The evidence on British social mobility makes for tragic reading. The chance of escaping a poor upbringing is considerably lower than in other European countries and has actually worsened over time. This has become a modern 'social science fact' first discovered by researchers at the Centre for Economic Performance at the London School of Economics (Blanden et al, 2002) that has rapidly been absorbed into the political DNA of both main parties.

The aim of the Government to "create a Britain that is economically successful because it is socially mobile" is, when placed in this context, a significant challenge. Social mobility measures the degree to which people's economic and social status changes between generations.

It is seen by many as a measure of the equality of life opportunities, reflecting the extent to which parents influence the success of their children in later life or, on the flipside, the extent to which individuals can make it by virtue of their own talents, motivation and luck. In this essay we offer an anatomy of British social mobility and suggest what can be done to build meritocracy in this country.

What has happened to mobility?

Social mobility in Britain is well below Germany and the Nordic countries, but a bit better than America. Figure 1 shows the size of the link between sons' and fathers' earnings.

It is also the case that, when measured on a consistent basis over time, social mobility in Britain is lower now than it was in the past. After remaining constant for people born before the end of the 1950s, research based on two large birth cohorts, the first born in 1958 and the second in 1970, shows that mobility fell significantly. For example, in terms of sons' earnings and parental income, for the 1958 cohort 31 per cent of sons from the poorest quarter of families stayed in the poorest quarter. This percentage rose to 38 per cent for the 1970 cohort. Similarly at the top end of the distribution, whilst 35 per cent of sons born in the late 1950s stayed in the top quarter of the distribution, this rose to 42 per cent for the 1970s cohort. It is evident that social mobility slowed down when comparing these birth cohorts across time.

Why is social mobility so bad in the UK?

So why is UK mobility lower than in most countries, and how come it became worse over time? At least two factors are important. First, income inequality has risen very rapidly in Britain since the late 1970s. Consequently, the number of pounds sterling separating people on each rung of the social ladder has become wider, making the ladder that much harder to climb. This may also explain the bad American picture on social mobility, as the gap between rich and poor is much larger in the US than in Europe. Second, and contrary to many people's prior expectations, the expansion of higher education that occurred from the late 1980s onwards disproportionately benefited people from richer families. It therefore acted to reinforce and strengthen already existent inequalities and restricted social mobility.

The gap between the richest and poorest in society has opened up significantly since the late 1970s. Figure 2 shows how one measure of inequality, the 'Gini coefficient', rose, especially in the 1980s. The relevance for social mobility is clear – because the income distribution is now wider, an individual has to gain more income to move the same 'distance' in the income distribution than before.

Hence one reason for declining social mobility is simply increased income inequality. The Right have traditionally argued that the increase in income inequality that began in 1979 is justified because society is now more meritocratic so that it is easier for the poor to become richer if they are willing and able to work hard.

In fact, the opposite has occurred – there was actually a fall in the degree of social mobility for children who left the education system and entered the labour market after the big 1980s rises in inequality had occurred. Those children born to poor families were less likely to break free of their background and fulfil their potential than they were in the past.

The second contributing factor to declining mobility is higher education. In many quarters, education has been seen as a route to promote social mobility by helping people escape their poor family backgrounds. If increased education benefited people from poor backgrounds then it could help promote social mobility. But if it mainly benefits people from richer backgrounds then the opposite applies.

The recent British experience has seen significant expansion of higher education, but it seems to have had the latter impact and acted as a cause of reduced social mobility. The expansion of the higher education system can be seen in Figure 3, which shows participation rates since 1961. The sizable expansion of higher education that occurred from the late 1980s is crystal clear.

Unfortunately this expansion has not been shared equally across the income distribution. Figure 4 shows degree acquisition by the age of 23, broken down by family income (the richest fifth, the poorest fifth and the middle) and how this has changed over time. There is a very clear pattern showing that the relationship between degree acquisition and family income has significantly strengthened over time. This implies that the big expansion in university participation has tended to benefit children from affluent families more and thus reinforced immobility across generations. It seems that the strong relationship between family income and educational attainment is at the heart of Britain's low mobility culture.

However, it is also true that the strengthening of the education- income association was much more pronounced for the 1958 and 1970 cohorts, but has changed little across cohorts born in the 1970s. This suggests that the rise in educational inequality that went hand-in-hand with falling social mobility may have reached a plateau.

This levelling out is also confirmed in preliminary findings of Blanden and Machin (2007) who look at intergenerational associations by considering the relationship between maths and reading test scores of the children of the 1958 and 1970 cohorts and their parents.

Given this, although the fall in social mobility predated New Labour, the Government faces the challenge of how to respond to the legacy it inherited. What can be done to build meritocracy?

Our analysis suggests one clear policy response to tackle widening inequality and falling social immobility: improving the education of the excluded. The key intervention has to be to build the human capital of the 'hard to reach' who are being failed by the education system.

