financial
jiafeng1
article_2.pdf
417
The recent rapid growth in household debt has been
driven by high rates of both secured borrowing, through
mortgages, and unsecured borrowing, through personal
loans, overdrafts and credit cards. While the aggregate
stock of unsecured debt (£164 billion at end-2003 Q2)
is small in relation to the stock of mortgage debt
(£714 billion) and household sector gross wealth
(£5,547 billion, including housing), its fast growth has
raised questions about whether an increasing number of
people have borrowed more than they can easily afford
to repay. This would be more likely if unsecured
borrowing is increasingly concentrated among those
with relatively low incomes and few assets. Its build-up
could also leave others more vulnerable to unexpected
changes in their circumstances. These possibilities
might have implications for both monetary policy and
financial stability. An assessment of the extent of
current and potential problems associated with the
growth of unsecured debt requires some investigation of
how the debt is distributed among individual borrowers.
This article is solely concerned with the distribution of
unsecured debt and only discusses mortgage borrowing
to the extent that this is useful in distinguishing the
characteristics of borrowers.
An important source of information on the financial
position of individual adults and households is the
British Household Panel Survey (BHPS), which since
1991 has asked broadly the same group of people about
their economic and social circumstances. However, the
most recent information on unsecured debt in the BHPS
is for 2000. This was analysed in detail by Cox, Whitley
and Brierley (2002). They found that the households
with the highest levels of both mortgage and unsecured
debts tended also to have the highest levels of income
and net wealth in both 1995 and 2000. But they also
found that debt to income ratios were highest for
low-income households.
The survey In order to update this analysis, more recent evidence
has been obtained from a specially commissioned survey
from NMG Research, that in October 2003 asked a
nationally representative sample of 1,950 adults about
their unsecured debt. Using the same questions as in
the BHPS, people were asked about the types of debt
they had, the amounts they owed and whether they
considered the debt to be a burden to their
household.(2) A broad summary of the survey
The distribution of unsecured debt in the United Kingdom: survey evidence
The Bank recently commissioned a survey asking people about their unsecured borrowing and whether it is a burden to them. This article summarises the main results.(1) As of October, 34% of respondents had some form of unsecured debt, over and above that which they expected to pay off at the end of the month, and the average amount owed was around £3,500. Some people owed much more than the average: 26% of those with some debt owed more than £5,000. Around 10% of borrowers said that their unsecured debt was a heavy burden to their households, similar to earlier surveys. For purposes of comparison over time, the questions were based on those used in earlier surveys. The evidence suggests that the proportion of people with some debt has not changed since at least the late 1980s. While the average amount borrowed by debtors has increased, since 2000 the extra borrowing has been concentrated among those with household incomes above £17,500. Despite the rise in average debt levels in recent years, the proportion of people who consider their debt not to be a burden has increased. But, the amount borrowed and the share of unsecured debt accounted for by those who consider it a heavy burden have both increased.
(1) A brief summary was also reported in the Financial Stability Review, December 2003. (2) The BHPS asked people about the total amount borrowed on a range of debt instruments, the NMG Research Survey
also asked how much they owed on each debt instrument individually.
By Merxe Tudela and Garry Young of the Bank’s Domestic Finance Division.
418
Bank of England Quarterly Bulletin: Winter 2003
methodology and exact wording of questions is included
in the annex.
Only 11% of interviewees refused to say whether they
had any unsecured debt or not. But of those with debt,
33% did not say how much they owed. This may well
reflect uncertainty about the amount they owe rather
than a wish to conceal it. A recent survey, Citizens
Advice (2003), explicitly asked people whether they
knew how much they owed in total on all of their credit
commitments. It found that 31% of those owing
something on credits/loans did not know how much they
owed. Information from the NMG Research Survey
suggests that the perceived burden of debt among those
who do not say how much they owe is not different from
that among those who do report the amount they owe.
On this basis, we assume that the amount owed by those
who do not reveal their debt is not systematically
different from the amount owed by those who do reveal
it.(1)
Participation
Table A outlines some of the key findings concerning the
proportion of people using different methods of
borrowing (the participation rate). For each type of debt
instrument, people are asked not to include any
borrowing that they expect to pay off in full by the end
of the month. This is intended to exclude borrowing,
mainly on credit cards, that is done on a temporary
basis, because it is a convenient way of making
transactions.(2)
In total, 34% of respondents have some type of
unsecured debt. By instrument, 15% of respondents owe
money on a credit card, 13% have a personal loan, 8%
owe money on catalogue purchases and 7% have an
overdraft. In general, participation varies broadly
positively with the level of household income for most
types of borrowing, except DSS social fund loans, not
available to those on high incomes, and borrowing
through catalogues and mail order, which is used at a
similar rate throughout the income distribution.
