Key points

  • As it is measured in relation to median income, income poverty can be viewed as being about inequality in the lower half of the income distribution only.  In looking at what has happened to the whole of the income distribution, this indicator is therefore moving beyond poverty to look at income inequality more widely.  Income inequality can be analysed in a variety of ways, and the analyse included here covers:
    • Changes in income over time across the income distribution.
    • Shares of total income across the income distribution.
    • Ratios of the incomes between the top, middle and bottom.
    • Gini coefficient: a measure of overall income inequality.
  • The overall message from these various analyses is simple: income inequalities have been increasing, both recently and over longer time periods.  These inequalities have been increasing at both ends of the spectrum.  In other words, the poorest have fallen further behind the average, and the richest have moved further ahead.
  • Over the last decade, the poorest tenth of the population have, on average, seen a fall in their ‘real’ 1 incomes after deducting housing costs.  In other words, after adjusting for inflation, their incomes are, on average, slightly lower than a decade ago.  This is in sharp contrast with the rest of the income distribution, which, on average, has seen substantial rises in their real incomes.
  • The richest tenth of the population have seen much bigger proportional rises in their incomes than any other group.
  • Except for those in the top and bottom tenths of the income distribution, households with below-average incomes have, on average, enjoyed bigger proportional increases over the last decade than households with above-average incomes.
  • In absolute (as opposed to proportional) terms, the vast majority (four-fifths) of the extra money has gone to those with above- average incomes and half of this (i.e. two-fifths of the total) has gone to those in the richest tenth.
  • The poorest tenth of the population now have, between them, 1.3% of the country’s total income and the second poorest tenth have 4%.  In contrast, the richest tenth have 31% and the second richest tenth have 15%.  The income of the richest tenth is more than the income of all those on below-average incomes (i.e. the bottom five tenths) combined.
  • The proportion of total income going to the richest tenth is noticeably higher than a decade ago: 31% in 2008/09 compared with 28% in 1998/99.  The rest of the income distribution changed little over the last decade.
  • The ratio of median income for the whole population compared to the income at the bottom decile point of the income distribution (50:10 ratio) rose for the fourth consecutive year in 2008/09.  It is now higher than at any point since the mid-1990s.
  • The ratio of income at the top decile point compared to median income (90:50 ratio) has also risen in each of the last four years, albeit at a slower rate.  It is also now higher than at any point since the mid-1990s.
  • Note that the trends in these 50:10 and 90:10 ratios over time are less extreme than the trends in the income shares previously discussed.  This is because, unlike the income shares, they are not affected by changes in the incomes of those at the very top of the income distribution, where such incomes appear to have been increasing very sharply.
  • The gini coefficient measure of overall income inequality in the United Kingdom is now higher than at any previous time in the last thirty years. 2
  • Inner London is deeply divided: it has by far the highest proportion of people on a low income (29% in the poorest fifth) but also a high proportion of people on a high income (28% in the richest fifth).
  • The composition of the population varies considerably by level of income.  For example, whilst only 10% of the poorest tenth are of pensionable age, this proportion rises to 15% for the second tenth and 30% for the third tenth before gradually reducing down to just over 10% for the richest tenth.  Alternatively, whilst almost 50% of the poorest tenth live in workless, working-age families, the proportion falls to 40% for the second tenth, 20% for the third tenth and gradually reducing down to virtually 0% for the richest tenth.  The implication of all this is that the weights of the different population groups changes depending on the focus of concern.  More specifically:
    • If the focus moves from the bottom two deciles (in effect, the ‘poverty’ focus) to the lowest decile only (i.e. those with very low incomes), then those of pensionable age become less of a concern.
    • If the focus widens from the bottom two deciles to also include the third decile (as per some of the Scottish Government’s targets), then both those of pensionable age and those in lower income, working families become more of a concern.

Why this indicator was originally chosen

One consequence of tracking the numbers on low income in relation to median rather than mean income (that is, the income of the household at the mid-point of the income distribution rather than average household income) is that incomes in the upper half of the income distribution have no effect on the numbers. While this makes sense as far as measuring income poverty is concerned, it may be insufficient for a full understanding of it as the same developments in society may be influencing both low and high income.

Definitions and data sources

The first two graphs focus on the change in real net incomes by income decile whilst the third and fourth graphs focus on the share of total incomes by income decile.

The first graph shows the average percentage change in real (i.e. after adjusting for inflation) net incomes for each income decile over the period 1997/98 to 2007/08.

The second graph shows the shares of the total change in real net incomes since 1997/98 by income decile.

The third graph shows the share of the total income of the population for selected net income deciles (tenths), namely the two poorest deciles and the two richest deciles.  Clearly, the shares added up for all ten deciles would total 100% of the total income.

The fourth graph shows, for the latest year, the distribution of total income across the ten income deciles.

The fifth graph shows two income ratios.  The 50:10 ratio is the income of a household halfway up the income distribution (i.e. at the 50th percentile, or median) divided by that of a household one-tenth of the way up from the bottom of the income distribution (i.e. at the 10th percentile, or 1st decile).  The 90:50 ratio is the income of a household one-tenth of the way down from the top of the income distribution (i.e. at the 90th percentile, or 9th decile) divided by that of a household at the median.

The data for the first five graphs is from Households Below Average Income (HBAI), based on the Family Resources Survey (FRS).  The data in the graphs over time (i.e. the first two and the fifth) relates to Great Britain whilst that in the other graphs relates to the United Kingdom (FRS did not cover Northern Ireland until 2002/03).  Income is disposable household income after deducting housing costs.  All the data is equivalised (adjusted) to account for differences in household size and composition.

