Are income inequality and poverty in Britain rising or falling? The answers today are less clear-cut than in the past. Until the mid-1970s, incomes were very gradually becoming more equal, a trend some economic historians trace back several centuries. In the Thatcher decade, the trend was reversed, making incomes less equal than at any time since the second world war. In the 1990s, the trend flattened. Under New Labour, there has been a small increase in inequality but the first sustained decline in poverty for a generation.
Long-term trends in some of the forces that determine income distribution are now far from clear. We need to understand these forces better - not least because, for the first time in a generation, there is a cross-party consensus that relative poverty matters.
What are the recent trends?
The basic maths of poverty and inequality is not difficult. Poverty rates most commonly measure the number of people living below a particular income level, relative to society's norms (see "Measuring poverty," below). Inequality is a measure of the distribution of income, which calculates how far the rich are above average as well as how far the poor are below it.
The story of how inequality has risen but poverty fallen since Labour came to power is eloquently set out by the Institute for Fiscal Studies in calculations shown in Graph 1. The bars indicate how fast incomes rose in real terms for people at different points in the distribution. For example, the income of a person at the 19th percentile (someone whose income is greater than 19 per cent of the population) has grown by an average of 3 per cent a year since 1996-97. The biggest rise was at the very top: those on the top 1 per cent of incomes did best, and have consistently done well in the past two decades, as City and top professional earnings have risen sharply. This development has fed inequality. But those placed about 20 per cent from the bottom of the distribution have also gained. These are typically people who have found employment during a period of jobs growth, and who have benefited from the new tax credits. Some have risen above the poverty line as a result, so poverty is falling. Their gains have also helped to reduce inequality, but this effect is offset by the fact that the very poorest 10 per cent have fared relatively badly. These are typically childless households without work, whose benefits have not kept up with rising living standards, or pensioners who are not claiming means-tested top-ups to which they are entitled.
The annual publication of the poverty figures this March was widely reported. Although they showed further falls in child poverty, some commentators were sceptical about whether the government would meet its target of halving the rate by 2010. These figures have been published annually since 1990, but until recently were little remarked upon. For 16 years, Tory administrations managed to sustain the myth that a rising tide of economic growth had lifted all ships. The Joseph Rowntree Foundation's 1995 inquiry into income and wealth shattered that myth, showing that not only had relative poverty more than doubled, but the poorest groups had seen no real income gains from 1980s growth.
Tony Blair's 1999 pledge to end child poverty in a generation crystallised a new politics in which relative poverty does matter - but inequality higher up the scale matters less. Even the Tories have changed their tune. Their social security spokesman, David Willetts, is a "one nation" Tory, and praises the goal of reducing child poverty while criticising Labour's method of achieving it.
By 2002-03, the year covered by the latest data, child poverty had fallen by 700,000 (about a sixth) and pensioner poverty by 500,000 (about a quarter) since 1996-97. This followed the sharp rise in the 1980s and a levelling off in the 1990s. Labour even looks like doing what nobody had expected: hitting its initial target of reducing child poverty by a quarter between 1999 and the present financial year. This achievement will be all the more remarkable given that median incomes are growing quite quickly, so poor people's incomes will have had to rise even faster for them to escape relative poverty. Under Tony Blair, median income has risen in real terms by an average of 2.6 per cent a year, compared to just 0.7 per cent under John Major. Under Margaret Thatcher, income grew by 2.1 per cent annually in the middle, but only a fifth as fast at the bottom. It was assumed that, with market forces making things ever less equal, any redistributing chancellor would have to run to stand still if the poor were not to fall behind.
But while celebrating the fact that, under Labour, living standards have risen healthily for families in the middle, and faster for people near the poverty line, we should also consider more closely the factors that underlie these changes. They contain some hope but also some important warnings for progress on the road ahead. Interestingly, they also show that poverty and inequality may be more closely intertwined than first appears.
