When the Treasury published its analysis in April of the long-term impact on the British economy of leaving the European Union, it was immediately accused of inventing statistics to generate headlines claiming each UK household would be £4,300 a year worse off.
On Monday George Osborne published another report, this time on the immediate impact of a Brexit vote. And again, the ensuing headlines warning of a "DIY recession" were exactly as chilling as the chancellor hoped.
But, once again, all may not be as it seems with the numbers. A close reading of the latest document suggests that the Treasury has mysteriously excluded a scenario from its modelling that, in all probability, would have shown the UK not entering a recession if the public votes Leave on 23 June.
The Treasury document presents two scenarios. Under a "shock scenario", the British economy would lose 3.6% of GDP compared to remaining in the EU. The other scenario envisaged is "severe shock" – under which the economy would lose 6% of GDP compared to staying in.
Under both scenarios, the Treasury says, the economy would fall into recession.
But in the earlier Treasury document looking at the long-term economic impact of Brexit, three possible scenarios were presented: a European Economic Area deal similar to Norway's, a bilateral agreement with the EU, and World Trade Organization (WTO) membership:
Yet, while Monday's document references all three, as the table above shows, its modelling took into account only two: the negotiated bilateral agreement and the WTO option.
In this graph, from Monday's analysis, you can see that all three long-term scenarios are plotted on the right, but for the short-term impact, scenarios with only two central estimates are modelled:
This is potentially significant because the missing third scenario – a Norway-type deal – would probably not have shown a recession.
One of the drivers behind the Treasury analysis is something called a "transition effect" – an effect that explains that people adjust how they behave today based on future expectations.
The Treasury's assumption in the published scenarios is that many people will spend and consume less, and businesses invest less, because they would fear that in the longer-term the country would be poorer if Britain were to leave the EU.
Why would people think this? Basically, because the Treasury assumes they would believe the analysis it published in its assessment of long-term impact on Britain’s economy of leaving the EU. So, what the Treasury is saying here is that economic actors would behave this way because they will share the government's analysis of the UK's longer-term prospects outside the EU.
Of course, the short-term and longer-term analyses are separate pieces of work that use different methodologies. Moreover, transition effects are not the only driver in the Treasury's model. This also takes in uncertainty effects and financial conditions. For these, the Treasury will have averaged and weighted the uncertainty of all possible alternatives to EU membership to produce its "shock" and "severe shock" scenarios, because immediately after a Leave vote nobody would really know which path the UK would eventually take.
Monday's Treasury document is, however, clear that its transition effects are linked to Britain's future arrangement with the EU. It says, for example: "In the 'severe shock' scenario – which represents a credible risk – the size of the transition effect is linked to the estimate of Britain leaving the Single Market and defaulting to WTO membership."
So, if the Treasury had linked the transition effect to a Norway-type deal it would have found a smaller economic impact. Given that the recession forecast under the "shock scenario" is so mild – four quarters of -0.1% growth – it is quite probable that modeling for a Norway-type EEA deal would not have predicted a recession at all.
The Treasury's own logic and calculations show that the longer-term impact of an EEA type deal would have a substantially lower impact on the economy than a negotiated bilateral agreement or a WTO agreement (-3.8% vs -6.2% vs -7.5% on GDP compared with continued membership, and all else remaining equal).
Nevertheless, the Treasury analysis forecasts a recession as fact, and the claim has been central to how its document was presented.
The Treasury is not alone of course in predicting that the UK economy would be hit in the short term by Brexit. There is broad consensus on this issue – and most agree that a vote to leave the EU would hit the British economy in the immediate future. The OECD and the Bank of England have both reached a similar conclusion.
Even many on the Leave side of the debate agree with the notion that the economy would be hit because of uncertainty in the immediate aftermath of an exit vote. Gerard Lyons, a former economic adviser to Boris Johnson, has mentioned the idea of a “Nike swoosh”, under which the immediate impact of Brexit would a hit to GDP, but this is then followed by a large uptick.
But forecasting a recession as fact is much more controversial, and leaves George Osborne open to accusations, once again, that the Treasury is highlighting only the numbers that advance its case.
Alberto Nardelli is Europe editor for BuzzFeed News and is based in London.
Contact Alberto Nardelli at firstname.lastname@example.org.
Got a confidential tip? Submit it here.