Student loans are concentrated in the lowest-income
group, probably indicating that a high proportion of
borrowers are still studying. The overall rate of
participation in the unsecured debt market is highest for
those with annual household income of between
£25,000 and £34,999, around twice that for those with
household income below £9,500.
Average levels of debt by household income and debt instrument
Table B sets out the average amount of debt per
borrower. In total, the amount borrowed is also strongly
related to household income, with average debt of
around £1,800 for debtors with household income in
the range £4,500–£9,499 and nearly £6,000 for those
with household income in excess of £60,000. The
exception to this pattern is for debtors in the
lowest-income group, where average debt is £2,400,
reflecting a disproportionate number of people with
student debt in this group. While the participation of
low-income households as borrowers in the overdraft and
(1) The survey also asked about other individual and household characteristics, including household income. Around half of those interviewed did not provide information on their household income. While this is not the main purpose of the survey, information on income is important in assessing the affordability of debt.
(2) This is different to aggregate unsecured debt figures which include any unsecured debt outstanding at a point of time, regardless of whether it bears interest or not. This is discussed further in the box on page 421.
Table A Participation by debt instrument and household income Per cent
Household income of those reporting income:
Less than £4,500– £9,500– £17,500– £25,000– £35,000– More than Whole £4,500 £9,499 £17,499 £24,999 £34,999 £59,999 £60,000 sample
HP agreement 1 6 6 4 16 11 14 5 Personal loans 11 10 16 16 36 28 31 13 Overdraft 7 4 9 14 18 13 25 7 Credit card 5 6 22 22 33 35 20 15 Catalogue or mail order 7 9 13 9 8 8 7 8 Student loan 8 1 2 3 2 4 2 3 DSS Social Fund 4 7 4 1 1 0 0 2 Other loans 0 2 4 3 2 2 0 2 Any type of debt 33 29 45 42 62 56 48 34 Memo: Proportion of
sample in each income group 8 28 28 12 10 11 4 100
Note: Some people have more than one type of debt so the proportion of people with any type of debt is less than the sum of those with each type of debt. Individual responses have been weighted to reflect the UK population (see Annex for further details). Bank calculations.
Source: NMG Research Survey.
Fieldwork: October 2003.
The distribution of unsecured debt in the United Kingdom: survey evidence
419
credit card markets is relatively low, the average amount
they borrow using these products is relatively high. For
example, the average overdraft among those with income
of less than £4,500 is close to the overall average, while
the average credit card debt of those with household
income of between £4,500 and £9,499 is above the
overall average. By contrast, borrowing is strongly
increasing with income for personal loans and (to a
lesser extent) hire purchase (HP) agreements, possibly
reflecting their use as a method of financing irregular
large-ticket income-related spending. For these
instruments, borrowers have little discretion to increase
their debt without the permission of lenders. On
average, borrowing through student loans, personal
loans, HP agreements and credit cards is for larger
amounts than through overdrafts, catalogue or mail
order finance and DSS social fund loans. The overall
average debt of borrowers is £3,500.
Robustness
The averages in the individual cells in Table B need to be
treated with care because the sample size is small in
some cases (particularly for ‘other loans’), although the
overall averages for income groups and types of loans are
likely to be relatively reliable. More generally, the
robustness of the figures on participation and average
debt levels can be assessed by comparing them with
other evidence. Table C summarises information on
participation and average debt levels from other surveys.
Some difference is to be expected simply because each
survey is based on a small sample of a much larger
population. Moreover, the figures are not exactly
comparable because some surveys, such as that by NMG
Research, ask about the borrowing of individual adults
while others consider the position of households.
