On this page9 sections
People in the UK owed £1,969.3 billion at the end of May 2026. Americans held $1,351.1 billion in revolving consumer credit in June 2026, and were charged $160 billion in credit card interest across 2024. Those numbers get quoted constantly, and they are almost always used to say something they cannot support.
This report does two things. It sets out what households actually owe, using only figures we have verified against the publisher's own page. And it works out what a given balance costs per year at the interest rates lenders were actually charging in mid-2026 — because the total owed is not the story. The rate attached to it is.
The short version: aggregate debt is dominated by mortgages, which are cheap; the debt that hurts household budgets is a small slice of the total carried at three to seven times the mortgage rate. A £2,000 overdraft costs more per year than a £10,000 personal loan.
Methodology, and what is not comparable
Every figure below comes from one of these sources, at the vintage stated:
- Bank of England quoted household interest rates (series IUMCCTL, IUMHPTL, IUMBX67, IUMODTL, IUMBV34), July 2026 data published early August 2026.
- The Money Charity, Money Statistics, July 2026 edition — a monthly compilation of Bank of England series, covering May and June 2026 data. The report is sponsored by a lender (Vanquis), though the underlying numbers are official Bank of England series.
- StepChange Debt Charity, Statistics Yearbook 2025, published March 2026, covering 163,916 clients advised in calendar 2025.
- Financial Conduct Authority, Financial Lives 2024 survey, fieldwork February to June 2024, published May 2025.
- Registry Trust, Q4 2025 court judgment statistics.
- Federal Reserve Board, G.19 Consumer Credit release of 7 August 2026.
- CFPB, The Consumer Credit Card Market: Report to Congress, published December 2025 and covering calendar 2024.
- OECD, National Accounts at a Glance, August 2026 vintage.
- Experian, 2025 Consumer Credit Review — bureau file data as at September 2025, used for one chart of US per-product balances.
Five cautions, up front rather than buried:
Aggregate totals and per-household averages are different animals. £1,969.3 billion is what the country owes. £67,802 is that total divided by the number of households. It is not what a typical household owes, and it is not what a typical borrower owes. We divide only where the publisher has done the division itself, and we say so each time.
Mean is not median. Where a source publishes both, we give both. StepChange clients owed a mean of £19,701 and a median of £12,026 in 2025 — the mean is dragged up by a tail of very large debts.
Quoted rates are not rates paid. Bank of England quoted rates are weighted averages of what lenders advertise. They exclude 0% promotional balances, and nobody's actual statement rate is guaranteed to match. Similarly, the CFPB's 25.2% average APR on US general-purpose cards in 2024 and the Fed's 22.15% for accounts assessed interest in Q2 2026 are different measures of different populations, not a contradiction.
The UK and US card rates in this report are not the same measure, and should not be read against each other. The Bank of England's 24.71% is a representative advertised rate — what lenders quote to a notional customer. The Federal Reserve's 22.15% is the rate actually charged on accounts that were assessed interest. The CFPB's 25.2% is an average of stated APRs across accounts. Three different definitions on three different populations. We compare how each series has moved over time, which is legitimate, and we do not claim UK cards are dearer than US cards, which the data cannot support.
People in debt advice are not the general population. The StepChange figures describe people who have sought help. They are the right numbers for describing problem debt and the wrong numbers for describing the average household.
One gap worth naming: we could not verify the New York Fed's Household Debt and Credit series, because newyorkfed.org blocked our verification pass at the domain level. US totals here therefore come from the Federal Reserve's G.19, which excludes mortgages and any loan secured on property, and from the CFPB. Do not read the G.19 total as "total US household debt" — it is not the same measure.
What is owed, and what kind of debt it is
The UK total splits cleanly, and the split is the first thing that gets lost when a headline says "household debt hits £2 trillion".
UK personal debt, end of May 2026
£1,969.3 billion in total. Mortgages are 87% of it.
- Secured (mortgages)£1,714.5bn87%
- Unsecured consumer credit£254.8bn13%
Source: The Money Charity, Money Statistics, July 2026 edition, compiled from Bank of England data
Within that unsecured £254.8 billion, The Money Charity puts outstanding credit card debt at £79.9 billion in May 2026, up 7.33% — £5.5 billion — over the year.
