The Budget: when saving less really means borrowing

Alongside last month's Budget, the Office for Budget Responsibility released its forecasts for the UK economy. It expects the household saving rate, excluding pensions, to fall from around 5½ per cent in 2025 to just over ½ per cent by 2031. It is one of the most striking numbers in the forecast that underpins the Budget, and it should concern anyone who works with households in debt.

Household incomes are being squeezed by higher energy prices, higher interest rates and frozen tax thresholds, and are expected to recover only slowly. Yet over the next five years the OBR expects real household consumption to rise by around 9 per cent, three times as fast as real household disposable income. In the OBR's words, households will "gradually reduce saving to smooth their consumption over time".

In reality, many households have little or no savings to draw upon. Nearly a third of adults (31 per cent) have less than £1,000 in cash savings, and one in ten have none at all (FCA, Financial Lives 2024).

It is not that the OBR is ignorant of this fact. Its own updated small economic model assumes that around 70 per cent of households are 'liquidity-constrained', spending what they earned in the previous period, with no savings to fall back on. So why is it forecasting a reduction in the saving rate as the means of sustaining consumption?

Saving less really means borrowing more

The saving rate is not just a measure of money put by for a rainy day. It is the share of disposable income that households do not spend, and there is a difference between those things. What matters is how the national accounts treat debt. Interest is deducted before disposable income is counted, so it never appears as saving. Repayments of principal are not spending, so they count as saving: a household paying down its mortgage or credit card is 'saving', even if it has nothing in the bank.

A falling saving rate can therefore mean households are running down savings, repaying less, borrowing more or paying more interest. The OBR's forecast points to borrowing. It projects that households will move from lending around 1¾ per cent of GDP to the rest of the economy in 2026 to borrowing around 1¼ per cent by 2031, around £45 billion in that year alone. It also projects household debt, excluding student loans, rising by around £270 billion by the end of 2031, mostly secured on property.

Why it matters for growth

Every pound borrowed to keep spending up carries interest, and that interest comes out of the income households need for everything else. Household consumption accounts for around three-fifths of the economy, so the growth forecast depends heavily on households continuing to spend. If spending is to grow three times as fast as income for five years, the gap has to be filled from somewhere. Some of it may come from better-off households running down their savings. But with most households having little to draw on much of it is going to be borrowed.

Growth financed in this way is only as durable as households' capacity to carry the debt that pays for it. If that capacity runs out, spending falls back, and the growth goes with it. That is what happened after 2007: Bank of England research suggests that spending cuts by highly indebted households may have reduced aggregate consumption by up to 2 per cent.

Measuring household debt burdens

The OBR forecast includes a measure of household interest payments as a share of income, and on that basis the outlook appears manageable: for example, mortgage payments reach 4¼ per cent of household income by 2031, the highest since 2009 but not, on its own, a cause for alarm.

However, income-based measures take no account of the other things that income has to pay for. That is not a technicality. It is the difference between a measure that reflects a household's real ability to keep up its repayments and one that does not.

Our approach

In 2016, our Britain in the Red report for the TUC proposed a different approach: measuring household interest payments not against income, but against the surplus left once households have paid for their consumption. We developed this further in a 2020 paper for the German Institute for Economic Research (DIW Berlin), showing debt burdens rising sharply between 2015 and 2017 while the conventional measures continued to fall.

While the OBR measure looks at mortgage payments relative to income, ours asks whether households can sustain their current spending while meeting the interest on all of their debts, including consumer credit. Built from the same national accounts as the OBR's forecast, it can be applied directly to it, to answer the question the Budget leaves unanswered.

On the OBR's central assumptions, our measure rises from 44 per cent at the end of 2025 to around 81 per cent by 2031. That would take it to the top of its post-crisis range of 58 to 85 percent, but below the 100 per cent line at which interest absorbs the whole cash surplus and the 123 per cent reached before the run on Northern Rock. Taken at face value, the forecast describes a return to the post-crisis norm rather than a new crisis.

But that reading depends on the central forecast proving right. By 2031 households are expected to have so little leftover after consumption that modest shocks have a large effect. Any one of the following would, on its own, take the measure past 100 per cent:

• A further squeeze on real incomes of around 1.2 per cent of consumption (about £26 billion a year), only slightly larger than in the OBR's own severe energy scenario;

• A rise of just over one percentage point in the average interest rate on household debt, close to the rise in that same scenario;

• An additional £22 billion a year of borrowing to fund spending.

In combination, the effects compound. On our estimates, the OBR's severe energy scenario would take the measure to between 99 and 112 per cent, depending on how far households protect their spending; with an extra £10 billion a year of borrowing, between 108 and 124 per cent, back to pre-crisis levels. That is not implausible. Net consumer credit lending rose from £7.8 billion to £13.5 billion in the three years to August 2026; another £10 billion would take it to levels last seen in the early 2000s.

What needs to change now

The Budget forecast asks households to carry the adjustment to a lower-income economy by saving less and, in practice, borrowing more. That is a choice about the kind of growth the country pursues. Growth built on borrowing that households cannot afford does not last. It ends, as it did after 2007, in spending cuts, arrears and losses that the whole economy pays for.

Our recent report, In the Balance, found that 44% of credit-holding low-to-middle income households (around 3.5 million) are unable to meet the cost of their credit repayments and essentials. These are already cutting back on food and fuel, and often also falling into arrears with household bills. Around 1.15 million lower-income households were also lent more, or lent to for the first time, while already unable to meet their essential costs, mostly through mainstream credit cards, overdrafts and personal loans. A forecast that depends on more borrowing, under lending rules that already fail to prevent this, depends on more of the same harm.

Three changes would help immediately:

• The FCA should strengthen its affordability rules, requiring lenders to test explicitly for an established pattern of financial difficulty before extending credit;

• The OBR should publish household saving, consumption and interest payments under each of its alternative scenarios, and

• The OBR and the Office for National Statistics should report household interest payments against the household surplus, alongside the saving ratio.

A growth strategy that relies on households borrowing to sustain their living standards needs lending rules that protect their capacity to do so. We call on the FCA to act on irresponsible lending now, and on the OBR and the ONS to start measuring the burden the forecast relies on.