Evidence from the BHPS, which considers both, suggests
that participation rates are about 8 percentage points
higher for households than individuals. This is not the
only difference in that some surveys prompt the
interviewee about the types of debt they may have.(1)
Taking these differences into account suggests that, with
the exception of the KPMG and Citizens Advice surveys,
there is a consensus that around 40%–50% of
households and 30%–40% of adults have some form of
unsecured debt. The probable cause of the high rates of
participation found by the KPMG and Citizens Advice
surveys is that they do not specifically exclude loans that
people expect to pay off in full at the end of the month
in question.(2)
The average debt of debtors in the NMG Research
Survey is also broadly consistent with the findings of
other surveys. It is 8% higher than the average debt of
debtors in the 2000 BHPS. This difference is smaller
than might be expected given that the aggregate level of
unsecured debt was 35% higher in October 2003, the
date of the NMG Research Survey, than it was in
September 2000, when the BHPS was undertaken. This
difference could reflect sampling error. The NMG
Research estimate is less than the average debt of
£4,600 of individual debtors in the FSA survey carried
out in 2002. On a household basis, the NMG Research
Table B Average debt of debtors by debt instrument and household income £
Income levels of those reporting income:
Less than £4,500– £9,500– £17,500– £25,000– £35,000– More than Whole £4,500 £9,499 £17,499 £24,999 £34,999 £59,999 £60,000 sample
HP agreement n.a. 1,600 2,000 3,400 1,400 3,800 3,400 2,500 Personal loans 2,000 2,000 3,500 4,800 5,300 5,400 5,300 4,400 Overdraft 700 200 700 1,000 700 800 1,000 700 Credit card 800 1,800 1,000 900 2,300 2,800 1,100 1,400 Catalogue or mail order 200 400 300 300 100 400 n.a. 300 Student loan 5,600 6,500 4,000 4,500 6,100 4,300 n.a. 6,300 DSS Social Fund 300 300 200 n.a. n.a. n.a. n.a. 200 Other loans n.a. 300 500 6,100 n.a. n.a. n.a. 1,400 Any type of debt 2,400 1,800 2,400 3,800 5,100 5,600 6,000 3,500
Source: NMG Research Survey.
Fieldwork: October 2003.
Bank calculations. All figures rounded to nearest £100.
n.a. indicates that the number of debtors in this category is too small for the average to be sensibly calculated. Figures in italics are indicative only as the number of respondents in these categories is less than ten.
(1) The Survey of Low Income Families (SOLIF), analysed by Bridges and Disney (2004, forthcoming) includes loans from employers, family and friends and the ‘tally man’ in addition to loans from financial institutions.
(2) Some of the surveys also provide information on participation by type of debt instrument, which may be used to check the robustness of our findings. Kempson (2002) and FSA (2003) both find that around 19% of households have credit card debt outstanding, slightly higher than our estimate of 15% for adults. They both find that around 8% of households have overdrafts and some 15% have personal loans, close to our findings for adults.
420
Bank of England Quarterly Bulletin: Winter 2003
estimate of the average debt of debtors is equivalent to
£4,700, if the same relationship between the debt of
individuals and households applies as in 2000. This is
higher than the average figure of £3,500 given in
Kempson (2002). The average value of debt of
individuals is similar to the Citizens Advice estimate of
£3,900, although this is not strictly comparable as it
does not exclude balances to be paid off at the end of
the month, raising the participation rate substantially.(1)
The consistency of the survey-based figures with the
aggregate statistics is discussed in the box on
page 421.
Distribution of unsecured debt
Table D shows the amounts owed by debtors, using
information from the NMG Research Survey. This is
compared with other surveys containing similar
information.
The NMG Research Survey shows most debtors owe
relatively small amounts, with close to half of debtors
owing less than £1,000 and nearly two thirds owing less
than £3,000. But there is wide variation in the amounts
owed and 26% of debtors owe more than £5,000.(2)
These figures are close to those of other surveys,
although the KPMG survey finds higher debt levels. The
KPMG survey is distinctive in that it is an online survey,
whereas the others are carried out face to face. It is
representative in terms of gender, age, region and social
class, but it may be that the anonymity offered by an
online survey encourages people to reveal more about
their indebtedness. It could also be that an online
survey attracts a sample of participants more likely, for
whatever reason, to have larger debts.
The unequal distribution of debt across respondents in
the NMG Research Survey suggests that a large
proportion of outstanding unsecured debt is
concentrated among relatively few people. If the survey
Table C Evidence from surveys(a)
Date of survey Sample size Unit of analysis Excludes debt Additional Percentage with Average debt paid off at types of debt unsecured debt of debtors (£) end of month mentioned
Survey: Berthoud and Kempson 1989 2,212 Household Yes Store cards 48 Not given BHPS 1995 9,249 Individual Yes n.a. 35 2,088 BHPS 1995 5,031 Household Yes n.a. 43 2,872 Bridges and Disney 1999 4,659 Low-income No Loans from 49 Not given
families employers, friends, ‘tally man’
BHPS 2000 9,006 Individual Yes n.a. 35 3,242 BHPS 2000 4,916 Household Yes n.a. 43 4,375 Kempson 2002 1,647 Household Yes Store cards/ 47 3,500
accounts FSA 2002 3,200 Families Yes Rent arrears, Not given 4,600 (b)
store loan, store card, car loan
KPMG 2003 2,304 Individual No 71 (c) Not given Citizens Advice 2003 1,986 Individual No Includes some secured 74 3,900
loans and remortgaging
NMG Research 2003 1,950 Individual Yes n.a. 34 3,500
(a) See references for more details of the surveys. (b) Individual average. (c) Based on Bank calculations.
