Theme: Social and Welfare
The Households Below Average Income statistical report (HBAI) for the
period 1994/5-1999/00 is published today.
The report, the twelfth in the HBAI series, uses household disposable
incomes, adjusted for household size and composition, as a proxy for
material living standards. It principally gives information on the
income distribution in Great Britain from the financial years 1994/5
to 1999/00 using cross-sectional data from the DWP Family Resources
Survey (FRS). It also draws upon data from the British Household
Panel Survey (BHPS), run by the University of Essex, which unlike the
Family Resources Survey, tracks the same individuals over time to
give a picture of income mobility.
In order to allow comparisons of living standards between different
household types, income is adjusted to take into account variations
in the size and composition of the household in a process known as
equivalisation. This adjustment reflects the common sense notion
that a family of several people requires a higher income than a
single person in order for both households to enjoy a comparable
standard of living. A key assumption made in HBAI is that all
individuals in the household benefit equally from the combined
(equivalised) income of the household.
HBAI employs two measures of (net equivalised household) income:
Before Housing Costs (BHC) and After Housing Costs (AHC). Each
measure has imperfections as a guide to differences in, and changes
in, living standards, but the two are complementary. Results are
mainly presented excluding households containing full-time
self-employed individuals, as reported income is a poor guide to
living standards for this group.
Key findings in the report are as follows (tables referred to are
publication tables unless otherwise stated):
The income distribution in 1999/00 - actual money values
As explained above, household incomes in HBAI are adjusted or
equivalised in order to facilitate comparisons between different
family types. The majority of monetary amounts presented in HBAI are
on this equivalised basis and do not reflect actual money amounts for
any family type other than a couple with no children (the benchmark
used for equivalisation).
The table below shows the actual household incomes for several
examples of different family types, corresponding to various income
levels in the HBAI 1999/00 Before Housing Costs income distribution.
The income distribution in 1999/00
- Just over 60 per cent of individuals had incomes lower than the
national mean average. [Main table F1]
- Individuals in workless families were much more likely to live in
low-income households than those with one or more adults in
full-time work. Workless households with children had a particularly
high risk of low incomes. Families containing only part-time workers
were also more at risk of low incomes. [Tables 2.1, 2.3]
- Families with children were more at risk of low-incomes than their
childless counterparts, with individuals in lone parent families
being particularly at risk of low-incomes.
[Tables 2.2, 2.11]
- Single female pensioners were more at risk of low incomes than
single male pensioners or pensioner couples. [Table 2.5]
- Women were slightly more at risk of low incomes than men. [Tables 2.5, 2.11]
- Amongst ethnic minorities, individuals in Pakistani and Bangladeshi
families were particularly at risk of low income. [Table 2.6]
- Just under one in four individuals in low-income households did not
have access to a bank account. [Table 2.10]
Different social groupings
The Government's report on tackling and monitoring poverty and social
exclusion, Opportunity For All (OFA) contains three low-income
indicators for each of the mutually exclusive groups; children,
working-age adults and pensioners. These three indicators are:
i) A relative low-income indicator - the proportions of each group
that falls below thresholds of contemporary mean or median income.
ii) An 'absolute' low-income indicator - the proportions of each
group that fall below fixed thresholds of 1996/7 mean and median
income, uprated in real terms each year.
iii) A persistent low-income indicator - the proportion of each group
experiencing low-income for 3 out of 4 years.
Tables 1 to 3, summarising the above indicators, along with a similar
Table 4, for the whole population, can be found in Annex A. Key
results are as follows:
- There were large falls across the board in 'absolute' low-income
counts from 1996/7-1999/00,
- There was a fall in the number of children below 60 per cent of
contemporary median income from 1996/7-1999/00. However, this is not
as clear-cut in the mean-based indicator figures, which show less
- Relative low-income indicators for all individuals and working-age
adults were broadly constant from 1996/7-1999/00.
- Relative low-income indicators for pensioners have tended to
fluctuate from year to year and consequently show no clear trend.
Broadly speaking, the number of pensioners below contemporary
low-income thresholds has shown little or no change from
- It is not possible to draw conclusions about trends in persistence
from 1997, as this indicator is measured over a four-year period, and
the period 1996-1999 is the latest period for which data is
Other results for children, working-age adults and pensioners are
summarised as follows:
- From 1994/5-1999/00 there was a large fall in the numbers of
children below 'absolute' type low-income thresholds, that remained
fixed in real terms. Over the same period, there was little or no
change in the proportion of children below thresholds of contemporary
median income and a slight rise in the percentage of children below
thresholds of contemporary mean income. [Main tables H2, H6]
- Overall, children were over represented at the bottom of the income
distribution and were under represented at the top. [Table 5.2]
- Children in single parent families were much more likely to live in
low-income households than those in families with two adults.