Market inequalities at a plateau
Headline figures for income distribution are based on the net income of households, after taxes and benefits. This is determined first by how much households receive from earnings and other private sources of income, and second by the extent to which government redistributes that "original" income by taxing it and supplementing it with benefit and tax credits. The two lines in Graph 3 (page 53) show a long-term upward trend in the inequality of both original and final incomes, as measured by the Gini coefficient, an index of inequality. This graph shows that growing inequality in the 1980s and 1990s was driven by growth in market inequalities rather than by a reduction in the redistributive effect of taxes and benefits. The degree to which such state transfers reduced inequality was identical in 2000 and in 1980: in both cases, net income distribution was 16 Gini percentage points more equal than original income. (This is surprising, given the tax cuts of the 1980s, and is mainly due to the failure to uprate tax thresholds at the same rate as rising earnings.)
Why has inequality in income from private sources risen? Most of this income comes from wages, so pay inequality plays an important part in influencing the overall income distribution, and so does the distribution of work itself. How many hours of pay come into a household depends on several factors that have been changing recently, including the incidence of two-earner households, of part-time working and of no-earner households.
Between the mid-1970s and the mid-1990s, both pay and work became steadily less equally distributed. Among working-age families, those without income from work trebled to about 20 per cent. This growth of the "work-poor" was paralleled by a rise in "work-rich," two-earner households, which rose from about half to nearly two thirds of all couples.
Since the mid-1990s, however, these long-term drivers of inequality have, at the very least, paused. Wage inequality appears to have levelled off, while the number of non-working households has fallen modestly. Optimistic economists think the widening of the wage gap may have run its course. But to consider what may happen in the future, we need to understand what lay behind these trends.
The decline in manufacturing has influenced both the distribution of wages and the distribution of work. Today, six people work in services for every one in manufacturing. Service jobs tend to be more polarised - for example, between checkout assistants and investment bankers - than manufacturing jobs, where traditionally there have been more skilled and semi-skilled workers on middling incomes. Wider wage differences in services arise partly because trade unions find it harder to organise than in manufacturing, and because of the large number of people doing menial tasks, with low productivity.
This is not an immutable characteristic of service employment. Other European countries place greater emphasis on using well trained workers in jobs such as waiters or shop assistants. In Britain recently, it appears that improved educational levels might have reduced skill-based wage differentials by increasing the supply of labour for high-quality jobs. But they do not seem to have improved the quality of drudge jobs, so some of those leaving school with good GCSEs still end up packing shelves.
There is a lesson here for the future of wage inequality in Britain. Economists have linked our wider than average inequalities with the character of our labour supply: countries with less stratified education systems tend to have greater equality in the labour market. But the future will also depend on labour demand - the jobs that employers offer and the people they seek to fill them. It will still take a big cultural shift for employers to adopt the more evenly graduated wage rates that exist in countries such as Sweden and Japan. In some parts of the labour market - those making high use of migrant labour, for example - there remains a strong risk that inequalities could grow.
The manufacturing shake-out threw many out of work, and its effects are still felt strongly in certain communities. Here, the principal problem is no longer unemployment but "economic inactivity" - the large number of people who have dropped out of the workforce altogether, often claiming disability benefits. Today, 1.5m claim incapacity benefit - nearly twice as many as claim unemployment benefit. The jobs boom of the 1990s reduced worklessness rates overall, but this change was concentrated in the more prosperous south. For worklessness to fall decisively, prosperity needs to spread geographically far more than it has done in the past decade.
The number of workless households did not triple because of job losses alone. Social and demographic changes also played their part. Fewer than one in five families with children today follows the traditional model of a male breadwinner and a female homemaker. Many more jobs than in the past go to second earners, while there are more than twice as many lone parents as in the 1970s, half without jobs. In the traditional household, income was redistributed between earners and non-earners primarily within the family itself. Clearly that is not possible where there is no earner in the household. Hence the government's efforts to help lone parents into work, which have succeeded in bringing their employment rate back above 50 per cent, after falling to around 40 per cent in the early 1990s. If the government reaches its target of raising the rate to 70 per cent by 2010, this will help reduce child poverty further. (Much will depend on whether the quality of childcare can be improved - and whether society feels comfortable with encouraging or pushing mothers of small children into work. See Katharine Quarmby's The Politics of Childcare, Prospect, November 2003.)
