Poverty and inequality: introduction.

Portes, Jonathan

The themed section of this Review includes four papers that look, through different lenses, at the evolution of the UK income distribution--the components and dynamics of income over time. Each contains significant new contributions to our understanding; taken together with other contributions to this literature, they paint a much richer picture of this issue, and provide some pointers about what we should be looking for in the future.

The central stylised fact here--one which has, one way or another, dominated the recent political and policy debate about domestic microeconomic and social policy--is the remarkable changes in the overall UK income distribution over the past four decades. The very substantial increase in inequality that began in the 1970s and continued through the mid-1990s was followed by a more complex picture, with inequality increasing in some parts of the income distribution and decreasing elsewhere.

[FIGURE 1 OMITTED]

Two of the papers focus on understanding and decomposing these changes. Richard Dickens' paper, 'Child poverty in Britain: past lessons and future prospects', decomposes changes in child poverty over the past decade (primarily as defined by the government's relative income measure) into different factors: demographic change, changes in wages and work, and changes in benefit and taxes. It finds that demographic changes and changes in the wage structure had the effect of increasing poverty over this period; and while welfare reform did increase work among families with children, this did not translate into large falls in child poverty; those entering work relied on substantial increases in government benefits to lift them over the poverty line.

So the key driver of the fall in poverty over this period was changes to the tax and benefit system, which both made it more progressive and favoured families with children (especially lone parents) over those without. Dickens concludes that the policies of the coalition government, which continue those of the last government in emphasising the importance of welfare reform and getting adults in workless families into work, but also include substantial benefit cuts which are likely to disproportionately reduce the incomes of poor families, are likely to result in a substantial increase in child poverty, a conclusion also supported by recent analysis from the IFS, which predicts sustained rises not only in relative child poverty but absolute child poverty. In fact, the forecast squeeze on personal incomes means that, most unusually, it is possible that absolute poverty will in fact rise faster than relative poverty (Brewer et al. 2011).

Mark Stewart's paper focuses more narrowly on the labour market, looking at individual earnings rather than household incomes. This allows the use of larger data sets, so it is possible to disaggregate to different segments of the population. The main focus of the paper is on the differences between regions and sectors. Regional income differences in the UK are large, and there has been much debate over the role of the financial sector in the UK economy. So it is perhaps not surprising--but has not previously been documented comprehensively--that Stewart finds that the growth in earnings inequality has been driven primarily by that in London and the financial sector.

The other two papers look at the same issues through a more dynamic, or longitudinal lens. Stephen Jenkins' paper examines trends in the instability of personal incomes in Britain in terms of changes in the transitory variance and in volatility. He shows that, in contrast to the US, there is no significant upward trend between the early-1990s and the mid-2000s; there was no significant change in income instability in the UK over this period. One possible explanation is that the social security safety net in the UK is more developed, so changes in labour income are compensated for more than in the US. The lack of a trend in income instability in the UK masks significant trends in the instability of the components of income. Since the end of the 1990s (to 2003) the transitory variance of income from benefits and tax credits rose sharply.

Justin van de Ven's paper concentrates on trends--or the lack of them--in private saving. Occupational pension provision, widespread in the UK private sector in the second half of the last century, began to decline sharply in the 1990s. Rational households, given rising personal disposable incomes over the period since this decline began, might have been expected to respond to this by increasing personal saving out of disposable income (in pensions or other investment vehicles); particularly given the well-publicised continued increases in life expectancy over this period. However, van de Ven finds no evidence of such an offsetting effect; this suggests a substantial, and possibly growing, proportion of the UK working population continues to 'undersave' relative to what might be reasonable expectations for retirement incomes.

It is stating the obvious to say that the picture gets more complex the more you look at it. But what themes emerge from this and other recent contributions to this literature? I would like to draw out three.

The first is what looking at incomes over the life cycle, as opposed to the cross-section picture shown in the IFS chart above (figure 1), tells us about inequality in the UK. Briefly, my assessment is that there is little evidence to suggest that the picture changes very much in qualitative terms. It was argued by some at the time, particularly in the US, that the large increase in static inequality seen during the 1975-95 period was of little concern if it was true only in cross-sectional measures; ie, with considerable income mobility from year to year, few people were trapped in low incomes for long periods. So, for example, the President of the Federal Reserve Bank of Dallas argued, "that most lower income households do rise through the income distribution, with a healthy percentage of them making it all the way to the top" (Federal Reserve Bank of Dallas, 1999).

