UK Budget Preview #4: This is not normal

As we approach the day of the UK Budget, it’s worth reflecting on just how unusual the process so far has been and what a better one might look like.

Speculation about the content of the Budget is nothing new, and has provided endless fodder for newspaper and podcasts alike – and, let’s face it, our blog too.

But the spectacle of a Chancellor delivering a speech on 4 November specifically aimed at softening up the public at large for bad news – or ‘setting the scene’, as the official Treasury communications put it – was extraordinary in its own right. This is usually the first ten minutes of the Budget statement to the Commons minus the official OBR figures – not just another Tuesday inside number 9 Downing Street’s briefing room.

The Grand Old Duke of York, Treasury edition

It might not feature as a new entry in the Roud Folk Song Index, but Rachel Reeves really did seem to be taking UK taxpayers on a march of the income tax rise hill. The ‘Scene Setter’ speech all but confirmed it would happen, with the Chancellor saying “each of us must do our bit.”

Nine days later, however, the FT reported that was no longer the case. ‘Better than expected OBR forecasts’ was the explanation for this – and the Chancellor had harnessed this opportunity to ditch what would have been a straight-up break of the Labour manifesto pledges on tax.

Call us sceptic, but we think it’s implausible that the decision was made on the basis of said forecasts. Not least because while it’s true that the Chancellor had received a new round of forecasts on Monday, 10 November, those only incorporated the effects of her proposed measures. The last pre-measures forecast with new economic and fiscal data, including market determinants and tax receipts, had been sent to the Chancellor on 31 October.

The Chancellor herself had gone on BBC 5 Live on 10 November to say that not breaking manifesto pledges would require “things like deep cuts in capital spending.” So when the metaphorical 10,000 troops were marched down the hill, there were only two plausible explanations: either the plan really was to back down all along – a possibility, though one we are less convinced by as time goes on – or the scrambled retreat was forced by what was politically palatable with backbenchers, especially given the tumult of the 48 hours that preceded it.

Whatever the reason for the U-turn, this is not a good process

U-turn might be an inaccurate turn of phrase – more like violently turning the wheel completely the other way while still stationary.

One of the now accepted truism from the Liz Truss and Kwasi Kwarteng debacle of 2022 is that the surprise of bond markets at the package and the lack of OBR assessment was a major part of why things went so badly so rapidly. The consensus inside the Treasury now appears to be that any major policy change requires extensive ‘pitch rolling’, so that holders of UK government bonds aren’t spooked by decisions.

It is indisputable that market participants at that time were shocked by the size of the package, and that volatility was increased due the OBR being sidelined. But to learn this particular lesson from that sorry episode is to miss a much bigger point – that market participants dumped UK debt wholesale because of the policy itself, not its presentation or anticipation.

The UK essentially went from being committed to bringing the deficit down relatively quickly to increasing it perpetually by tens of billions of pounds and more than doubling it by the end of the forecast horizon. Debt went from being on a gentle downward path to increasing every year from a large starting point already. However anticipated, this sort of expansionary policy from its starting point was always likely to lead to significantly higher debt interest costs. As much as you announce that you’re running into a brick wall, it will still hurt when your skull crashes directly into it.

But learning this lesson has taken the forecast process from something that happens behind the scenes to being well publicised and choreographed. The Chancellor issues a public letter commissioning the forecast and the OBR publishes timetables of each of the rounds. This sets off three months of constant speculation, with journalists – quite understandably – keen to understand what is likely to happen, and government sources using the media to brief out policies well in advance of Budget Day. In 2013, news articles said George Osbone had been embarrassed by the Evening Standard leaking his beer duty cut 30 minutes before his speech; these days, it would be a wonder to see a major policy that hadn’t been leaked 30 days in advance.

The asymmetry of the OBR’s place in the forecast process

Once the date of the Budget or Spring Statement is known to the public, the OBR essentially becomes unable to comment on forecasts. This is completely understandable: doing so would be market-moving, and reveal privileged information.

But the OBR produces documents for the Treasury throughout, and on top of that, has lots of interactions with officials across government. The OBR doesn’t actually hold the models that are used, for example, to produce income tax forecasts, as it doesn’t have powers to access confidential taxpayer data. That’s done by HMRC – on the basis of OBR assumptions, but it still means officials in government and ministers have access to a lot of privileged information.

The OBR also doesn’t cost policies, despite what is usually reported. Its role is to certify the costings done by government officials as central and reasonable, and to estimate any broader second-round effects they might have on the economy if they are large enough.

The OBR has since it was created been a very leak-proof institution. This is unsurprising – its leadership is not made of politicians whose job is to sell the forecasts to the public or the markets, and therefore it has much less incentive to leak. Its staff are civil servants and its leadership public appointees; in each case, confirmed leaking would carry severe penalties that do not apply to minister or political appointees.