Another factor in the distribution of work is the length of working lives. People live longer, work less and live in less extended families than they used to. A growing population of retired people relies heavily on income transfers, via the state. Even though occupational pensions have become more important, 70 per cent of pensioners still receive most of their income from the state. Most retired people end up on below-average incomes, so population ageing tends to increase inequality.
Older people's employment rates are, however, rising - reversing a century-old trend. Seven million people over 50 are working, compared to 5m a decade ago. This is partly because the population of fiftysomethings has grown, but also because a higher proportion are in work. This may not in itself reduce overall inequality, since part of the cause is a reduction in opportunities for early retirement, often among the better off. But a general improvement in older people's job prospects could help the 1m-plus over-50s on incapacity benefit - many of them in depressed areas.
Fiscal swings and roundabouts
Can chancellors of the exchequer affect income distribution or are their actions always outweighed by market influences? The overall effect of fiscal policy (taxes and benefits) has varied in recent years. The total amount of fiscal redistribution - the gap between gross and net inequality - has varied between about 16 and 20 percentage points on the Gini index. It was highest in 1984 and 1993, not years when chancellors were the most socialist, but at the peaks of unemployment, when the largest numbers of people had no earnings and relied on state benefits. In the ensuing recoveries, inequality has risen faster for final than for original incomes, because more poor people have been able to provide for themselves through work, so claims on the benefit system fall.
This does not mean that Gordon Brown redistributes income only as much as his Tory predecessors. In practice, chancellors can do much at the margin to affect the incomes of particular groups, even though the influence of fiscal policy on income distribution may be modest relative to the market.
This can be seen by contrasting what happened under the two most distinctive recent chancellors, Nigel Lawson (1983-89) and Gordon Brown (1997 onwards).
The Lawson period probably marked the most dramatic change in the distribution of disposable income in British history. The Gini coefficient rose by 8 points in six years (1984-90). To put this in perspective, in the previous quarter century it had remained within a 3-point band, and it went on to fluctuate by little over 2 points in the 1990s.
To what extent was this effect produced by Nigel Lawson, the tax-cutting chancellor? Lawson's tax cuts undoubtedly redirected income to the rich. However, the most dramatic cut, of the top rate from 60 per cent to 40 per cent, came near the end of his chancellorship, in 1988. Much of the growth of inequality in the 1980s arose from what was happening to market incomes. One of the biggest fiscal factors was what Lawson failed to do. At a time when average incomes were growing at an unprecedented rate (they grew by a third in real terms over this short period), he continued to peg state benefits to rises in prices rather than earnings, so they remained frozen in real terms. Relative poverty shot up.
While Nigel Lawson's focus was on improving life for people on middle and upper incomes, Gordon Brown's eye is focused on the poor. He took no action to correct the strong relative rise in top incomes during the late 1990s, so overall inequality remained around record levels. (This was the first sustained rise in higher incomes under the 40 per cent top tax rate, so to some extent Brown reaped what Lawson had sown.)
However, inequality would have risen much more had Brown and his immediate Tory predecessors preserved the tax structure bequeathed by Lawson. The better off are now footing a much greater proportion of the income tax bill than even before Lawson's changes. The top 10 per cent contribute 52 per cent of all income tax compared to 39 per cent in 1986-87. This is because the chancellor has not raised the threshold for paying the higher rate in line with the earnings of the better off. Since Lawson left office, the number of higher-rate taxpayers has doubled to over 3m, while those already paying the higher rate do so on greater slices of their income. Thus, just as Lawson hit the poor by not uprating benefits in line with earnings, so Brown has offset the gains of the rich by not uprating tax thresholds in line with earnings. There is a neat symmetry in these passive redistribution strategies.