The same report performs the division we warned about, and it is worth quoting precisely because of how it is usually misread. Average total debt per UK household, including mortgages, was £67,802 in May 2026. Of that, £59,030 is secured. So the average unsecured debt per household is about £8,772, and the average credit card debt per household is £2,751. These are arithmetic averages across all households, including the millions with no mortgage and no card balance. They describe the country, not a family.
The US picture is a different measure again. Federal Reserve G.19 data for the second quarter of 2026 shows $1,858.2 billion in student loans and $1,571.5 billion in motor vehicle loans outstanding, with revolving credit at $1,351.1 billion in June and total consumer credit at $5,166.9 billion. Every one of those is an aggregate. We are not dividing them by a household count, because we do not have a sourced household count from the same publisher on the same basis — and that division is exactly where most "average American debt" claims go wrong.
Where a per-borrower average does exist, it is narrower than it looks. Experian's bureau file data gives an average balance per consumer who holds that product:
US average balance by credit product
Per consumer holding that product — not per household, and not per adult.
Source: Experian, 2025 Consumer Credit Review, As of September 2025
Debt against income: the one genuinely comparable table
Comparing debt levels across countries in pounds or dollars is meaningless. The OECD's household debt-to-net-disposable-income measure is built to be comparable, and it is the single most useful table in this report.
Household debt as a share of net disposable income, 2024
Loans and other accounts payable, measured against annual net disposable income.
Source: OECD, National Accounts at a Glance (measure LES1M_FD4), 2024 calendar year, August 2026 data vintage
Two things to take from it. First, the ranking tracks mortgage markets more than credit card habits — the Netherlands, Australia and Denmark sit at the top because of housing debt, not consumer borrowing. That is the same force that makes shelter the largest line in every national spending survey, though where each statistics agency files mortgage interest moves the housing share by nine points before any household behaviour is involved. Second, and less widely reported: the ratio fell in most countries between 2023 and 2024. Australia went from 214.0% to 209.6%, the UK from 137.1% to 130.8%, the US from 102.6% to 98.9%. There is no rising-debt-crisis trend in this table, and we are not going to manufacture one.
If you want to run the same calculation on your own numbers rather than a national aggregate, the debt-to-income calculator uses the conventional monthly-payment version of the ratio, which is the one lenders apply.
The rates that actually apply
Here is where the household budget lives. All five of these are Bank of England quoted rates for the same month, July 2026, so they are directly comparable with each other.
UK quoted borrowing rates, July 2026
Weighted averages of advertised rates. Bank Rate was 3.75%.
Source: Bank of England, quoted household interest rates, Series IUMODTL, IUMCCTL, IUMBX67, IUMHPTL, July 2026; Bank of England, Interactive Statistical Database, Series IUMBV34, July 2026
The cheapest line on that chart is also the one most often left to drift. The average UK revert-to rate — where a fix lands when it expires — was 6.60% in the same month against 4.79% on a new two-year fix, which on an illustrative £216,000 loan is about £236 a month for doing nothing when a deal ends.
That £5,000-versus-£10,000 gap is real and it is counterintuitive: borrowing less costs a higher rate. The standard caveat applies, though — borrowing more than you need to reach a cheaper tier only saves money if you genuinely do not spend the extra.
Rates on cards have barely moved, while loan pricing has swung. Over eighteen months the Bank's representative credit card rate travelled six basis points; the £10,000 loan rate fell to 6.27% in February 2026 and then climbed every month.
Card rates barely move. Loan rates do.