Table D The distribution of the amount owed by debtors Per cent
NMG Citizens KPMG Kempson Research Advice
Under £1,000 43 46 25 67 £1,000–£1,999 14 13 £2,000–£2,999 7 7 33 £3,000–£3,999 5 6 17 £4,000–£4,999 5 5 £5,000–£7,499 8 4 20 £7,500–£9,999 5 3 7 £10,000–£14,999 9 6 17 10 £15,000–£19,999 2 3 £20,000 or more 2 2 7
Note: KPMG and Kempson figures have been recalculated by the authors as proportions of debtors. Citizens Advice figures have been adjusted to take account of refusals. Figures may not sum to 100 due to rounding.
(1) There is also a broad degree of consistency across the surveys in estimates of the average amounts of debt for individual types of debt. For outstanding credit card balances, Kempson finds households owe an average of nearly £1,600 while the FSA finds families owe £2,200, both broadly consistent with the NMG Research estimate of £1,400 for adults. Similarly, average personal loans are estimated to be over £5,500 by the FSA and £5,000 by Kempson, compared with our estimate of £4,400 for adults. Overdrafts are also of a similar order across the surveys, with the average in Kempson’s sample put at £450 compared with £900 in the FSA sample and £700 here. Interestingly, Kempson notes that the average overdraft in the 1989 survey reported by Berthoud and Kempson (1992) was £1,700 in 2002 prices, whereas the average credit card balance in 1989 was £550 in 2002 prices, indicating a broad shift in the method of borrowing over the past 14 years. The average student loan debt of £6,300 in the NMG Research Survey is consistent with Callender and Wilkinson (2003) who find that the average student loan debt of students was £5,500 at the end of the 2002–03 academic year. This is of course much higher for final-year students.
(2) This is equivalent to 6% of respondents, including those who do not reveal how much unsecured debt they have, and 9% of respondents for whom the level of debt is known or estimated.
The distribution of unsecured debt in the United Kingdom: survey evidence
421
results are grossed up to the UK adult population, then
of the total unsecured debt of £56.8 billion, around
£29 billion is owed by the 4.3% of the population with
individual debt of over £10,000.
Distribution of debt by household income of the borrower
Table E looks at the breakdown of debt levels according
to the household income of the borrower. This is
derived from the NMG Research Survey and grossed up
to the UK adult population. It provides a guide to the
affordability of debt, given that individuals with more
debt and less income are most likely to experience debt
problems. In general, it shows relatively few cases where
debt levels are high relative to household income. For
example, of the 770,000 individuals with debt over
£10,000 whose income is known, only 185,000 have
household income of less than £17,500, while 340,000
have income in excess of £35,000.
Characteristics of borrowers and those who report debt to be a burden
The NMG Research Survey also provides qualitative
evidence, comparable to that in the BHPS, of the extent
Consistency with aggregate figures
While the NMG Research Survey evidence is broadly
consistent with other similar surveys, there is a
substantial difference between the grossed-up
amount of unsecured debt claimed in survey
responses and that reported in official statistics.(1)
At the end of September, total UK consumer credit
was £168.4 billion, whereas the grossed-up figure
from the NMG Research Survey was only
£56.8 billion.
There are a number of ways of reconciling these
figures.
First, there is a difference between what the surveys
and official figures intend to cover. Official estimates
are derived from lenders and cover all forms of
consumer credit outstanding at a particular date,
including that which does not bear interest. By
contrast, survey respondents are usually asked not to
include credit card and other bills being paid off in
full in the month of interview. So part of the
difference between survey and official figures is that
surveys are stating what people perceive to be their
normal stock debt position, whereas official aggregate
figures are also picking up credit used temporarily to
make transactions. This is an issue particularly for
credit card borrowing, given that many people now
use credit cards for transactions in preference to cash
and other means of payment, either because of
greater convenience or to take advantage of insurance
and cash-back special offers. The Citizens Advice
(2003) survey suggests that around half of all people
who use credit cards aim to pay them off in full at the
end of the month.(2) Unpublished figures from the
British Bankers’ Association suggest that about three
quarters of Visa and Mastercard affiliated credit card
balances bear interest.
Second, despite the consensus between surveys, it
may be that respondents have a persistent tendency
to underreport their debt, perhaps because they do
not think of goods being paid for in instalments as
debt. There could also be confusion within
individual families as to the amount borrowed by
other family members. This should be avoided in the
NMG Research Survey, which asked people about
their own borrowing. To the extent that people
included the debts of other family members in their
replies, this would have led to an over rather than an
underestimate of the total when grossed up by the
number of adults.