[Tables 5.1/2, 5.7]
- Children in workless families were much more likely to live in
low-income households than those with one or more adults in full-time
work. [Tables 5.1, 5.7]
- Children in families with three or more children were much more
likely to live in low-income households than those in families with
one or two children. [Tables 5.3, 5.7]
- Children in families where the head of the family was aged under 30
were more likely to live in low-income households than those where
the head was in older age groups. [Table 5.7]
- Children in families containing one or more disabled persons,
either a child or an adult were more likely to live in low-income
households than those in families with no disabled persons. [Tables
- Children living in households headed by someone from an ethnic
minority community were more likely to live in low-income households.
This was particularly the case for those headed by a member of the
Pakistani or Bangladeshi communities. [Tables 5.5, 5.7]
- The percentage of working-age adults below contemporary low-income
thresholds showed little or no change from 1994/5 -1999/00. The
proportion of this group below 'absolute' type low-income thresholds,
that remained fixed in real terms, showed a fall over the period,
although less so than for pensioners and children. [Main tables H3,
- Overall, working-age adults were under represented at the bottom of
the income distribution and over represented at the top. [Table 6.1]
- Working-age adults in workless families were much more likely to
live in low-income households than those in families with one or more
adults in full-time work. [Tables 6.1, 6.9]
- Working-age adults with children were more likely to live in
low-income households than their childless counterparts. This was
particularly true for lone parents. [Tables 6.2, 6.9]
- Working-age adults with three or more children were more than twice
as likely as those with one child to live in low-income households.
[Tables 6.2, 6.9]
- Men without children had a similar distribution of income to women
without children. Women with children were more likely to live in
low-income households than were men with children. [Table 6.3]
- Working-age adults in families where the head of the family was
aged under 24 or over 54 were more likely to live in low-income
households than other age-groups. [Table 6.4, 6.9]
- Working-age adults living in households headed by someone from an
ethnic minority community were more likely to live in low-income
households. This was particularly the case for those headed by a
member of the Pakistani/Bangladeshi communities. [Tables 6.5, 6.9]
- Working-age adults living in families containing one or more
persons who were disabled - in the sense of having a long term
illness, disability or infirmity that limited their activity in some
way - were more likely to live in low-income households. The effect
was especially marked for those living in households containing both
a disabled adult and a disabled child. [Tables 6.6, 6.9]
- Working-age adults who were disabled or had a disabled partner,
accounted for one third of those in households below 50 per cent of
mean or 60 per cent of median income. [Table 6.8]
- Working-age adults with no educational qualification were more
likely to live in low-income households. [Tables 6.7, 6.9]
- The proportion of pensioners below 'absolute' low-income
thresholds, that remained fixed in real terms, showed a large fall
between 1994/5 and 1999/00. Over the same period, there was
fluctuation in proportion of pensioners below low-income thresholds
that vary with mean or median income, meaning that no consistent
trend could be identified. [Main tables H4, H8]
- Overall pensioners were over represented in the lower part of the
income distribution and under represented at the top. However
pensioners were not over represented in the bottom fifth of incomes
after housing costs. [Table 7.1]
- When the After Housing Costs measure was used there were generally
fewer pensioners in the bottom quintile and more in the higher
quintiles than on a Before Housing Costs basis. [Table 7.1]
- Older pensioners were more likely to live in low-income households,
although single pensioners over the age of 80 were slightly less
likely than those aged 70-80. [Table 7.1]
- Single female pensioners as a whole were more likely to live in a
low-income household than their male counterparts. [Tables 7.4, 7.7]
- Pensioners in a household headed by someone from an ethnic minority
community were more likely to live in low-income households.
HBAI provides an insight into the income distribution of residents of
the countries and regions of Great Britain. It should be noted that
no adjustment has been made for regional cost of living differences,
as the necessary data is not available. In the summary that follows
it is therefore assumed that there is no difference in the cost of
living between regions, although the After Housing Costs measure will
partly take into account differences in housing costs.
