Image © iStock
Over recent years, there have been growing calls for the UK to alter its fiscal framework to focus more on improving the net worth of the public sector, and less on simple debt targets.
With a general election due next year, and against a backdrop of widespread concern about the parlous state of UK public finances, it is therefore timely for the Institute for Fiscal Studies to review this topic in its annual Green Budget. Chapter 8, by Ben Zaranko, does just this, by addressing the question of whether Public Sector Net Worth (PSNW) should feature as a (or more precisely, in the rather limited hypothesis examined by the author, the sole) fiscal target.
Its conclusion is that this would be a "long walk for a small sandwich"; the potential benefits, in the view of the IFS, are marginal.
This is an important topic, the tone of the chapter is balanced and thoughtful and many of its observations are very interesting. However, overall the chapter falls short in its understanding of the basic premise for using a measure of government Net Worth as a fiscal target, the details of the issues which it discusses, or the practical possibilities that this creates for better government. It is perhaps telling that the footnote next to the author’s name at the top of the chapter acknowledges comments of “several civil servants”; it is tempting to suppose that the chapter’s arguments might have been shaped by the views of officials who are aligned with the status quo that one might argue the IFS should be seeking to challenge.
Accounting is key to understanding and managing public finances, and fiscal rules should reflect this
Perhaps one reason why the IFS fails to get at the heart of the question is that it is approaching it from the wrong direction: that of top-down fiscal policy. Our premise is that government financial management as a whole should be based upon accounting principles, and that fiscal targets should follow on from these. As Net Worth is the “bottom line” measure of the health of a balance sheet, so should it feature as a key (but not sole) fiscal target.
Our starting point is the rather obvious observation that governments are the largest and most important organisations in the world. As citizens, we depend on them for protection, for provision of public services, and for the ability to intervene at times of stress – illustrated on multiple occasions over the past 15 years or so.
We also depend on their ability to anchor national financial systems, and collectively to ensure global financial stability. All this makes us highly dependent on the overall financial health of governments, on the efficiency of their management, and on their being accountable to their "stakeholders" – the electorate.
Similar, though lesser, concerns relate to private or public sector corporations, institutions, not-for-profit organisations and others, and a major way in which these concerns are met is through legislation which requires the production of timely, audited accounts. These in turn are compiled from the management accounts that are used to run the enterprises concerned. There is at all times a straight-through line from the information that informs the actions taken by managers to the financial reports of the enterprise.
The threshold conditions for needing to produce accounts relate to the scale and importance of an organisation. However, despite their much greater scale and complexity, governments around the world see accounting as at best an interesting retrospective, mainly to be ignored. Only New Zealand does what every government requires every other enterprise to do, and puts accounting at the heart of the financial management of its activities.
So when we advocate use of Net Worth in fiscal rules (a topic to which we will return), we are doing so not because of its theoretical attractions in isolation, but rather because Net Worth is the ultimate expression of the financial health of an organisation which describes and manages itself through the application of accounting principles. These will never be perfect, and there will be areas where accounting fails fully to capture the essence of what governments seek to do. But this is a very poor reason for abandoning the attempt, as we will aim to demonstrate in response to some of the points made in the Green Budget chapter.
We are also firmly of the view that an effective fiscal framework (whether or not embodied in formal “fiscal rules”) needs to use multiple measures to assess financial health and financial capacity. This too has its parallels in the private sector. For example, a debt rating agency or prospective lender will look not just at a company’s Net Worth, but also at a range of other measures on both its balance sheet and income statement in order to assess its creditworthiness. Why should governments – with their much greater complexity and importance – rely on a single target?
Intergenerational fairness – a key argument for measuring and managing Net Worth
Net Worth has one further attraction. More than any other single measure, it describes the legacy that each cohort of taxpayers passes on to the next. On a year-to-year basis that might not matter much. But over decades, or generations, it is very important, and indeed represents a crucial measure of intergenerational fairness.
This is a major reason why the currently quite strongly negative public sector net worth position of the UK is important: we have apparently lost the moral compass that might otherwise guide us to leave more for next generations than we received ourselves. At a time when populations are ageing and healthcare and pensions costs rising, that is an especially unfair position to take.
Why it is important to understand liabilities as a whole, not just debt
One of the drawbacks of the IFS analysis is that it expounds on the disadvantages of a simple Net Worth-based fiscal rule operated in isolation without pausing to consider the systematically undesirable effects of our current simple debt-based fiscal targets, which focus on two metrics: net borrowing in-year (PSNB), and the net debt (PSND)-to-GDP ratio.
Those familiar with accounting will recognise that this means making key financial decisions based upon information about a small fraction of the balance sheet. Assets are ignored in their entirety, as are non-debt liabilities. In the case of the UK, non-debt liabilities – mainly public sector pensions – exceed PSND. The rules we use to impose financial discipline address less than half of our liabilities, and in paying no attention to assets also ignore the uses to which borrowing is put. This is not only profoundly illogical; it is dangerous.