However, at least in the UK, while incomes, especially of those on low or no incomes, do change from year to year (measures of 'persistent poverty' published by the Department for Work and Pensions show that many people move in and out of poverty from year to year), it is clear that the increase in inequality was roughly similar if incomes were measured over longer periods; earnings mobility if anything fell. This is shown in an earlier paper by Dickens and McKnight (2008). This paper also shows that the slight fall in inequality in the 2000s was mirrored by a slight rise in earnings mobility (or volatility), consistent with the results found in Stephen Jenkins' paper in this issue. Similar results are found in Savage (2011). (The regional patterns in this paper are also consistent with Mark Stewart's results.)

It seems to me that we can therefore come to a reasonably firm conclusion about the broad 'story' of inequality in the UK over the past forty years; consistent with the static analyses, inequality measured, if not over the life cycle, over reasonably long periods, rose sharply in the period up to about 1995, driven by growth in earnings inequality over the entire earnings distribution, demographics and other structural change, as well as a less progressive tax and benefit system. After 1995, and especially after 2000, the picture became more mixed. Structural changes, especially at the very top of the earnings distribution, continued to drive greater inequality; those in the financial sector and high earners in London did particularly well. However, a much more progressive tax and benefit system, and the National Minimum Wage, helped reduce inequality at the lower end and in the middle of the distribution. Throughout this period female labour force participation increased, driven by societal trends and also recently by welfare policies. However, once in the labour force, a disproportionately large number of women are low-paid, part-time and on the edge of poverty at best, with limited opportunities for advancement.

The second question naturally arises from this--what does this tell us about the future? In the short to medium term, it is difficult to be optimistic. Three of the papers in this issue argue for pessimism. Most directly, the analysis in Dickens suggests that much of the progress in reducing child poverty came from increases in the progressivity of the tax and benefit system; but changes made by the coalition government are likely to reverse this. The government argues that welfare reform, and in particular helping workless families into work, is the key to escaping poverty; but Dickens' analysis suggests that even relatively successful welfare reforms under the previous government contributed only modestly to progress. In particular, while, because of improved work incentives and active labour market policies, the lone parent employment rate increased by about 10 percentage points--reversing the previous long-term downward trend--the impact on child poverty was limited.

It is difficult to see why this should change in future; elsewhere, I have argued (Portes, 2011) that it is implausible to suggest that the government's flagship welfare reform measure, the Work Programme, will result in employment increases that are significant at the level of the population as a whole, even if the programme is reasonably successful. Meanwhile, Stewart suggests that structural factors (regional and sectoral disparities, as well as the usual story of skill-biased technological change) are the key underlying drivers of growing earnings inequality. It is difficult to see why this should reverse; while one might hope that rising educational participation would over time mitigate the rise in the skill premium, there is little evidence of this in the data as yet (OECD, 2010). And van de Ven's analysis, taken together with the other references above, suggests we are accumulating a worrying problem of non-accumulation--that is, insufficient saving, particularly for the same sort of 'precarious' workers, especially in the private sector (low paid, mostly women, often part-time, with relatively little opportunity to move up the earnings ladder).

Finally, turning from the short to medium term to the longer term, what light does all this shed on the ongoing debate about intergenerational social mobility? The much cited paper by Blanden et al. (2001) has had a huge impact in policy and political circles. It has often been misinterpreted as saying something about contemporaneous trends in social mobility, with some arguing that the measured fall in intergenerational mobility between the 1958 cohort (outcome measured in 1991) and the 1970 cohort (outcome measured in 2000), in some way reflects on more recent developments. See, for example, Conservative Party (2007), which argued that "Social mobility is lower today than ten years ago".

Blanden et al. (2001) has also been challenged on methodological grounds by a number of analysts, for example Saunders (2010). However, while the consensus among economists is clearly that, while the study is by definition only a snapshot of two particular years, and the raw numbers should not necessarily be taken literally, by any measure there was a significant fall in intergenerational social mobility (or more precisely a significant rise in the correlation between the incomes and educational attainment of parents and children) over roughly this period.

There is, however, no definitive consensus on what explains this fall, although the obvious explanation--the rise in static measures of earnings and income inequality for the families in which these cohorts were growing up --seems eminently plausible. The 1958 cohort would have entered the labour market between, roughly, 1974 and 1980; the 1970 cohort between 1986 and 1992; as discussed above, the major share of the rise in static measured inequality took place between these two periods.

One explanation widely cited by politicians and the press can, however, be dismissed--the (near) abolition of grammar schools; Boliver and Swift (2011) show that for the 1958 cohort comprehensive schools performed overall at least as well as grammar schools in terms of social mobility, individual examples to the contrary notwithstanding. The OECD, on the basis of the cross-country evidence, concurs that selective education is overall damaging to social mobility, concluding that, "early selection into different institutional tracks is associated with larger socio-economic inequalities in learning performance without being associated with better overall performance" (OECD, 2009).

