Public finances – Critical Finance

Rob Calvert Jump and Jo Michell

Just over a year ago, ahead of Jeremy Hunt’s first Autumn Statement, we published a report on how UK public finances are managed and discussed. At the time, the media was awash with claims of a ‘black hole’ in the public finances. We pointed out that this was incoherent because the so-called black hole was nothing more than the difference between an arbitrary fiscal rule and an uncertain forecast.

We also pointed out that forecasts of public debt are highly sensitive to the assumed path of variables such as nominal GDP growth. The latest Autumn Statement and its accompanying OBR projections provide a case in point.

In the run up to the Statement, the media focus had switched from black holes to ‘fiscal headroom‘. This was reported to be around £20bn, and is a measure of the size of the fall in the public debt in the fifth year of the OBR’s forecast – a ‘black hole’ with the sign reversed.

The figure below shows the last three forecasts of public debt from the OBR, along with the historical data published at the time of the forecast. Since the Autumn Statement a year ago, the level of public debt in the final year of the forecasts has fallen from 97.3% of GDP in the November 2022 forecast, to 92.8% of GDP in the most recent one. This forecast revision amounts to nearly £150bn of 2028-29 GDP.

Is this £150bn improvement in the forecast the result of policy actions taken by the government? And if so, why is all the talk of £20bn of headroom, rather than £150bn? The answer to the first question is a firm ‘no’. These shifts in the forecast have nothing to do with policy, and are driven entirely by data revisions and changes in the OBR’s macroeconomic forecasts.

Recent revisions to GDP have shown a stronger recovery from the pandemic than previously thought, alongside higher than expected inflation. As a result, nominal GDP is substantially higher than it was a year ago, and so the debt to GDP ratio is lower.

In the space of a year, data revisions and revised expectations about inflation have, therefore, swamped any change in the debt level driven by policy. Shifts in the positions of the series are far greater than the marginal change between the final two data points – the so-called ‘headroom’ which receives so much attention.

What about the second question? Why are we not talking about £150bn of headroom? The answer to this is unclear, but it is probably the case that the Chancellor of the Exchequer is not actively considering every possible trajectory for the public debt that satisfies his fiscal rules. If he did so, he could substantially increase public spending in a manner that would leave debt falling, as a percentage of GDP, by the end of the OBR’s forecast period.

Consider, for example, the OBR’s November 2023 forecasts for the main fiscal aggregates, displayed in the table below. Public sector net debt, as a percentage of GDP, falls from 93.2% in 2027-28 to 92.8% in 2028-29. Public sector net borrowing is less than 3% of GDP in 2028-29. As a result, the government’s two fiscal targets are met.

year	gdp	gdp_centred	psnd	psnb	gilt_rate	psnd_pct	psnb_pct
2022-23	2552	2650	2251	128.3	3.13	84.94	5.03
2023-24	2726	2761	2458	123.9	4.5	89.03	4.55
2024-25	2798	2841	2603	84.6	4.52	91.62	3.02
2025-26	2887	2938	2724	76.8	4.55	92.72	2.66
2026-27	2995	3051	2845	68.4	4.62	93.25	2.28
2027-28	3106	3162	2947	49.1	4.74	93.2	1.58
2028-29	3218	3274	3039	35.0	4.88	92.82	1.09

OBR forecasts in the November 2023 Economic and Fiscal Outlook.

Now, consider a counterfactual increase in public investment, by £10bn in 2024-25, £15bn in 2025-26, £20bn in 2026-27, £28bn in 2027-28, and then £10bn in 2028-29. This looks something like the Labour Party’s proposed green new deal, in which annual public investment would increase to £28bn over a single parliament. For simplicity we assume multipliers of zero and no impact on inflation, so that nominal GDP is unchanged from the OBR forecasts.

How would this affect public sector borrowing? In 2024-25, it would increase (relative to the OBR’s baseline forecast in table 1) by £10bn. In 2025-26 it would increase by £15bn plus interest payments on the previous year’s £10bn. In 2026-27 it would increase by a further £20bn, plus interest payments on the previous two years’ borrowing, and so on. The resulting time paths for borrowing and debt are displayed in table 2, below.

year	gdp	gdp_centred	psnd	psnb	gilt_rate	psnd_pct	psnb_pct
2022-23	2552	2650	2251	128.3	3.13	84.94	5.03
2023-24	2726	2761	2458	123.9	4.5	89.03	4.55
2024-25	2798	2841	2613	94.6	4.52	91.97	3.38
2025-26	2887	2938	2749.46	92.26	4.55	93.58	3.2
2026-27	2995	3051	2891.63	89.58	4.62	94.78	2.99
2027-28	3106	3162	3023.84	79.31	4.74	95.63	2.55
2028-29	3218	3274	3129.59	48.75	4.88	95.59	1.51

Counterfactual government spending path based on OBR forecasts.

Public sector net debt, as a percentage of GDP, now peaks at a higher level: 95.63% rather than 93.25%. But it is still falling between 2027-28 and 2028-29, and public sector net borrowing is still less than 3% of GDP in 2028-29. As a result, the government’s fiscal targets are still met.

This counterfactual trajectory of debt-to-GDP, alongside the OBR’s recent forecasts are plotted in the figure below. Our counterfactual trajectory is not dissimilar to the OBR’s forecast from March 2023. The end of forecast debt/GDP is around two percentage points of GDP lower than in the November 22 forecast, in which Hunt had defeated the black hole and met his fiscal rules. Given that this was regarded as a success only a year ago, on what basis could our counterfactual trajectory be rejected?