For a practical example of this problem in action, consider the National Health Service Pension Scheme (NHSPS), the largest UK public sector pension scheme, representing about 40% of total public sector pension liabilities.
In 2021-22, the impact of the NHSPS on public finances is described in three ways, according to its annual report:
Net cash effects: a £4bn benefit to the Exchequer (£17bn contributions, £13bn payouts).
In-year accruals of service benefits totalling £35bn and consisting of a £52bn in accrued benefits offset by £17bn in employer and employee contributions; an item in the Department for Health’s Supply Estimates, as approved by Parliament.
An actuarial assessment of the present value of the Scheme’s liabilities: this takes into account cash flows, market-derived discount rates provided by HM Treasury as well as estimates of mortality, inflation etc. On this measure NHSPS liabilities increased by £112bn – almost as much as that year’s PSNB of £120bn.
At the very least, points 2 and 3 suggest that very large amounts of value are being deployed outside the fiscal rules, on an ongoing basis. This means that some (very large) costs, and their long-term consequences, are not being effectively managed.
Is Net Worth too volatile to be a useful measure?
Rather than ponder on the way that current fiscal rules saw the NHSPS as a net benefit to fiscal targets whilst the underlying financial truth was very different, the IFS focuses on what it sees as the volatility of Net Worth.
In part, this is right; Net Worth will vary in ways that are outside government control. Indeed, any useful assessment of any entity’s financial condition will vary through time, including for external reasons. The fact that these variations fall outside the domain of present debt-based fiscal targets does not in itself make these targets superior; more likely the reverse is true.
But it is also important to think carefully about what exactly is meant by Net Worth, and this will depend both on the nature of the government involved, and also in the way that the measure is calculated. For example, a federal government will lack direct control over state or provincial entities, and should therefore manage to a narrower definition of its balance sheet; a unitary government such as the UK can more reasonably manage on the basis of a full public sector balance sheet. Hence our use of “Net Worth”, without defining the exact scope.
In illustrating the volatility that it is concerned about, the IFS discusses various ways of valuing debt obligations and then uses an IMF-endorsed measure of PSNW that looks at all levels of government and state owned enterprises, and uses market values of debt obligations, rather than face values.
This means that changes in nominal interest rates result in (potentially quite big) changes in the value of outstanding debt and hence PSNW. The IMF framework compounds this problem by expressing debt as a percentage of GDP. As both interest rates and GDP move with inflation, this approach has the effect of making the debt measure vary inversely with the square of the GDP deflator, which appears to account for most of the volatility that concerns the IFS.
In our view, the IMF is to be applauded for its work in highlighting the importance of public sector balance sheets, and in gathering hard-to-find data to allow comparisons to be made; the fact that there is no other easily available comparative data perhaps explains why it is employed in this chapter. But in this particular respect, we believe that its methodology is lacking.
A better approach, more consistent with IFRS accounting standards for organisations such as governments, would use face values of debt, rather than market values. This would also result in lower reported PSNW volatility, and also more accurately reflect the nature of the liabilities themselves.
Looking at other liabilities, it would of course remain true that when real interest rates – which are less volatile than nominal ones – rise, pension liability values fall, other things being equal. But at the same time, debt refinancing costs go up. So governments gain on the balance sheet swings, but lose on the income statement roundabouts. As we have noted above, an intelligent fiscal framework will take both factors into account, and avoid the Green Budget chapter’s conclusion that higher interest rates allow higher borrowing under a Net Worth-based fiscal rule.
Why we need to understand asset values
The IFS is particularly concerned about the difficulties of measuring the values of non-financial assets. We would argue that it is exaggerating the problems, and that valuable information is lost when the government fails to measure or account for the assets that it holds.
For example, experience in Europe and North America demonstrates a very large disparity between reported values of government property holdings and their actual market values. This reflects, in part, an accounting standard which allows governments to value properties (if they value them at all) on the basis of their current use or historic cost. The scale of this undervaluation is large – perhaps as much as 100% of GDP. How can taxpayers (or for that matter civil servants or politicians) have any idea whether these resources are being used in a way which best meets public needs if they don’t know what they are worth, or if these values are not incorporated in the financial management framework?
The IFS suggests that because the government is unlikely to close the rail network and sell off the underlying property there is no need to value that property. This is a false premise; the UK government has sold plenty of rail assets, often, one might think, unwisely. But that is not an argument for not valuing the assets; it might be that understanding asset values will expose opportunities for their more efficient or effective use, as has been illustrated in other countries. And new infrastructure investment comes at a known cost that should be recorded in government accounts, and written down where it exceeds the value of what has been produced.