But it has been at the bottom end of the distribution that Brownite redistribution has been most clear cut. The last seven budgets have been pro-poor, focusing particularly on children and pensioners. On average, the poorest tenth of families with children have gained over ?50 a week from these budgets, compared to ?5 a week for the top tenth. This redistribution has helped contain rises in inequality fed by increases in executive and professional pay. The Institute for Fiscal Studies recently estimated that the small widening of inequality under Labour would have been twice as great without this direct transfer of resources to the poor. Government transfers accounted for about half the reduction in child poverty, the other half coming from parents getting jobs.
Can government largesse continue to contribute to poverty reduction over the longer term? Even under Labour, the answer is uncertain. While Gordon Brown has given generous help to some groups, he continues to link rises in most out-of-work benefits and pensions to prices, not living standards. Means-tested tax credits for families and pensioners are currently linked to earnings rises, but it is uncertain how long this will continue.
Why is the state so reluctant to raise out-of-work incomes in line with the growth of earnings - a failure that has compounded the market causes of rising poverty and inequality? In the Thatcher years the argument hinged on incentives and rewarding merit. Today, it is a question of affordability - especially given the rising number of pensioners over coming decades. To maintain the relative living standards of a non-working group that is expanding, the working generation needs to sacrifice a greater portion of its income - whether in taxes or pension contributions.
Yet redistribution from young to old looks more like long-term self-interest for each of us than does simple redistribution to the poor. Nor need it involve a cut in the real living standards of those who are earning. In the 20 years before the next demographic crisis starts to hit, average incomes are likely to rise by 40-50 per cent in real terms. Only a small portion of these gains would need to be set aside to pay for adequate pensions. Indeed, foregoing just a bit of future affluence could eliminate poverty entirely without real pain to the better-off. If 5p-7p in each ?1 of this growth were diverted to the poor, nobody need be living on an income below 60 per cent of the median by the mid-2020s.
Expanding the anti-poverty armoury
Progressive politicians have adopted a much more selective approach to combating the uneven distribution of income than in former times. No longer is the aim to transform the underlying forces in society that produce paupers and millionaires, but rather to ensure that these forces do not produce excessive casualties: nobody should fall too far behind the norm. This priority is not just a more cautious form of social justice. Its espousal by non-socialist parties such as Labour under Tony Blair and the Conservatives under Michael Howard owes much to shared concern about the damage which poverty and its associated problems of drugs, crime and dysfunctional families can do to the rest of society.
The problem is that it is hard to lift people out of relative poverty without addressing the fundamental causes of inequality. Two weapons have helped Labour to win early battles in its war on poverty: getting people into jobs and selectively improving benefits and tax credits for the poor. But these are unlikely to produce a final victory.
The employment effect will produce diminishing returns. While work is the best route out of poverty for many people, there are limits to how far rising employment will continue to make inroads into low incomes. Jobs growth is bound to wax and wane, and it will become ever tougher to reduce the number of workless households, both in areas with long-term economic difficulties and among groups least likely to re-enter work, such as non-working men over 50.
Direct redistribution can certainly play an important part, helping to improve the relative incomes of poor people both in and out of work. However, there are great difficulties associated with such redistribution where it is not accompanied by improvements in market incomes. Gordon Brown has addressed working poverty and the incentive to work successfully by giving generous tax credits to working families with low incomes. However, by withdrawing such support rapidly as earnings rise he has created a new low-income trap. More gradual withdrawal rates would draw more people into the trap, as eligibility for benefits moved up the income scale.
A big danger over the long term is that dependence on tax credits helps to entrench a large low-paid sector, where earnings do not rise in line with the average. Ever greater public support would be needed to help these low-paid workers escape poverty.
Some pessimists point to the experience of the Speenhamland system of poor relief in the late 18th and early 19th centuries, under which public subsidy maintained labourers' incomes at a minimum level to combat falling wages. The result was that employers felt able to lower wages further, producing greater dependence on public support. Eventually that support became too much of a burden and was reduced, leaving labourers worse off. There are several reasons to think that today's tax credits will not have such a clear-cut effect on employer behaviour. Yet a long-term answer to poverty cannot ignore what is happening to wages as well as to public transfers, especially since the limits to public tolerance for fiscal redistribution are all too evident.
