Can you trust Treasury forecasts when they struggle with basic arithmetic?

The Treasury are playing a key role in the referendum campaign. They have published two reports which explicitly campaign for remain. The first looked at the long term impact. The second report published yesterday, looks at the immediate impact. Both reports are all doom and gloom. They claim our economy will quickly be pushed into a recession and we will be £4,300 poorer in the long term.

Like the Government’s referendum leaflet (which cost £9m to publish), this analysis isn’t balanced. It is openly designed to promote Project Fear. It also cost taxpayers’ money to produce. Civil service staff are being diverted to fight the remain campaign. These are vast resources which aren’t at the disposal of the other side, essentially circumventing the funding limits. A team of 20 treasury wonks would cost around £1m a year. 

Experts often make errors. From Corbyn’s recent election as Labour Leader, to the 2015 Tory majority or 2007-8 financial crisis – the establishment often get predictions horribly wrong. Philip Tetlock, a professor of political science at the Wharton School of Business is one of the world experts on forecasting. From 1984 to 2003 he tested 284 experts from government, economics and wider academia. On the 28,000 predictions they made, they were "roughly as accurate as a dart-throwing chimpanzee". We have little reason to believe the Treasury predictions are any better.

As a concerned taxpayer, I submitted an FOI request to the Treasury. We have a right to know how much this all costs. I asked them how much they are spending on the EU referendum and in particular for staffing numbers. Their (late) reply speaks for itself. 

“It is not possible to identify full time equivalent staff numbers involved in the production of HM Treasury’s analysis because of the range of staff who contributed on an ad-hoc basis from across the Department. We have not yet received the publication invoice.” Information Rights Unit, HM Treasury.

It is not possible to count up the team involved. But it is possible to predict the next 30 years to the nearest pounds and pence, complete multivariate macroeconomic modelling, and work together as a pan department team to quickly turn this around into 290 pages of reports. Either they can't do basic arithmetic or they are hiding the amount they are throwing at the campaign.

Naturally I have asked for a review, and will be writing to the information commissioner in due course. Having worked for the government in various forms, I know that they have this basic data. If the number is so great as to defy a count, they could ask HR.

Whilst people may have provided support on an ad-hoc basis, this is true of most organisations, projects and reports. They could still count the number of authors and core team members involved. They could even estimate the wider support. If the treasury is incapable of approximating the number of staff involved in an internal project, what can be said for assumptions used in their reports?

290 page reports aren’t conjured out of thin air, a team would have been created, or existing teams would have taken on the task. Often departments have to approve a business case and set aside funding before diverting resources to a major project. 

The Treasury’s forecasts have been criticised from many angles – their poor forecasting record (consistently missing borrowing targets), use of fishy assumptions, focus on grabbing headlines and use as ‘independent statistics’ whilst being instructed by a biased government. Either way, the public has a right to know the facts, including how much taxpayers’ money is being spent on promoting remain.