But this means that the OBR ends up in a very uncomfortable position. The Treasury and Downing Street have access to confidential information, which they can and do choose to selectively leak to the press. This is unverifiable and the OBR has no right of reply, and therefore any blame or reason can be laid at the OBR’s feet, in the knowledge that there will be no countervailing statement. This is a perverse set of incentives, and one which undermines trust in the process and in institutions.

Time to (re)think?

The OBR and the forecast process in its broad current guise were first set up in 2010, as part of the Coalition Government’s reforms. A small number of staff were moved from the Treasury and given the short time frame of a few weeks between the Government taking office and its first Budget, the set-up made sense: the OBR would run an already existing macroeconomic model and send the figures to the Treasury, which would then decide on its policy and cost it. The OBR would then assess the impacts of those policies and how reasonable the costings were, before putting it all together into a set of tables that constituted the final forecast.

This may have been fine for a small organisation navigating a fast-moving situation, but – as with many other parts of the UK’s system of government – it relies on slightly murky interactions. It’s not immediately obvious that the OBR should have as much interactions as it does with the Treasury – there has always been a risk of Treasury officials presenting data and arguments in a more convincing but slightly convenient way, and it puts the OBR in the position of trying to sniff out what is the full truth from what officials told them.

The OBR’s embedded place into the forecast process is actually very unusual among independent fiscal institutions (IFIs). Many do not produce economic forecasts and scenarios themselves, but rather issue assessments of the risks and the reasonableness of their Finance Ministry’s forecasts, often weeks after budget bills are introduced.

That would obviously be a major weakening of the OBR’s standing, and one that would probably not be acceptable – and to be clear, not one we’d feel positive about. The OBR has carved out an important voice in cutting through the Government’s spin by presenting its assessment as a counterweight to the Chancellor’s speech.

But there are other models that could provide some inspiration for the future of the OBR and that shouldn’t be dismissed out of hand. For example, the CPB – the OBR’s sister body in the Netherlands – is much larger and does both forecasts and costings of government policies themselves – among other roles such as manifesto costings. The CPB is the official forecaster that the Dutch Government has to use, but it is much more separate from policymaking than in the UK, while still maintaining a publication and voice on Budget Day.

The Congressional Budget Office in the US has a different role, providing forecasts and assessments of the costs of policies but not forming part of the official Budget presented to Congress – which is itself much less structured as a budget process than in the UK, often relying on continuing resolutions and one-off bills.

Even in the UK, there are different roles for forecasters – for example, the Scottish Fiscal Commission runs an income tax model themselves, while also costing Scottish Government policies.

These are but three examples, but they illustrate that the OBR is not the only model that exists for running an IFI. Its timeliness is a great asset to fiscal transparency in the UK, but its reliance on Treasury and HMRC models is a significant weakness, as is the fact that government sources can and do brief out its purported forecasts without a right of reply.

What would a better process look like?

These two issues – lack of a voice in the briefing wars and reliance on Government departments for its modelling capabilities – should be a priority to solve going forward.

Taking the latter first, it is undeniable that the OBR’s position of not having in-house expertise undermines its independence. The OBR’s staff has increased in size over the years, but this has not been used to change the way it does its forecasting and policy costings, but rather to expand analysis in areas that are of interest to the Treasury such as second-round effects.

Of course this would come at a cost, but the OBR is a tiny organisation in the grand scheme of the UK Government: it employs around 50 people and costs less than £6 million a year to run. Even if it increased to the size of the CPB (200 people) it probably would cost less than £25 million – a drop in the ocean of the £1.1 trillion (£1,100,000 million) the UK Government spends every year, and a small price to pay for more fiscal credibility.

As for the issue of how briefing undermines and pins the blame on the OBR – this is a harder question, but one where there might be some radical ideas. For example, what if the OBR produced and published its pre-measures Budget forecast on a set day, and the Chancellor had, say, three weeks to prepare the Budget and then present it to the Commons?

This would put an end to the possibility of blaming the OBR about movements in the forecast that not only are not public at the time, but are never made public. The Treasury’s briefing and pitch rolling is based on interim forecasts that never see the light of day, and are therefore impossible to verify.

This would give the public transparency about what the actual economic situation faced by the Chancellor is. It would also provide a more level-playing field to the Opposition, which would have the option to thoughtfully decide and present how they would react to the economy, rather than awkwardly fumbling through their impossible job of reacting to the Chancellor based on no information or time to digest the announcements.

Of course there are complications – for example, what would happen to policy costings? One option might be to allow the Government to present them, and let the public and the markets judge how credible they are. It would also shed some light on the weird games the Treasury plays to add up the numbers. And it would stop the Treasury’s attempts to perennially convince the OBR of the second-round effects of policies. If the Government is really willing to bank that they will come to fruition, then it should pursue those policies – and the OBR could reflect in the next forecast whether that is really the case or not in their view and, in many cases, with the data to back it up.

Of course some of these are big changes to the way the OBR and the Treasury operate. But what’s been happening is not normal. The current system isn’t working, and the circus of the last three months is proof of that. 15 years down the line, it’s time to improve the process and increase public confidence.