Finally, accounting properly for assets matters not just because it tells us something about public sector wealth, but also because it gives us a measure of the rate at which those assets are being consumed. For the long term, this is as important to know as the amount of cash that is being spent; if we are running down assets to save money today, we are storing up costs for the future. This is a key feature of accrual-based accounting, and one that is embedded in accounting-based measures of Net Worth.
Do Net Worth-based fiscal rules “ignore the biggest state assets and liabilities”?
The IFS, using language very redolent of HM Treasury orthodoxy, suggests that a “fundamental critique” of using Net Worth-based fiscal rules is that it ignores the state’s biggest asset – its ability to tax – and its most important liabilities – the implicit promise to provide public services.
This is a very odd position to take. Taken literally, the IFS appears to be arguing that the present values of these future in- and out-flows are so much greater in scale than the current balance sheet that there isn’t much point looking at where we stand today, and what matters is where we are going.
We agree that the “intertemporal balance sheet” is very important. This is the sum of two things: the present balance sheet, and (the present value of) projections of future revenues and costs. Certainly, targets for shorter-term government finances – for example, embedded in fiscal rules – need to reflect the long-term outlook. But they also need to take into account a full picture of assets and liabilities. This is the essence of what we propose: that governments use accounting tools fully to describe their financial position, and fiscal rules that ensure that this tracks the course necessary to meet long term constraints.
Failing to measure the balance sheet, and managing through debt-based fiscal rules, as the IFS seems to suggest we should, will neither tell us where we are beginning the journey, nor what direction we need to go in.
How embedding accounting in financial management and fiscal rules can help increase government revenues
There is a further potentially very important set of benefits that accounting can unlock when embedded in fiscal rules, which is not covered in the chapter - perhaps because it sits well outside the current HM Treasury fiscal framework. This relates to the way that assets and liabilities are managed.
First, assets. Under current arrangements, as we have seen, little attention is paid to the value of the assets that government holds, even where these have easily realisable alternative uses, as is the case for much of the government’s property portfolio. Even if the values of these assets are known, and there is no systematic process for ensuring that that value is disclosed, and that managers are seeking to maximise the value of the property portfolio for the public good.
The IMF describes property holdings as “public commercial assets” and estimates that better management of such assets can typically yield revenues to government of around 1.5% of asset value. On the premise that assets whose value is not disclosed or maximised are likely to be undermanaged, our estimates that property is undervalued by amounts equivalent to up to 100% of GDP suggests potential revenue gains of 1-1.5% of GDP are available through better management of the public sector property portfolio.
Second, liabilities. We have seen how a debt-based fiscal rule encourages the incurrence of non-debt public sector pension liabilities as these are ignored by such a rule. But these will need to be paid out just as surely - with the added kicker that they are inflation-linked and so unlike much government debt can’t be inflated away.
A policy of borrowing the amount needed to fund these liabilities and investing in a diversified global portfolio could reasonably be expected to earn an annual return of 3% above the government’s cost of borrowing over time. For a country with pension liabilities equivalent to 100% of GDP, such as the UK, this in turn would increase government revenues by the equivalent of 3% of GDP, when fully implemented. In addition, the balance sheet would benefit from a large diversified portfolio which would greatly improve national resilience in the face of economic or financial stress.
Because the liabilities that are being funded already exist, the only incremental risk associated with this strategy would be the investment risk – and financial theory, supported by centuries of data, suggests that this is negligible for a programme in which government borrowing and diversified investing takes place on a continuous basis.
Of course, it would be better to have started by investing as liabilities were first incurred, but we have not done that, presumably because of our not-fit-for-purpose fiscal framework. But it still makes sense to start doing it, and present pension liabilities could reasonably be matched by assets well within a generation. Moreover, by providing a truer indication of the cost of these obligations and embedding this in the fiscal framework, it is possible that this discipline would encourage the adoption of less expensive public sector pension arrangements.
How long a walk, and how big a sandwich?
To get full value for the accounting-driven framework that we propose would no doubt require considerable evolution of existing, non-accounting driven financial management systems across government. But this should not be impossible, indeed many of the elements are already in place as demonstrated by the UK’s 10-year-plus record of publishing Whole of Government Accounts.
To continue the walking metaphor, there are also numerous shortcuts available, including the much more up-to-date aggregate statistics produced by the ONS, which (with some amendments better to reflect accounting standards and asset values) could form the basis for measuring performance versus Net Worth-based rules before the full framework was in place.
The very substantial opportunities for improving finances through better management can also be exploited from the outset; it is only the existing rules that are holding these back. By better management of assets and liabilities, we estimate that there is scope for the UK government to increase government revenues by 4%+ of GDP - very close to what the OBR estimates is needed to sustain existing levels of public services as the population ages.
Overall, the walk is shorter, and the sandwich a lot bigger, than the IFS believes.
Look out for a feature in the January/February issue of PF in which John Crompton and Ben Zaranko will set out the arguments for or against.