- Overall, individuals in Wales were more likely to live in
low-income households than those in Scotland and England. [Tables
- Children in Wales were more at risk of low incomes, with there
being little consistent difference between those in Scotland and
England. Children in Scotland were more likely to be in the upper
part of the income distribution than their counterparts in England or
Wales. [Tables 3.3/4]
- Among working-age adults, those in Wales were most likely to live
in low-income households, followed by those in Scotland. [Table
- In Wales, benefits made up a larger proportion of total gross
income than in both England and Scotland, whereas earnings were a
larger component in England than for the other two countries.
- Individuals living in the North East, North West and Merseyside,
and Yorkshire and the Humber were over represented towards the bottom
of the income distribution. On the After Housing Costs income measure
only, so were those in London. [Table 3.10/11]
- Individuals living in the Eastern region, the South East and London
were over represented towards the top of the income distribution.
- Children and working-age adults in the North East and Yorkshire and
the Humber had a relatively high risk of low incomes Before Housing
Costs. In addition, those in London and the North West and Merseyside
had a high risk on an After Housing Costs basis.
- Children and working-age adults in the South East and Eastern
regions were least at risk of low incomes. Working-age adults in the
East Midlands also had a relatively low risk. [Tables 3.14/15]
- Pensioners in Yorkshire and the Humber and the East Midlands tended
to have a relatively high risk of low incomes, with those in London
having a relatively high risk on an After Housing Costs basis only.
[Tables 3.16/17] Changes in income levels from 1994/5 to 1999/00
Table 5 in Annex A shows how real income has changed throughout the
income distribution since 1994/5.
- Over the period 1994/5-1999/00, there was significant growth spread
fairly evenly across the whole of the income distribution.
- Mean income grew by 12 per cent BHC between 1994/5 and 1999/00, and
15 per cent AHC.
- Median income grew by 10 per cent BHC, and 13 per cent AHC for the
distribution including the self-employed. Median growth for the
distribution excluding the self-employed was similar.
- Overall income inequality rose slightly between 1994/5 and 1999/00.
There was no clear change between 1998/9 and 1999/00.
Dynamics and persistence of low incomes
Analysis elsewhere in the report is based on the Family Resources
Survey (FRS), which is a cross-sectional survey. Therefore, the same
people are not interviewed in consecutive years and it is not
possible to trace the incomes of particular individuals over time.
The extent to which individuals living in low-income households at a
point in time are simply experiencing transitory fluctuations in
income, or are persistently experiencing low income, is an important
issue, that can only be addressed using longitudinal data. The
question of income mobility and the extent of persistence of low
incomes experienced by different groups can be addressed using data
from the British Household Panel Survey (BHPS).
The results presented provide a picture of income mobility in the
1990s and are based on data from nine waves of the BHPS, from 1991 up
to, and including, 1999 and show that:
- The proportion of children experiencing persistent low incomes
(below thresholds of median income for 3 out of 4 years) was broadly
constant from 1992-1999, as was the proportion for working age
adults. The proportion of pensioners below low-income thresholds
showed evidence of a rise over this period. [Table 8.6]
- More than half of those in the bottom quintile of the income
distribution in 1991 had moved up the income distribution by 1999.
- Over the period 1991-1999, half the population had spent at least
one year in the bottom 20 per cent of the income distribution, and
just over 60 per cent one or more years in the bottom 30 per cent.
- For both of the periods 1991-1994 and 1996-1999, nearly half the
population spent at least one year in the bottom 30 per cent, and
one-third in the bottom 20 per cent of the income distribution.
- For both of the periods 1991-1994 and 1996-1999, 7 per cent of
individuals were in the bottom quintile for all four years and around
15 per cent in the bottom 30 per cent for all four years. [Table
- One quarter of the population spent at least 5 years and one sixth
of the population spent 7 years, between 1991 and 1999, in the bottom
30 per cent of the income distribution. [Table 8.5]
- Only around 6 per cent of the population spent all 9 years, from
1991 to 1999, in the bottom 30 per cent of the income distribution.
- Women and children were more likely to experience relatively low
income and persistent relatively low income than men. [Table 8.8]
- Pensioners, individuals living in single parent families, workless
households, the social rented sector or those with no qualifications
were significantly more likely to experience persistent relatively
low incomes than other groups. [Table 8.8]
- Around seven out of ten individuals in workless households and over
half of pensioners and lone parents experienced relatively low
incomes for at least one year out of every four. [Table 8.8]
- Whilst there was no significant change in the persistence of
relatively low incomes for the population as a whole between 1991 and
1999, lone parents had a decreased risk and pensioner couples an
increased risk of relatively low incomes by the end of the period.