There is some evidence that the gap between the
lender-based and borrower-based estimates of
aggregate unsecured debt has been increasing over
time. Using information from the BHPS, the
proportion of the official, lender-based aggregate
estimate of unsecured debt that is accounted for
by survey-based information declined from 52% in
1995 to 41% in 2000. According to the NMG
Research Survey, this has fallen further to 34% in
2003.
(1) Note that this does not apply to BHPS estimates of labour income, housing wealth and secured debt which, when aggregated, largely agree with aggregate estimates. Financial assets, however, are substantially underrecorded.
(2) The recent CAB survey asked those in their sample that had at least one credit or store card (around 1,200) how much they usually pay off each month and found that: 2% pay off less than the minimum payment, 10% make the minimum payment, 29% pay more than the minimum payment but less than the full balance, 49% pay the full balance, 9% have no outstanding balance, 2% refuses or do not know.
422
Bank of England Quarterly Bulletin: Winter 2003
to which individuals consider the repayment of their
unsecured debt to be a burden to their household. This
is useful in making comparisons with previous years and
in assessing the consistency of answers in the survey. It
finds that 10% of individuals consider their unsecured
debt to be a heavy burden, 22% say that it is somewhat
of a burden and the remainder do not consider it a
problem. The survey enables us to examine the
characteristics of those who report different degrees of
financial distress. This is summarised in Table F.
Table E The distribution of unsecured debt and income (millions, estimates grossed up from survey respondents)
Household income:
Less than £4,500– £9,500– £17,500– £25,000– £35,000– More than Missing Whole £4,500 £9,499 £17,499 £24,999 £34,999 £60,000 £60,000 sample
Debt: None 1.2 4.3 3.6 1.6 0.9 1.1 0.4 14.5 27.4 Under £1,000 0.3 0.9 1.1 0.3 0.2 0.1 0.1 1.3 4.1 £1,000–£4,999 0.1 0.3 0.8 0.3 0.3 0.4 0.1 0.7 3.0 £5,000–£9,999 0.1 0.0 0.2 0.1 0.3 0.2 0.1 0.2 1.2 £10,000 or more 0.0 0.1 0.1 0.1 0.2 0.2 0.1 0.4 1.2 With debt but missing value 0.2 0.4 0.6 0.3 0.5 0.4 0.1 2.5 4.9 Missing 0.1 0.6 0.4 0.2 0.1 0.1 0.1 3.8 5.4 Total 1.9 6.6 6.7 2.8 2.4 2.6 0.8 23.4 47.2
Source: NMG Research Survey.
Fieldwork: October 2003.
Bank calculations.
Table F The characteristics of adults with different degrees of debt problems
Characteristics of each group: Contribution of group to overall shares:
Share of Participation Mean Proportion of Share of those Share of those Share of those Share of those Mean debt population rate debt to debtors with positive reporting debt reporting debt reporting debt for those
income reporting debt to be no to be somewhat to be a heavy reporting ratio of debt to be a problem of a problem burden debt to be debtors heavy burden a heavy
burden
Overall 34 0.24 10 34 68 22 10 6,900
Age group 15–24 15 33 0.45 11 16 14 18 17 5,300 25–34 19 54 0.19 13 29 26 33 37 6,600 35–44 18 50 0.18 10 25 25 25 25 5,900 45–54 16 39 0.19 9 17 18 15 15 12,400 55–64 13 26 0.13 5 9 11 6 5 Not given 65 plus 20 7 0.75 5 4 6 2 2 12,600
Income group Less than £4,500 4 33 1.03 24 4 4 3 10 4,300 £4,500–£9,499 14 29 0.29 15 12 9 18 18 3,300 £9,500–£17,499 14 45 0.19 13 19 18 20 25 6,700 £17,500–£24,999 6 42 0.18 7 8 8 7 5 15,800 £25,000–£34,999 5 62 0.17 7 10 11 8 7 9,700 £35,000–£60,000 5 56 0.13 0 10 11 9 0 n.a. £60,000 plus 2 48 0.07 11 3 3 1 3 13,500 Missing income 50 26 n.a. 9 35 37 34 32 7,300
Social class AB 19 36 0.14 6 19 21 18 11 7,500 C1 25 41 0.35 7 30 32 30 21 14,500 C2 23 33 0.26 8 22 23 21 17 12,800 DE 34 30 0.20 17 29 25 32 50 2,400
Debt group (for debtors only) Under £1,000 43 4 43 49 35 18 500 £1,000–£1,999 14 10 14 13 15 15 1,300 £2,000–£2,999 7 13 7 7 6 10 2,200 £3,000–£3,999 5 8 5 3 14 5 3,300 £4,000–£4,999 5 13 5 6 2 8 4,000 £5,000–£7,499 8 8 8 7 10 6 5,500 £7,500–£9,999 5 11 5 5 3 6 8,000 £10,000–£14,999 9 23 9 6 11 20 12,200 £15,000–£19,999 2 16 2 1 2 3 18,000 £20,000 or more 2 40 2 2 11 9 23,800
With financial assets 35 39 0.21 5 40 46 33 19 6,400
Housing status With mortgage 33 47 0.17 4 45 50 44 17 17,700 Own outright 26 18 0.28 5 14 17 9 7 9,600 Rented local authority 23 30 0.24 20 20 18 17 41 2,200 Rented private 10 42 0.49 17 12 9 18 21 10,200 Housing association 7 45 0.18 18 8 6 11 15 7,800
Source: NMG Research Survey.