General increases in higher education have partially backfired because they have disproportionately benefited the wealthy. Expanding universities is a boon for national productivity, but a university degree also buys a big lifetime salary bonus. Therefore, students should be made to pay the vast majority of the costs of their education with top-up fees coupled with highly generous bursaries for the less well-off. Furthermore, whilst some of the

Government education initiatives in schools have raised achievement levels, in general they have had little impact for children at the bottom of the achievement distribution (see the evidence from the Excellence in Cities programme in Machin, McNally and Meghir, 2007).

In the medium term, improving the education of low income children will reduce inequality as it will boost the supply of skills and will dampen down the relative wage premium given to the highly educated. Even though skill demands will keep on rising due to globalisation and technological change, better education can keep a lid on increasing inequality.

More importantly, in the long-run, improving the quantity and quality of education of the poor will help them escape the social immobility trap. These interventions have to be heavily targeted at schools with many low income pupils. We suggest a five point program that runs with the grain of current reforms but radicalises them by targeting the worst off.

First, evidence shows that teachers are hugely influential in school performance. To attract the best teachers in key subjects like maths there should be large increases in pay to high performing teachers prepared to work in tough schools. This will require breaking the current pay system and giving much larger rewards to successful individual teachers in inner city schools. The rewards given to teachers living in high cost areas is much too low.

Second, we currently have selection for schools brokered through the housing market as school quality drives up local house prices 'pricing out' low income parents (Gibbons and Machin, 2006). One way to alleviate this is to follow Brighton's lead and introduce lotteries. Another is to widen catchment areas and relax the geographic criteria for selecting pupils. The fear that middle class parents will exit the system into the private sector is real, but could be mitigated by addressing the question of the charitable status of private schools.

Third, the effective supply of good schools, i.e. those providing the largest increases in value added, needs to increase. This could be facilitated by making it easier for high performing schools to expand and underperforming schools to contract. This is the key mechanism through which productivity grows in the private sector and needs to be emulated more in the public sector. Additionally, improving the 'takeover' market so that successful head teachers can take control of failing schools is also the right way to go.

Fourth, the moves towards increased autonomy of schools in terms of admissions and governance in general are a good thing. Offering head teachers more scope to be innovative allows schools to offer better options to parents for their children's education.

This of course ties in with the issue of access where parents from all income levels need to be given fair access to schools of their choice. Finally, the extensive information on school quality now available is a valuable tool in helping parents monitor school performance and choose where to send their kids. But this information is used disproportionately by the middle classes. We need to have personalised 'state school consultants' to mentor poorer parents into navigating around the maze of information to help them choose the right school for their kids. This would be helped by economic evidence on the cash pay-offs in the labor market to a higher quality education.

Conclusions

By international standards, the UK's generational game of inequality is bad and has got worse over time. The rise in overall inequality and the growth in higher education are the two main culprits. The solution is to aggressively target educational resources on the hard-to-reach groups. We suggest a policy agenda that builds upon school choice through precision information and promotes reallocation of resources from worse to better schools. We also suggest more radical interventions into the way teachers are paid and pupils effectively selected.

Even with these reforms in hand, the task of building meritocracy is formidable as there needs to be help for families outside of school hours. These deep interventions into family lives are even harder than re-engineering the school system.

For more research from the CEP on inequality and productivity see http://cep.lse.ac.uk

Notes on figures

[NOTE: FIGURES NOT YET ON THIS PAGE. THEY WILL APPEAR HERE SHORTLY.]

Figure 1: The extent of social mobility is the correlation between sons' and fathers' earnings for the following birth cohorts of sons: Norway, 1958; Denmark 1960-73; Sweden, 1962; Finland, 1958-60; W. Germany, 1960-73; Britain, 1958; US, 1954-70. Sources – Bjorklund et al (2006) Table 3, Blanden (2005) Table 3.3, additional estimates from Blanden, Gregg and Machin (2005).

Figure 2: The Gini coefficient (1970-2005) measures income inequality (0 = perfect equality, 1 = perfect inequality). Calculated using Before Housing Cost incomes using FES and FRS data.

Figure 3: Age Participation Index. The number of home-domiciled young (aged less than 21) = initial entrants to full-time and sandwich undergraduate courses of higher education expressed as a proportion of the average 18- to 19-year-old GB population. Higher Education Initial Participation Rate (HEIPR): participation by English domiciled first time entrants to HE courses at UK HE institutions and English, Welsh and Scottish FE colleges. Source – DfES.

Figure 4: 1981 corresponds to the 1958 birth cohort, NCDS; 1993 to the 1978 birth cohort, BCS; 1998 to the 1975 (average) birth cohort, BHPS; 2002 to the 1979 (average) birth cohort, BHPS. Sample sizes: NCDS 5706; BCS 4706; BHPS 725 and 363. If information on degree graduation is not available at age 23, degree attainment by age 22 is used instead.