A more plausible alternative explanation is that the fall in social mobility is partly due to the increase in higher education participation being focused on those from higher income families. Blanden et al. (2006), conclude that the primary driver is "the increased influence of parental income in determining educational attainment, especially higher education, and labour market attachment".

Of course, it is far too early to come to a definitive conclusion on developments in intergenerational social mobility over the past decade or so. But early indicators suggest--consistent with the picture above on inequality --that there is unlikely to have been any further deterioration, and there may be some (rather modest) improvements, with some narrowing of gaps in educational performance (see Gregg and MacMillan, 2009). More recent data, analysed by Chris Cook in the Financial Times (2011), found a significant narrowing of the gap in educational attainment in state schools, leading Simon Burgess (CMPO and University of Bristol) to conclude, "We may have here the first evidence of a turning of the tide...declining social mobility is not an immutable force, but can be changed. Indeed, it seems that it was changed by the education policies of the previous government".

This seems somewhat optimistic. While the narrrowing of the gap at 16 does appear real, on the other hand, if anything, the youth labour market appears to have become more polarised (with an increasing proportion of young adults attending university, but also a rise in the number not in education, training or employment), which does not bode well for later outcomes.

Overall, the recent evidence is summed up well in the Cabinet Office (2011) document, 'Opening Doors, Breaking Barriers'.

What then can we say about prospects for the future? On the basis of the papers in this issue and others in the literature, it is difficult to be optimistic. To recap:

* the prospect is that income inequality is likely to rise again, driven both by structural change and government policies;

* increases in static inequality measures are not offset by greater income mobility (or intragenerational social mobility);

* the general stylised facts suggest that social mobility is inversely correlated with income inequality (see figure 2 and the references in Cabinet Office, 2011), although the relationship is clearly not deterministic.

[FIGURE 2 OMITTED]

As Crawford et al. (2011) put it: "We know that countries with higher income inequality tend to have lower social mobility (at least when using income-based definitions of mobility)., it is likely to be very hard to increase social mobility without tackling inequality".

As an antidote to this gloomy picture, however, it is worth setting out the government's counter arguments. To oversimplify somewhat, the government argues that the increase in progressivity of the tax and benefit system, while reducing measured inequality, was just papering over the cracks, and failing to deal with the structural causes of greater inequality and reduced social mobility. The priority should be early years education, to reduce educational underperformance among more disadvantaged groups, and to tackle entrenched worklessness among some groups, especially young adults with low qualifications.

In principle, there is much to commend in this approach, and it is in no way inconsistent with the majority of the evidence set out above. Indeed, it seems plausible that, if successful policies could be implemented in these areas, over the longer term they would at least contribute to reducing inequality and eventually increase social mobility. It is, however, worth noting that in a number of areas the specific policy changes announced so far do not seem to be based on strong evidence of what might contribute to greater social mobility:

* the reduction in funding available for SureStart (although Sure Start was notionally protected in cash terms in the Spending Review, in practice the removal of the ringfence means that it is subject to the same reductions as other unprotected local authority services) does not seem consistent with the priority attached to the early years--a point made strongly by the Frank Field Review (2010).

* By contrast, the extension of childcare to relatively disadvantaged two-year olds may have some positive impact. In both cases, however, it should be noted that while there is very strong evidence that early outcomes are important for later educational attainment, and hence very probably for social mobility, the evidence base for particular interventions such as Sure Start is less strong (National Evaluation of Sure Start Team, 2010).

* the impact of the introduction of free schools is obviously difficult to predict at this stage but existing evidence suggests it is likely to be negative for social mobility. The OECD argues, citing multiple references, that "research has shown that school choice, and by extension school competition, is related to greater levels of segregation in the school system, and consequently, lower levels of equity" (OECD, op cit). In Sweden, the closest direct analogy, the impact of similar changes on overall performance remains controversial, although it does not appear to have been substantial; but most evidence suggests it has resulted in some rise in segregation, consistent with the OECD view.

* the abolition of the Education Maintenance Allowance, despite strong evidence that it significantly increased staying-on rates and attainment among the target group (16-18 year olds from poorer families) is likely to have a negative impact on social mobility, given the impact both on further educational attainment (e.g. university participation) and on earnings. See the summary of the evidence on EMA (Chowdry and Emmerson, 2010).

* on a more positive note, the Wolf Report on Vocational and Technical Education makes some important and evidence-based recommendations designed to help, in particular, the most disadvantaged young people: "Among 16 to 19 year olds, the Review estimates that at least 350,000 get little to no benefit from the post (16) education system". In particular, the focus in the Report on ensuring that all teenagers achieve at least a grade C in GCSE English and Maths; and, beyond that, on ensuring that young people get qualifications which are actually valued in the labour market, could if successfully implemented contribute to reducing the growing polarisation of the youth labour market described above (Wolf Report, 2011).