Table 1: Low-income indicators for children
Table 2: Low-income indicators for Working-age adults
Table 3: Low-income indicators for pensioners
Table 4: Low-income indicators for all individuals
Table 5: Changes in income levels from 1994/5-1999/00
1. 'Households Below Average Income - 1994/5 to 1999/00' is published by Corporate Document Services on behalf of the Department for Work and Pensions and is a National Statistics publication.
2. This is the twelfth edition.
3. HBAI is available on the DWP website.
4. The income definition used in the main analysis in HBAI, derived
from the Family Resources Survey (FRS) is as follows: net earnings;
profit or loss from self-employment after income tax and NI; all
social security benefits and tax credits, including Social Fund
grants; occupational and private pension income; investment income;
maintenance payments; top-up loans and parental contributions for
students, educational grants and payments; the cash value of certain
forms of income in kind such as luncheon vouchers, free meals/food
from employers, free coal and coke and free milk (where data
available). HBAI income is net of: income tax payments; NI
contributions; contributions to occupational and personal pension
schemes and personal pensions; council tax; maintenance and child
support payments made; and parental contributions to students living
away from home.
5. Housing costs are made up of: rent (gross of housing benefit);
water rates and community water charges; mortgage interest payments
(net of tax relief); structural insurance premiums (for owner
occupiers); and ground rent and services charges.
6. Information from the DWP survey, the Family Resources Survey, was
introduced into the HBAI series after a review of methodology in
7. The Family Resources Survey does not collect information on
individuals living in institutions, e.g. nursing homes, barracks or
jails; and homeless people living rough or in bed and breakfast
accommodation. Consequently HBAI does not cover these individuals.
Survey data are grossed to national totals and an adjustment is made
using the Inland Revenue's Survey of Personal Incomes (SPI) to ensure
very high income cases are correctly represented in HBAI.
8. HBAI is the only official statistical series that allows
consistent comparisons of disposable income over time, for different
parts of the income distribution. The income measures used have
been up-rated to enable one year's data to be compared with another.
This edition focuses on results which cover the period since 1994/5.
Some results are presented from the Family Expenditure Survey which
look at the period between 1979 and the mid 1990s. Following a
consultation exercise with outside academics and other Government
departments, results from the two survey sources are presented
together, using agreed methodologies to provide a longer time series.
9. HBAI aims to rank people in each year according to their standard
of living. To do this HBAI uses household income. Using household
income assumes that all income into the household is pooled and that
all members of the household benefit equally from the pooled income,
and so can be assigned a common income level. To avoid ranking a
four person household on£201 a week above a single person on£200 a
week the level of household income is adjusted, or 'equivalised', for
household size and composition.
10. In previous years, analysis comparing household expenditure
levels using the Family Expenditure Survey (a survey run for the
Office for National Statistics which captures Household expenditure)
indicated that the standard of living of people in the bottom decile
is no worse than that of people in the second decile; and the
expenditure levels of zero or negative income cases place nearly half
of them in the top half of the expenditure distribution. Many of
these latter households were self-employed. Selected results have
therefore been presented excluding the self-employed and main tables
are provided including and excluding the self-employed. Even so,
there remain a number of non-self-employed cases of very low incomes
recording relatively high expenditures. For these reasons, results
for the bottom end of the income distribution should not be
interpreted as relating unambiguously to the bottom end of a
distribution of living standards.
11. Neither BHC or AHC income is ideal as a guide to differences in,
and changes in, living standards, but the two are complementary. A
single figure is sometimes quoted when the difference between these
two measure of income does not materially affect the statement.
12. Information on income mobility is derived from the British
Household Panel Study, a survey run by ESRC UK Longitudinal Studies
Centre at the University of Essex. The income mobility estimates
presented here have been weighted in an attempt to remove biases that
can arise due to differential attrition. However, it is possible
that some bias remains. In addition, measurement error may be
affecting some of the mobility estimates.
13. The government's annual report on poverty and social exclusion,
'Opportunity for All', published in September 2000, gave details of
success measures and indicators related to poverty and social
exclusion. These include indicators based on percentages of
individuals below income thresholds. One of these indicators,
persistence of low income for children, is also a Sustainable