The sixth graph shows how a widely used measure of overall income inequality – the Gini coefficient – has changed over time since 1979.  Gini coefficients can range from 0 to 100, where it would have the value 0 if all households had identical incomes and a value of 100 if a single household had all the country’s income.  So, the lower the number, the lower the income inequality.  See Wikipedia for the precise technical definition.

The data from the sixth graph is HBAI, based on FRS since 1994/95 and the Family Expenditure Survey (FES) for earlier years, with the data being obtained via the Institute of Fiscal Studies website.  The data relates to the United Kingdom, except for the period 1994/95 to 2001/02 when it relates to Great Britain.  Income is again disposable household income after deducting housing costs and all the data is equivalised (adjusted) to account for differences in household size and composition.

The seventh graph shows how the proportion of the population whose income is in the lowest and highest United Kingdom net income quintiles (fifths) varies by region.  Inner and outer London are presented separately as the results are so different.  To improve its statistical reliability, the data is the average for the latest three years.

The eighth graph shows how composition of the population varies by household net income decile.  For each of the ten deciles, the shares of the population are shown separately for children, working-age adults and adults of pensionable age, with the data for children and working-age adults further divided into those where no one in the family works and those where at least one of the adults in the family works.  To improve its statistical reliability, the data is the average for the latest three years.

The data for the seventh and eighth graphs is from HBAI, based on FRS.  The data relates to the United Kingdom.  Income is again disposable household income after deducting housing costs and all the data is equivalised (adjusted) to account for differences in household size and composition.

Overall adequacy of the indicator: high.  The FRS is a well-established annual government survey, designed to be representative of the population as a whole.

External links

Relevant 2007 Public Service Agreements

No directly relevant.

The numbers

Graph 1 and 2

Decile Percentage change Numeric change per week (2008/09 prices) Share of total change
Poorest tenth -12% -£7 0%
2nd 25% £34 4%
3rd 29% £49 5%
4th 26% £55 6%
5th 24% £62 7%
6th 23% £68 8%
7th 22% £79 9%
8th 22% £93 10%
9th 23% £121 13%
Richest tenth 37% £365 40%

Graph 3 and 4

Year Share of total income
Poorest tenth 2nd 3rd 4th 5th 6th 7th 8th 9th Richest tenth
1994/95 2% 4% 5% 6% 8% 9% 11% 13% 15% 28%
1995/96 2% 4% 5% 6% 8% 9% 10% 12% 15% 28%
1996/97 2% 4% 5% 6% 8% 9% 11% 12% 15% 28%
1997/98 2% 4% 5% 6% 8% 9% 10% 12% 15% 28%
1998/99 2% 4% 5% 6% 7% 9% 10% 12% 15% 28%
1999/00 2% 4% 5% 6% 7% 9% 10% 12% 15% 29%
2000/01 2% 4% 5% 6% 7% 9% 10% 12% 15% 30%
2001/02 2% 4% 5% 6% 8% 9% 10% 12% 15% 28%
2002/03 2% 4% 5% 6% 8% 9% 10% 12% 15% 29%
2003/04 2% 4% 5% 6% 8% 9% 10% 12% 15% 29%
2004/05 2% 4% 5% 6% 8% 9% 10% 12% 15% 29%
2005/062% 4% 5% 6% 7% 9% 10% 12% 15% 30%
2006/071% 4% 5% 6% 7% 9% 10% 12% 15% 30%
2007/081% 4% 5% 6% 7% 9% 10% 12% 15% 31%
2008/091% 4% 5% 6% 7% 9% 10% 12% 15% 31%

Graph 5

Year 50:10 ratio 90:50 ratio
1994/95 234% 213%
1995/96 227% 214%
1996/97 238% 212%
1997/98 238% 213%
1998/99 234% 217%
1999/00 236% 213%
2000/01 234% 214%
2001/02 229% 212%
2002/03 227% 208%
2003/04 234% 209%
2004/05 228% 211%
2005/06236% 212%
2006/07244% 212%
2007/08246% 214%
2008/09249% 217%

Graph 6

Figures are as shown in the graph.

Graph 7

Region Percent
In the poorest fifth In the richest fifth
East 18% 22%
East Midlands 20% 16%
inner London 29% 28%
outer London 24% 28%
North East 21% 14%
North West 21% 16%
Northern Ireland18% 14%
Scotland17% 20%
South East 17% 27%
South West 18% 19%
Wales 20% 16%
West Midlands 22% 17%
Yorkshire and The Humber 21% 15%

Graph 8

Decile Adults of pensionable ageChildren in workless families Working-age adults in workless families Children in working families Working-age adults in working families Total
Poorest tenth 11% 12% 34% 14% 30% 100%
2nd 16% 15% 23% 18% 28% 100%
3rd 28% 6% 13% 21% 33% 100%
4th 25% 3% 10% 22% 41% 100%
5th 22% 1% 7% 22% 48% 100%
6th 21% 1% 6% 20% 53% 100%
7th 19% 0% 4% 17% 59% 100%
8th 16% 0% 4% 15% 64% 100%
9th 14% 0% 3% 14% 69% 100%
Richest tenth 13% 0% 2% 13% 71% 100%
1. ‘Real’ incomes are incomes  adjusted for inflation. 
2. Gini coefficients can range from 0 to 100, where it would have the value 0 if all households had identical incomes and a value of 100 if a single household had all the country’s income.  So, the lower the number, the lower the income inequality.  See Wikipedia for the precise technical definition.