To conclude, it is difficult enough to state with any degree of confidence what has happened to social mobility in the past decade; prognosticating about the future is at best courageous. The coalition government has a clear commitment to implementing policies designed to increase social mobility, and its broad policy focus on addressing educational underperformance among disadvantaged groups is sound and evidence-based. Nevertheless, it is difficult at present to be optimistic about either broader structural trends or about the majority of the specific policies implemented so far.

REFERENCES

Andrews, D. and Leigh, A. (2009), 'More inequality, less social mobility', Applied Economics Letters, 16, 1489-1492, http:// people.anu.edu.au/andrew.leigh/pdf/InequalityMobility.pdf.

Blanden, J., Goodman, A., Gregg, P. and Machin, S. (2001), 'Changes in intergenerational mobility in Britain', London School of Economics, Centre for Economic Performance, http:// cep.lse.ac.uk/pubs/download/DP0517.pdf.

Blanden, J., Gregg, P. and Macmillan, L. (2006), 'Explaining intergenerational income persistence: non-cognitive skills, ability and education', University of Bristol, CMPO Working Paper Series No. 06/146, April.

Boliver, V. and Swift, A. (2011), 'Do comprehensive schools reduce social mobility?', British Journal of Sociology, March, http:// onlinelibrary.wiley.com/doi/10.1111/j.14684446.2010.01346.x/full.

Brewer, M., Browne, J. and Joyce, R. (2011), 'Child and working-age poverty from 2010 to 2020', Institute for Fiscal Studies, October, http://www.ifs.org.uk/publications/5711. Cabinet Office (2011), 'Opening doors, breaking barriers', Cabinet Office, May, http://download.cabinetoffice.gov.uk/socialmobility /opening-doors-breaking-barriers.pdf.

Chowdry, H. and Emmerson, C. (2010), 'An efficient maintenance allowance', IFS, December, http://www.ifs.org.uk/ publications/5370.

Conservative Party (2007), 'Raising the bar, closing the gap', http:/ /m.conservatives.com/~/media/Files/Green%20Papers/ Schools_Policy_Paper.ashx?dl=true.

Crawford, C., Johnson, P., Machin, S. and Vignoles, A. (201 I), 'Social mobility: a literature review', Department for Business Innovation and Skills, March, http://www.bis.gov.uk/assets/ biscore/economics-and-statistics/docs/s/11-750-socialmobility -literature-review.

Dickens, R. and McKnight, A. (2008), 'Changes in earnings inequality and mobility in Great Britain 1978/9-2005/6', London School of Economics, Centre for Analysis of Social Exclusion, CASE/ 132, October, http://eprints.lse.ac.uk/28245/1/ CASEpaper132.pd.

Federal Reserve Bank of Dallas (1999), 'By our own bootstraps, economic opportunity and the dynamics of income distribution', http://www.dallasfed.org/fed/annual/1999p/ ar95.pdfi

Financial Times (2011), 'Poorer children close education gap', 30 September, http://www.ft.com/cms/s/0/d82fc3cc-eab3-11e0aeca-00144 feab49a.html#axzz1ZYz7DPD4.

Frank Field Review (2010), The Foundation Years: Preventing Poor Children Becoming Poor Adults, HM Government, December,http://webarchive.nationalarchives.gov.uk/ 20110120090128/http://povertyreview.independent.gov.uk/ media/20254/poverty-report.pdf.

Gregg, P. and Macmillan, L. (2009), 'Family income and education in the next generation: exploring income gradients in education for current cohorts of youth', University of Bristol, CMPO Working Paper 09/223.

National Evaluation of Sure Start Team (2010), 'The impact of Sure Start Local Programmes on five year olds and their families', Birkbeck University of London, Institute for the Study of Children, Families and Social Issues, http:// www.ness.bbk.ac.uk/impact/documents/RR067.pdf.

OECD (2009), PISA 2009 Results: Overcoming Social Background: Equity in Learning Opportunities and Outcomes, Paris, OECD.

--(2010), Education at a Glance, Paris, OECD.

Portes, J. (201 I), 'Will the work programme succeed', Cabinet Office, Working Brief, July.

Saunders, P. (2010), 'Social mobility myths', Civitas, June, http:// www.civitas.org.uk/pdf/SocialMobilityJUNE2010.pdf.

Savage, L. (2011), 'Snakes and ladders: who climbs the rungs of the earnings ladder', Resolution Foundation, http:// www.resolutionfoundation.org/media/media/downloads/ Snakes and Ladders_Final_Report.pdf.

Wolf, A. (Wolf Report) (2011), Review of Vocational Education, March, https://www.education.gov.uk/publications/ eOrderingDownIoad/The%20Wolf%20Report.pdf.