Fieldwork: October 2003.
Bank calculations. Figures may not sum to 100 due to rounding.
The distribution of unsecured debt in the United Kingdom: survey evidence
423
By age, 37% of those who find debt to be a heavy burden
are between 25 and 34. This is about double their
weight in the grossed-up sample as a whole and also
higher than their share in the debtor population. These
individuals do not have a particularly high unsecured
debt to income ratio, but their finances are likely to be
put under strain by mortgage borrowing and other
financial demands at a stage in their life when they are
starting families and buying homes.
By income, 28% of those who find unsecured debt to be
a heavy burden have household income of less than
£9,500. While their participation rate is relatively low,
they have an above-average unsecured debt to income
ratio. Over half of those who say that unsecured debt is
a heavy burden are from the DE social class.(1) This is
much higher than their 29% representation in the
debtor population. Nonetheless, the average level of
debt for those from low income or social class
households who report their debt to be a heavy burden
is low relative to that owed by other groups.
By debt level, over 32% of those whose debt is perceived
to be a heavy burden have unsecured debt of over
£10,000. This is substantially larger than their 13%
share in the population of debtors. The average debt of
people who say it is a heavy burden is £6,900, compared
with £3,900 for those for whom it is somewhat of a
burden and £2,900 for those who say it is not a
problem. It is not surprising that those who report their
unsecured debt to be a heavy burden also tend to have
more of it. This suggests that unsecured debt in
aggregate is to some extent concentrated among those
who experience difficulties in repaying it. Almost 20%
of unsecured debt is owed by those who consider it a
heavy burden, and 25% by those who say it is somewhat
of a problem.
It is clear from the survey that unsecured debt is less of
a problem for those with other assets. Only 19% of
those reporting unsecured debt to be a heavy burden
have financial assets compared with 40% in the debtor
population as a whole. Only 17% of those who report
unsecured debt to be a heavy burden have a mortgage
compared with 45% in the debtor population as a
whole. By contrast, 41% of those for whom debt is a
heavy burden live in local authority rented
accommodation, compared with 20% in the debtor
population as a whole.
These figures on the proportion of people reporting
debt to be a heavy burden indicate that a significant
minority of debtors are struggling with their unsecured
debt. This is consistent with other evidence from debt
counsellors and the Citizens Advice Bureaux.
Comparisons over time
One of the key reasons for asking the same questions in
our survey as in the BHPS is to ensure, as far as possible,
that the results are comparable over time. These
combined results show that there has been no overall
change in the participation rate between 1995 and
2003. This confirms Kempson’s (2002) analysis of the
1989 to 2002 period that ‘the large increase in
consumer borrowing is not due to a larger proportion of
the population owing money’. This may seem surprising
in view of the substantial increase in credit cards in
circulation in recent years, but this does not necessarily
imply that a larger proportion of people borrow using
these cards.(2) Moreover, Kempson finds that greater use
of credit cards has substituted for other forms of credit
such as mail order catalogues and hire purchase
arrangements. Chart 1 plots the participation rate for
different age groups between 1995 and 2003. This
shows very little change in participation rates across age
groups in recent years. It also suggests that the NMG
(1) Social class is defined as follows: AB: Professionals, directors, self-employed people employing more than 25 staff, employees in senior positions with professional qualifications and/or a degree; C1: Office workers without a degree, employees in junior positions with professional qualifications and/or a degree, self-employed people employing 1–24 staff, students in full-time education; C2: qualified skilled manual workers; DE: semi-skilled and unskilled manual workers, shop assistants, cleaners, unemployed, retired on state pension only.
(2) The number of credit cards issued rose from 28.3 million in 1995 to 58.8 million in 2002 according to APACS (2003).
Chart 1 Unsecured debt participation rates by age group
0
10
20
30
40
50
60
16–24 25–34 35–44 45–54 55–64 65+
BHPS 1995
BHPS 2000
NMG 2003
Per cent
Sources: BHPS, NMG Research and Bank calculations.