A third, essential weapon in any long-term fight against poverty must be an improvement in low-paid workers' wages. The minimum wage was introduced at a very low level compared to other European countries, but has been making up ground. The rise to ?4.85 an hour this October will represent a 35 per cent increase in the first five years; considerably faster than pay in general. But more important than an adequate minimum wage will be the way in which jobs are structured. In Britain and the US a group of low-paid workers are stranded in low productivity tasks, and there is little pressure to raise either their wages or productivity. In other European countries with higher average pay, the pressure is on to use labour more efficiently. Minimum wages contribute to this, as does the higher proportion of young people who gain a basic qualification at the end of secondary school. However, employment cultures are not changed easily or quickly.
The recent pause in the rise in wage inequality shows that globalisation and the loss of manufacturing jobs may not make pay inexorably less equal. If the wage structure were to become less polarised, this is likely to reduce inequality and poverty at the same time, by helping a wider section of people on low incomes. If so, the recent experience of poverty falling while inequality rises is likely to be an exception. However, if wage inequalities resume their rise, inequality is likely to grow, and the fall in poverty could be halted.
Perhaps the safest prediction is this: while the deterioration of the 1980s was alarmingly rapid, any improvement will be far more gradual. That is why poverty reduction is only effective where it is a long-term mission shared by all. Labour's biggest contribution to this cause may be to have made the distribution of incomes an issue that a future Tory government can no longer ignore.
Measuring poverty
There is no single way of measuring poverty. Surveys show that most people in Britain accept the idea of a minimum acceptable living standard relative to prevailing conditions, rather than an absolute threshold required to avoid hunger or disease. The simplest way of measuring relative poverty is to count how many people fall below an income poverty line calculated as a percentage of the norm ? for example, how many have less than half of average income. This means that the poverty line rises with overall income growth, and poverty increases if low-income groups do not get a proportionate share of rising living standards.
It is wrong to imagine that a relative measure of poverty means that, by definition, some people will always be poor. Although some people will always be below average, it is quite possible to have nobody below half average income.
The main relative poverty threshold used in Europe today is 60 per cent of median income ? the income of the person in the middle of the distribution who is better off than half the population and worse off than the other half. This threshold measures poverty relative to a typical person, rather than relative to an average that can be skewed by changes in the incomes of the super-rich. The British government uses 60 per cent of the median, but has recently announced that it will also be looking at whether the number of people below an absolute income threshold (uprated only with prices) is rising, and at deprivation levels.
The latter refers to the number of people unable to afford particular items which a majority of the public deem essential. This is a useful extra measure, but because the definition of "essential" will only be revised occasionally, it will effectively track poverty against an absolute threshold. The government says that its aim is for the figures to fall on all three measures, but critics suspect that it will take credit for small reductions in absolute poverty to soften the political blow if relative poverty stops declining.
Measuring inequality
Income inequality can be calculated in many different ways. To measure the gap between the most privileged and least privileged groups, one could compare average incomes among the top 10 per cent and the bottom 10 per cent of the population. For a wider measure one might compare the top with the bottom half. Or to compare typical living standards near the top and the bottom, while ignoring the extremes of rich and poor, one might look at the income of an individual a tenth of the way from the top of the distribution alongside that of someone a tenth of the way from the bottom.
Economists have come up with a single measure that sums up the distribution of income all the way along the range. The "Gini coefficient" is notoriously complex to calculate, but essentially looks at the share of all income each person receives and summarises how unevenly these shares are distributed. If everyone got the same share (all incomes were identical), the coefficient would be zero. If one person got everything and everyone else got nothing, the coefficient would be one. In the real world, the Gini coefficient measures where inequality lies between these extremes.
This coefficient can help compare inequalities across time and among countries, although historical and international comparisons are imprecise since income is not always measured in consistent ways. On international comparisons, the Gini coefficient is, as one might expect, much lower in Sweden (around 0.25) than in Britain (around 0.35) or the US (around 0.45). More surprisingly, France?s rate used to be much higher than Britain?s. It is now at a similar level, because inequality grew much faster in Britain than in most other industrialised countries during the past 20 years.
