 However, on average we can say that the period analysed in this report covers the period during which individuals move towards their peak earning potential.
This paper focuses on relative mobility; the position of individuals in the earnings distribution relative to their peers.
For every person who moves up the earnings ladder, someone has to move down because the number of positions on the ladder are fixed. This means that, although we might expect to see an increase in absolute earnings during this period of people’s working lives as they progress to their point of peak earnings, this analysis looks at the factors that help people move up relative to their peers as their careers progress or put them at greater risk of moving down (see Box 1).
Participants in the NCDS cohort aged from their thirties to their forties during the 1990s while the BCS cohort did so during the 2000s. This allows us to compare mobility across the two most recent decades. During the remainder of this report the NCDS cohort will be referred to as the ‘1990s cohort’ while the BCS will be referred to as the ‘2000s cohort.’ For the purposes of assessing changes in earnings mobility we divide the samples into quintiles based on their hourly earnings. Mobility is defined as movement up or down the distribution by at least one quintile. The analysis in this report has been carried out using binary logistic regression. This identifies the independent influence of different variables such as education and region, holding all other variables in the model constant. This means that, for example, in the discussion of the effect of gender on upwards mobility in the 1990s, the effects of education, occupation, region, and so on are controlled for. The analysis does not permit us to assess the relative importance of one variable over the others. It simply identifies whether a specific variable has an influence over upwards or downwards mobility in the 1990s or 2000s, all other things held constant.
The birth cohort studies have some limitations. They are irregular, sometimes with large intervals between surveys.
However, they have been used repeatedly in analyses of social mobility and are generally considered to be of good quality.11 Most importantly, the birth cohort studies contain a wide range of data that is not available in alternative datasets. For example, the Annual Survey of Hours and Earnings panel data does not include information on individuals’ education levels.