What are the different ways to measure public debt?
How is public debt measured and what does this tell us about fiscal sustainability?
Politicians and commentators often refer to government or ‘public’ debt. It is important to understand how indebted a government is for many reasons. Voters need to be given a sense of how far the current government is offering today’s population benefits but deferring the costs to future taxpayers. Investors need to be convinced that the state can afford to repay any money it loans.
Ever since Gordon Brown adopted the UK’s first formal set of fiscal rules in 1997, the UK government has imposed on itself some kind of debt limit. But there are many different measures of indebtedness, covering many different types of assets and liabilities. This explainer sets out the most common measures and gives a brief summary of the pros and cons of each. Another explainer sets out how these different measures are forecast to evolve over the next five years.
Public sector net debt (PSND)
This is the measure of debt that has been most widely adopted by chancellors as the target of UK fiscal rules since 1997. It includes all debt held by the public sector, most of which is held by central government. It also nets off the value of some assets, but only those classified as ‘liquid’. This includes cash and foreign exchange reserves, but excludes most assets, including bonds and other financial assets.#
Another way to understand PSND is the cumulative difference between public sector cash receipts (mostly taxes) and cash outlays (such as spending on public services) over time. This is financed through the issuing of gilts (government debt contracts). 4 In other words, PSND is the stock equivalent of public sector net cash requirement.
Public sector net debt excluding the Bank of England (PSND ex BoE)
This was the measure of debt targeted in fiscal rules between 2022 and 2024. It follows the same methodology as PSND but excludes the Bank of England’s balance sheet. As a result, the level and time profile of PSND ex BoE differs from PSND. There are two main reasons for this.
First, the former excludes the effects of the Term Funding Scheme, which were loans provided by the Bank to commercial banks from 2016 onwards to ensure cuts to the base rate were passed through to lending rates. This scheme increased PSND when loans were issued but PSND fell again by a similar amount as the loans were repaid. The TFS, and the effect it has on PSND, was the primary rationale for adopting PSND ex BoE: repayments were tending to reduce PSND in 2022, even though this did not reflect an inherently improved fiscal position for the government.
Second, the effects of unwinding Quantitative Easing (QE) are different for PSND and PSND ex BoE. Under QE the Bank of England purchased gilts: the difference between the price at which the gilts were purchased and the price they are sold at as QE unwinds represents either a profit or loss to the public sector. The OBR forecasts that the public sector will ultimately make a loss on these transactions because the price of gilts has generally fallen since the Bank purchased them. The Treasury indemnifies all losses arising from QE – in other words, it transfers funds equivalent to the loss made on QE to the Bank. This payment is made at the point when the Bank sells the gilt, and so the loss is realised.
This payment increases PSND ex BoE by the full amount of the transfer at the point when the gilt is sold.
However gilts purchased under QE are valued differently in PSND: at the point they were purchased, the gilts were revalued at their ‘redemption price’ (that is, the amount the Treasury commits to pay the owner of the gilt when it matures). The redemption price was in all cases lower than the purchase price, and so PSND increased when the gilts were purchased. That means PSND is currently at a higher level than PSND ex BoE but will increase less as QE is unwound because some of the ‘loss’ associated with QE has already been accounted for in PSND.
The effects of QE are especially large, leading to a £128bn gap between PSND ex BoE and PSND in March 2023 – a gap that was forecast to fall to £58bn by 2028/29.
General Government Gross Debt (GGGD)
General government debt is a narrower measure than PSND, including only the liabilities held by central and local governments, thus excluding the wider public sector. This means this measure excludes, for example, the value of loans made by public sector investment banks, such as the UK Infrastructure Bank and the new National Wealth Fund. It also does not net off any assets.
This measure is used more widely in other countries. It is, for instance, the target used for the EU’s fiscal rules, which require that member states’ public finances be on a path towards GGGD being below 60% of GDP.
However, UK governments have tended to account for debt across the whole public sector as much debt held by public corporations is effectively (even if not legally) underwritten by central government.
Public Sector Net Financial Liabilities (PSNFL)
This measure is also referred to a ‘net financial debt’ by the government. PSNFL is a broad measure of indebtedness and is the measure currently used by the government in its ‘investment rule’. Compared with the previous, widely-used debt indicator, public sector net debt (PSND), PSNFL includes some additional future liabilities that the government has committed itself to, but also nets off more assets where the government would expect a future financial return from previous spending.
The main additional liability accounted for is funded pension schemes (such as local government schemes), while the assets held by these schemes are also counted on the other side of the balance sheet.
The most important other assets included in this measure that are not covered in PSND are outstanding loans (including student loans) and equity stakes in private companies. Relative to PSND, the additional assets accounted for are greater than the additional liabilities, so PSNFL is lower than PSND.
An alternative way to understand PSNFL is as the stock version of public sector net borrowing, the government’s main borrowing measure.
Public Sector Net Worth (PSNW)
Public Sector Net Worth is even more comprehensive, taking account of more assets and liabilities than PSNFL.
The main additional assets included are non-financial, including the value of public sector land and buildings. The most important additional liabilities are unfunded public sector pension schemes.
Valuing non-financial assets is more difficult than valuing financial assets, especially as many public sector assets will not have an easily comparable market price – for example, land owned by Ministry of Defence is not often bought or sold. The ONS uses standard accounting practices to calculate these values, meaning that the value of an asset depends on the amount spent to produce it and assumed depreciation. As a result, these may not reflect true economic value: a road that was built less efficiently (and so more expensively) than an equivalent efficient project would be accounted for as a more valuable asset.
Overall public sector liabilities are greater than assets, so PSNW is negative.
What do these measures tell us about fiscal sustainability?
All these measures are relevant for understanding how sustainable the public finances are. PSND, the measure most commonly used previously in public discourse about the public finances in the UK, captures the amount the government has borrowed from financial markets, and as such changes in this measure represent demand for additional finance. However, PSNFL (the measure chosen by the new Labour government) and PSNW account for the fact that not all debt is equal. Loaning money to a student that you expect will be repaid over time is different to spending on teachers’ salaries. Likewise, spending on assets, like roads and hospitals, that will be used in the future is different to spending on current consumption like salaries and grants.
In the past, the UK government has ‘gamed’ fiscal rules by selling assets like the student loan book, which reduces PSND because financial assets are not netted off that measure. PSNFL and PSNW would not be affected by selling an asset for what it is worth, which should have no effect on fiscal sustainability.
Fiscal policy should be made with all these factors in mind: if PSNW is increasing, but only via non-liquid assets that cannot be sold, this might overstate how secure the finances are. But failing to differentiate between debt accrued for current consumption and debt accrued in exchange for a physical or long-term financial asset will also give a misleading view.
UK fiscal rules have used PSND (or PSND ex BoE) as a target rather than PSNFL and PSNW previously. In part, this reflected the difficulty of accounting for some assets and liabilities – in particular the current value of pension liabilities, which can change hugely depending on the discount rate applied, which is largely out of the government’s control and can change rapidly. The ONS has a consistent series for PSNFL from 1998 and has revised its methodology in recent years.