424
Bank of England Quarterly Bulletin: Winter 2003
Research Survey is comparable with the earlier BHPS
results.
The surveys suggest that the average unsecured
debt to income ratio of debtors doubled between 1995
and 2003, with unsecured debt levels rising from 12% to
24% of income.(1) The increase between 1995 and 2000
was fairly well-spread across the income distribution,
albeit with the largest increase in the lowest-income
group. By contrast, the increase since 2000 has been
concentrated in households with income of £17,500 and
above. Indeed there has been a sharp fall in unsecured
debt among the lowest-income group since 2000.
Chart 2 shows average unsecured debt levels by income
group in the 1995 and 2000 waves of the BHPS and in
the 2003 NMG Research Survey.
Despite the increase in average debt levels and
debt to income ratios, comparable figures suggest that
this has not added to the proportion of households
experiencing debt problems. Chart 3 shows that the
proportion of debtors reporting that unsecured debt is
not a problem has increased from 58% in 1995 to 68%
in 2003, while the proportion reporting that it is
somewhat of a burden has declined from 31% to 22%
over this period. The proportion reporting debt to
be a heavy burden has been broadly stable at around
10%. It is not clear from this evidence what has caused
the decline in the proportion of people reporting debt
to be somewhat of a burden, but it is likely to be
associated with the fall in effective interest rates on
unsecured debt and the fact that, in aggregate,
unsecured debt has remained small relative to household
wealth.
Changes in the concentration of debt
There is some evidence that the concentration of debt
among riskier borrowers has increased over time, despite
the fact that the proportion of people reporting debt
problems has declined. Chart 4 shows that the average
Chart 2 Mean debt by household income groups
Sources: BHPS, NMG Research and Bank calculations.
1,000
2,000
3,000
4,000
5,000
6,000
7,000
Up to 4,499
4,500–11,499
11,500–17,499
17,500–24,999
25,000–34,999
35,000–59,999
60,000+
Debt (£)
0
BHPS 1995
BHPS 2000
NMG 2003
Income (£):
Chart 3 Trends in the burden of debt
Sources: BHPS, NMG Research and Bank calculations.
58 62 62 60 63 62 64 68
10 11 9 10 9 9 1011
30272931 27 29 27 22
0
10
20
30
40
50
60
70
80
1995 96 97 98 99 2000 01 02 03
Heavy burden
Somewhat of a burden
Not a problem Per cent
Chart 4 Changes in average debt by attitude to debt
0
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
Not a problem Somewhat of a
burden
Heavy burden
1995
2000
2003
£
Sources: BHPS, NMG Research and Bank calculations.
(1) This is broadly consistent with the increase in the aggregate unsecured debt to income ratio from 12.6% to 21.9% of household income between 1995 Q2 and 2003 Q2 although, as pointed out in the box on page 421, the level of aggregate debt is higher than is implied by the survey evidence. The level of aggregate income is also higher because the income of non-debtors is included.
The distribution of unsecured debt in the United Kingdom: survey evidence
425
amount of debt held by individuals who consider it to be
a heavy burden has increased by more than it has for
those who are more relaxed about their unsecured debt.
This reflects a sharp increase in the unsecured
debt to income ratio of those reporting debt to be a
heavy burden and implies that the level of borrowing at
which debt becomes a problem is higher now than in
the recent past. It also suggests that debt has become
more concentrated among riskier borrowers.
Summary
The main results of the NMG Research Survey analysed
in this article are as follows:
● Around 34% of adults have some type of unsecured
debt. Participation in the unsecured debt market
is greatest among the higher-income groups. The
participation rate peaks at 62% for those with
annual household income of between £25,000 and
£34,999. It is about half this rate for those with
household income below £9,500.
● Unsecured debt is strongly related to household
income, with average debt of £1,800 for debtors
with household income in the range
£4,500–£9,499 and £6,000 for those with
household income in excess of £60,000. The
overall average debt of borrowers is £3,500. The
average for households is estimated at about
£4,700.
● The aggregate amount of unsecured debt implied
by surveys of borrowers is substantially less than
lender-based official figures. This partly reflects
the exclusion of debt used for transactions from
borrower-based surveys.
● Most debtors report that they owe relatively small
amounts, with close to half of debtors owing less
than £1,000 and two thirds owing less than
£3,000. Nevertheless, there is considerable
variation in the amounts owed and 26% of debtors
report that they owe more than £5,000.
● The unequal distribution of debt across the sample
suggests that a large proportion of outstanding
unsecured debt is concentrated among relatively
few people. Around half of the identified aggregate
obtained by grossing up the survey results is owed
by around 4.3% of adults with individual debt of
over £10,000.