Bank of England quoted rates, selected months
- Credit card (representative)
- Personal loan, £10,000
Show the data
| Period | Credit card (representative) | Personal loan, £10,000 |
|---|---|---|
| Jan 2025 | 24.7% | 6.7% |
| Jun 2025 | 24.7% | 6.7% |
| Dec 2025 | 24.7% | 6.5% |
| Jan 2026 | 24.7% | 6.4% |
| Mar 2026 | 24.7% | 6.3% |
| Apr 2026 | 24.7% | 6.3% |
| Jun 2026 | 24.7% | 6.7% |
| Jul 2026 | 24.7% | 6.9% |
Source: Bank of England, quoted household interest rates, Series IUMCCTL and IUMHPTL
The US card rate has moved far more. The Fed's G.19 measures the rate on accounts actually assessed interest — the right measure for the cost of revolving, because it excludes cardholders who clear the balance and are charged nothing. That also makes it a different animal from the Bank of England's advertised rate above, so read the two charts as separate trend lines rather than as a UK-versus-US price comparison.
US credit card APR on accounts assessed interest
Annual averages. The 2026 Q2 reading was 22.15%.
Source: Federal Reserve Board, G.19 Consumer Credit, Terms of Credit, release of 7 August 2026
The CFPB measures something different and gets a higher number: an average APR of 25.2% on general-purpose cards and 31.3% on store cards in 2024, based on offered rates across accounts. Both are correct for what they measure.
What a balance costs per year
APR already accounts for compounding within the year, so a balance held flat for twelve months costs roughly balance × APR. The figure below applies the July 2026 quoted rates to a fixed £2,000. It is a worked example, not measured data, and it assumes you make no repayments at all — a deliberately simple ceiling on the cost, not a forecast.
Annual interest on £2,000, by product
£2,000 held for twelve months at July 2026 quoted rates. Illustrative.
The same arithmetic on the US average balance per cardholder: the CFPB puts it at about $5,300 in 2024, rising to roughly $8,700 for prime-scored cardholders. At the Q2 2026 rate of 22.15%, $5,300 held for a year would cost about $1,174 — but that is a ceiling on a hypothetical, not a typical bill. The $5,300 is averaged across all cardholders, including the roughly half who clear their statement every month and are charged nothing, while the 22.15% applies only to accounts that were actually assessed interest. The two are being deliberately forced together here to show the mechanism; nobody's real position is the average of both. Multiply the same logic across the market properly and you get the CFPB's headline finding — US consumers were assessed $160 billion in credit card interest in 2024, up from $105 billion in 2022. The Bureau attributes the rise to higher APRs, 9.5% more cardholders, and an 18% increase in the average balance per cardholder.
For the UK, The Money Charity runs the equivalent projection: at May 2026 debt levels and rates, twelve months of interest on UK personal debt would total £89.4 billion, an average of £3,077 per household — about £256 a month. That figure includes mortgage interest, and it is a forward projection from a single month's data rather than an observed outturn, so treat it as an annualised estimate.
Nothing here is a reason to panic about the total. It is a reason to care intensely about which line the balance sits on. If you want to see the effect of an extra £50 a month against a specific balance and rate, the credit card payoff calculator does the month-by-month version, and compound interest, explained covers why the same mechanism that builds savings also builds a card balance.
The minimum payment problem
This is the single most expensive default setting in consumer finance. The Money Charity models a credit card at the average interest rate, repaid at the legal minimum — interest plus 1% of the balance — and finds it takes 27 years and 10 months to clear. On the £2,751 average household card balance, the first month's minimum is £79. Fix the payment at that same £79 rather than letting it shrink each month and the debt clears in 4 years and 11 months.
Same balance. Same rate. Same first payment. Twenty-two years and eleven months of difference, entirely because minimum payments fall as the balance does.
It is not a rare behaviour. The CFPB found about 15% of US general-purpose cardholders and 20% of store card holders paid only the minimum in 2024, the highest share since at least 2015. Roughly half of all US card accounts revolve a balance. In the UK, the FCA's Financial Lives 2024 survey found 19% of adults — 10.1 million people — had revolved a card balance, and 5% (2.8 million) were in persistent debt, paying more in interest, fees and charges than they paid off the balance.
Who is actually in trouble
Averages hide the distribution. Three sourced measures of the tail:
Average debt of people entering UK debt advice
Mean arrears and unsecured debt per new StepChange client
Source: StepChange Debt Charity, Statistics Yearbook 2025, 163,916 clients advised in 2025
Credit cards are the most common debt in that population: two in three StepChange clients had card debt in 2025, averaging £8,021, while personal loans were held by 48% and carry the highest average balance at £9,789. Encouragingly, the share of clients in a negative budget — spending more than they earn even after advice — fell from 30% in 2024 to 28% in 2025.