It is clear that ‘headroom’ as reported in the media is not simply a measure of the amount of money that the Chancellor could spend without breaching his fiscal rules. In fact, given its complicated nature, there is no single number that summarises the amount of extra spending consistent with a headline fiscal rule defined by the rate of change of debt-to-GDP at a future point in time. It depends on the distribution of extra spending over the forecast period, as well as the time path of interest rates.

Moreover – and this is, perhaps, more important – it depends on the volatility of the forecasts themselves. The difficulties involved in forecasting economic and fiscal aggregates over a five year horizon is illustrated by pre-budget forecasting exercises published by the Institute of Fiscal Studies and the Resolution Foundation. Their estimates of public debt as a share of GDP at the end of the forecast period differ by nearly 10 percentage points – over £300bn. Minor adjustments to the assumptions that generate these forecasts lead to outcomes an order of magnitude greater than the ‘headroom’ which attracts so much attention.

This is not a rational basis on which to conduct the planning of long-term spending and taxation. It is clear that Hunt’s budget is an exercise in gaming the system. Current nominal tax cuts are ‘paid for’ by creating ‘headroom’ which results from imprecisely specified cuts to government spending towards the end of the forecast period. Moreover, the widely-quoted ‘headroom’ figures have no correspondence whatsoever to the amount of extra money the Chancellor could spend while meeting his rules, and any policy effects are swamped by revisions to the data and forecasts.

As Paul Johnson, head of the Institute for Fiscal studies says, “Aiming for debt to fall in a particular year is not a good fiscal rule”. Simon Wren-Lewis puts it even more bluntly: “falling debt to GDP is a silly rule”.

We are entering fiscal silly season. As the budget approaches, we should brace for impact with breathless reporting of context-free statistics about inflation, interest rates and government debt.

The story is likely to go something like this. Inflation is rising. This raises costs on government debt because some of it (index-linked bonds) pays an interest rate linked to inflation. Costs associated with quantitative easing (QE) will also increase because QE is financed by central bank reserves which pay Bank Rate (the Bank of England’s policy rate of interest). Since inflation is rising the Bank will have to raise interest rates to control it. This will increase the financing costs of QE and the cost of issuing new debt for the Treasury.

The conclusion — sometimes implied, sometimes explicit — is usually some version of “the situation is unsustainable therefore the government will have to make cuts”.

While each part of the story is technically correct in isolation, the overall narrative — debt is out of control and the situation is going to get worse because of inflation — doesn’t stand up to scrutiny.

These stories are rarely presented with sufficient context. Instead, journalists tend to rely on statistical soundbites such as “public debt is the highest since … ”. This is rarely if ever accompanied by the fact that debt/GDP is a fairly meaningless number.

The problems associated with government debt essentially boil down to the fact that debt involves redistribution. In the case of the government this means redistribution in the form of transfers from tax payers to bond holders. This is politically difficult. (This is also why “but currency issuer …” responses to these issues are largely beside the point — the problems of debt management are ultimately political not technical).

The ratio of debt to GDP tells us very little about the current political difficulties arising from debt servicing. Instead, the relevant magnitudes are total interest payments and tax revenues.

Total interest payments are equal to the debt stock multiplied by the effective interest rate on government debt. Focusing on the debt stock in isolation is thus equivalent to representing the area of a rectangle by the length of one side.

A better indicator of the risks associated with public debt is the ratio of government interest payments to tax revenues, as plotted in the figure below.

Interest payments on government debt have indeed risen recently. A spike in June triggered media articles about the highest interest payments on record. In context, such statistics are shown to be meaningless. Interest payment have risen to around 6% of taxation over a four quarter period, compared with all-time lows of about 5.3%. (Calculated on a 12 monthly basis this rises to around 6.5%). It is hard to see signs that the sky is falling.

In fact, this indicator overstates current interest costs. This is because much of the interest paid by the Treasury is paid to the Bank of England which holds a substantial chunk (currently around 37%) of UK government debt as a result of QE (see chart below). Most of this interest is returned directly to the Treasury. Since the start of QE, this has saved the Treasury over £100bn in interest costs.

Adjusting for this reduction in interest payments produces the figure below: net interest payments sum to around 4.7% of tax revenues over the last four quarters (or 5.2% on a rolling 12 monthly basis).

What of the dangers ahead? It is true that if inflation rises, then interest costs will rise, all else equal. But the scale of these rises is not predetermined, and will be affected by policy.

First, persistent inflation is far from a certainty. If if inflation does persist in the short term, the Bank does not need to raise interest rates. Hikes in response to price pressures due to pandemic reopening and supply side bottlenecks will do more harm than good — instead the Bank should wait until the economic recovery is clearly underway. In this context, interest rate increases would likely be a good sign, and would be offset by rising tax revenues. Further, the Bank could introduce a “tiered reserve” system which would serve to hold down the rate paid on a substantial proportion of outstanding debt. Short term and index-linked debt can be rolled over at longer maturities, delaying the point at which higher rates would feed into higher interest payments.

In summary, simple claims such as “a one percentage point rise in interest rates and inflation could cost the Treasury about £25bn a year” are not useful without context and explanation of the long list of assumptions required to produce such a figure. The policy conclusions derived from such claims should be taken with a large pinch of salt.

Season’s Greetings and enjoy the festive period!

Category: Public finances

Hunt versus headroom

Fiscal silly season