This issue has flared up again following the row about the Conservatives’ document that claimed Labour will “increase your taxes by £2,094” (a figure still being cited on X today).

There’s a lot to unpick here, so be prepared for a relatively nuanced answer!

In case you’ve been lucky enough to miss all this, the Conservatives published a document last month, launched by Chancellor Jeremy Hunt, which identified a “black hole of over £10 billion a year by 2028-29 or nearly £38.5 billion over the next four years”.

The £2,094 figure simply divided the £38.5 billion by the number of working households (thus excluding households where no-one works, though this is not unreasonable as they are less likely to pay a large amount of tax).

Prime Minister Rishi Sunak then used that as the basis for his claim in ITV’s election debate on 4 June that a Labour government would mean “£2,000 of tax rises” for working families.

Life is too short to go through every item in the original report. Suffice to say that Labour has disputed that the spending measures are actually their official policies, while the Conservatives have argued that the assumptions made are cautious. Instead, I want to focus on how the numbers have been presented.

As a starting point, it is potentially misleading to describe the £2,000 figure as “costed by Treasury officials” or “by independent, impartial civil servants”. This is for two reasons.

First, where officials have produced costings, these will have been based on assumptions provided by Special Advisers who, by definition, are neither independent nor impartial.

Second, some of the costings were drawn from other sources, with no input from officials. This was an important part of the pushback from the Treasury’s top civil servant, Tom Bowler, who has (rightly) stressed that “any costings derived from other sources or produced by other organisations should not be presented as having been produced by the Civil Service.”

He also wrote, “as you will expect, civil servants were not involved in the production or presentation of the Conservative Party’s document Labour’s Tax Rises or in the calculation of the total figure used.”

In addition, the Office for Statistics Regulation (OSR) has raised the concern that people hearing the £2,000 figure might think it refers to their tax bill in a single year, rather than an estimate summed over four years. This is a valid criticism too.

Nonetheless, I do not think this justifies the more extreme headlines that this row has generated, including claims that the Treasury has “rubbished” the Conservatives numbers, or that Rishi Sunak has “lied”. Here there are three points.

First, most of the numbers were indeed signed off by Treasury officials – albeit at an earlier stage. You can find a long list here.

Second, where the estimates did not come from the Treasury, they came either from other reputable sources – or from Labour itself (notably £23.7 billion for the Green Prosperity Plan, which is the biggest single element of the Conservatives’ costings)

Third, while I’m not usually keen on ‘whataboutery’, Labour is just as guilty. They have produced a document called “Conservatives’ Interest Rate Rise” which contains some ridiculous assumptions about Conservative policies. For example, the document assumes that the Tories would abolish National Insurance *completely* in the *first* year of the next parliament, which is definitely not official policy.

What’s more, while Labour did not (and could not) claim that their analysis had been endorsed by the Treasury, they did butcher some Treasury analysis of the impact of changes in the fiscal stance to claim that Conservative policies would add 2.5 percentage points to mortgage interest rates. That figure does not stand up to a simple test of plausibility, let alone any serious scrutiny.

The OSR’s criticism of the presentational trick of adding up tax or spending over several years to give a bigger figure applies equally to Labour too.

So, to the original question, should civil servants be involved in costing opposition policies – in any way?

I’m not as squeamish about this as some. This has been standard practice for decades. In this link you can find examples from 2009, under the last Labour government.

What’s more, officials routinely provide costings of policy changes to MPs and peers (of all parties and none), who are free to use them as they wish. It seems odd that ministers should not be able to ask for these costings too.

Gemma Tetlow, Chief Economist at the Institute for Government, has made a strong case that the job of costing opposition policies should be removed from ministerial influence, by transferring it to the Office for Budget Responsibility (OBR).

But I’m not convinced that this is necessary – as long as the basis on which the figures have been calculated is made clear. There is also a danger that the OBR’s verdict is treated as ‘gospel’, giving this single body even more influence than it already has.

Finally, of course, there are already plenty of independent ‘fact checkers’ who can assess the reliability of any numbers that politicians are tossing around.

Here, for instance, is Full Fact’s judgement on the £2,000 claim: “This figure is unreliable and based on a number of assumptions. It comes from a Conservative estimate of Labour’s “unfunded spending commitments”, but Labour has said this is “flawed” and many of the costings behind the calculation are uncertain. Even if the figure was right, we can’t be certain this money would be collected by raising taxes, and if it was, families are unlikely to be affected equally.” There are other examples here and here.

In short, as an ex-HMT economist, I would not have been happy either with how the £2,000 figure has been presented. But that in itself does not invalidate it. There could also still be some role for expert civil servants in producing these costings, provided this is explained properly. And if politicians do use dodgy stats, they will soon be found out.

• Applied economics
• Fiscal Policy
• Politics
• Taxation
• UK economy