Two other markers. The FCA found 13% of UK adults (7.3 million) described keeping up with bills and credit commitments as a heavy burden in May 2024, down sharply from the 21% peak recorded in January 2023 — though that 2023 reading came from a smaller cost-of-living recontact survey of 5,286 people rather than the full ~18,000-person wave, so the size of the fall is less precise than it looks. It remains above the 11% recorded in both 2017 and 2020. A separate FCA measure put 24% of adults on low financial resilience, statistically unchanged since 2022. And Registry Trust recorded 996,261 new consumer County Court Judgments in England and Wales in 2025, up 11.8% on 2024 — with 39% of the roughly 4.1 million unsatisfied consumer judgments on the register for amounts below £500.
That last detail matters more than the headline count. Most of these are small debts that escalated, not large ones that collapsed. If you are anywhere near that territory, the UK debt solutions guide covers what free regulated advice can actually do, including Breathing Space.
What this means for your budget
Four things follow from the data above, none of which require you to care about national aggregates.
Rank your debts by rate, not by size or by how much they annoy you. The July 2026 spread — 34.55% on an overdraft against 6.85% on a £10,000 loan — means the same £1,000 of repayment does five times as much work depending on where you point it. That is the arithmetic case for the avalanche method, laid out against the snowball in debt snowball vs avalanche.
Treat a persistent overdraft as the emergency. UK overdraft pricing has been a single flat APR since the FCA's 2020 reforms, and it is now the most expensive mainstream credit there is. The FCA found 8% of UK adults — 4.3 million — are constantly or usually overdrawn by the time they next get paid. That is not a debt problem you can outrun with a better spending month; it is a structural gap between income timing and outgoings, and it needs a written monthly budget to fix.
Fix your payment amount instead of accepting the minimum. The 27-years-to-4-years finding is the highest-return action in this entire report and it costs nothing. Set a standing order for a fixed amount and leave it there as the balance falls.
Keep a small buffer even while repaying. Paying every spare pound at a 24.71% card is mathematically optimal right up until the boiler goes, at which point the repair lands back on the card. That is not a rare scenario: 12% of US adults told the Fed they could not cover a $400 expense by any means at all, and the surveys measuring who has a buffer disagree by nine points on the same question. The emergency fund guidance covers the sizing; the practical version is that a modest buffer first, then aggressive repayment, usually beats pure optimisation. The debt-free guide sequences the two.
Limitations
We have said most of this already, but it belongs in one place.
These are national and aggregate statistics. None of them describe your household, and several describe populations — debt advice clients, cardholders, adults — that differ from each other. Quoted rates are advertised averages, not what any individual pays; effective rates on existing balances differ, and 0% promotional balances sit outside the quoted series entirely. The illustrative annual-cost figure holds a balance flat for a year, which nobody does, and therefore overstates the cost of a balance you are actively repaying while understating the cost of one that grows.
The UK and US figures are not directly comparable. Only the OECD table is built for cross-country comparison, and even there the ratio reflects mortgage market structure as much as household behaviour. The US totals here exclude mortgages, because the G.19 release does. The interest rates are not comparable across the two countries either — the UK series are advertised rates, the US ones are rates actually charged or stated on live accounts, which is why we never place a UK card rate beside a US one and call one higher. And the CFPB data is from calendar 2024 — the most recent published, but roughly eighteen months behind the UK rate figures.
Finally, several figures are calculations by their publishers rather than observed outturns: The Money Charity's £89.4 billion annual interest projection and its 27-year minimum-repayment model are both explicitly modelled. We have flagged each one where it appears.
If a number in this report made you want to check your own, that is the right instinct. iBudget tracks what you actually owe and what you actually pay against it, across both partners in a household if that applies. The budget calculator is free and needs no account.
About iBudget
iBudget helps couples and families take control of their finances with simple, collaborative budgeting tools. Track spending, set goals, and build wealth together.
Start Your Budget