● About 10% of debtors say that their debt is a heavy
burden. Around half of these people are in social
class DE, around 40% live in local authority rented
accommodation, few have other assets, and a
disproportionate number are between 25 and 35.
● There does not appear to have been any upward
trend in the participation rate over time. But the
debt to income ratio of borrowers has doubled
between 1995 and 2003. Since 2000, the increase
has been concentrated among those with
household income above £17,500.
● There has been an increase since 1995 in the
proportion of households who consider their debt
not to be a burden despite the general rise in
unsecured debt. This may reflect falling interest
rates on unsecured debt and the relatively small
share of unsecured debt in household sector
wealth. But the debt levels of those reporting it to
be a burden have increased disproportionately.
In summary, the recent growth of unsecured debt has
not as yet caused widescale debt problems. Moreover,
the survey evidence suggests that most unsecured
borrowing is arranged through personal loans and HP
agreements where borrowers have little discretion to
increase their debt without the permission of lenders.
There is, however, evidence of large unsecured
borrowing relative to income by a small proportion of
individuals. This may not necessarily be a problem
currently, although it could become one if these
borrowers experienced adverse financial shocks,
such as unexpected increases in interest rates or falls in
income.
426
Bank of England Quarterly Bulletin: Winter 2003
NMG Research runs a monthly omnibus survey,
MarketMinder, of the investments, savings and pensions
markets. To compile this omnibus survey, a nationally
representative sample of 2,000 adults aged 15 and over
are interviewed each month by Ipsos Capibus.
Interviews are conducted face to face, in respondents’
homes, using Computer Assisted Personal Interviewing.
Interviews are carried out in around 120 areas of the
country selected using a two-stage stratified random
location design. Different local areas are selected for
each month.
In order to correct for minor deviations in terms of the
generated sample profile week on week, Capibus uses a
rim-weighting system which weights to their National
Readership Survey defined profiles for age, social grade,
region and working status—within sex.
Three questions on unsecured debt were added to this
omnibus for interviews carried out between 3 and
9 October 2003. The questions were:
● I would like to ask you now about financial
commitments you may have apart from mortgages
and housing-related loans. Do you currently owe
any money on the following types of loan or credit
agreement? (please exclude any borrowing that will
be fully repaid at the end of the month, eg the
settling of your credit card in full).
1 Hire purchase agreement.
2 Personal loan.
3 Overdraft.
4 Credit card.
5 Catalogue or mail order agreement.
6 Student loan.
7 DSS social fund loan.
8 Any other loans.
● Can you please tell me how much money you
currently owe to the following types of lender (but
please exclude any borrowing that will be fully
repaid at the end of this month, eg the settling of
your credit card in full).
1 Hire purchase agreement.
2 Personal loan.
3 Overdraft.
4 Credit card.
5 Catalogue or mail order agreement.
6 Student loan.
7 DSS social fund loan.
8 Any other loans.
● To what extent is the repayment of such debts and
the interest a financial burden on your household?
Would you say it is:
1 A heavy burden.
2 Somewhat of a burden.
3 Not a problem.
Table 1 summarises the number of responses to the
questions about debt and household income.
Table 1 Summary of survey responses
Has no debt Declares debt Has some debt Refuses Total level but does not
say how much
Declares income 496 285 96 62 939 Does not
declare income 637 120 100 154 1,011 Total 1,133 405 196 216 1,950
Source: NMG Research Survey.
Fieldwork: October 2003.
Bank calculations.
Methodology
The distribution of unsecured debt in the United Kingdom: survey evidence
427
References
APACS (2003), Plastic Card Review.
Berthoud, R and Kempson, E (1992), Credit and debt: the PSI Report, Policy Studies Institute, London.
Bridges, S and Disney, R (2004), ‘Use of credit and arrears on debt among low income families in the United
Kingdom’, Fiscal Studies, forthcoming.
Callender, C and Wilkinson, D (2003), ‘2002/03 student income and expenditure survey: student’s income,
expenditure and debt in 2002/03 and changes since 1998/99’, Department for Education and Skills’, Research Report
RR487.
Citizens Advice (2003), Financial over-commitment survey, Research Study conducted by MORI for Citizens Advice,
July 2003.
Cox, P, Whitley, J and Brierley P (2002), ‘Financial pressures in the UK household sector: evidence from the British
Household Panel Survey’, Bank of England Quarterly Bulletin, Winter, pages 410–19.
Financial Services Authority (2003), Financial Risk Outlook.
Kempson, E (2002), Overindebtedness in Britain: a report to the Department of Trade and Industry.
KPMG (2003), Debt crisis—an attempt to make ends meet or a frivolous approach to spending?, Press Notice describing
survey carried out by YouGov on behalf of KPMG.
