Some economists have tried to identify the impact of the departure from the EU by using a ‘synthetic control’ method. This relies on a computer algorithm to select a weighted combination of economies whose performance best matched that of the UK before Brexit.

The actual performance of the UK economy since Brexit is then compared to this control group, or ‘doppelganger’, and the difference taken as a proxy for the impact of leaving the EU.

This approach is fine as far as it goes, but that is not very far. There are a number of problems with all these studies, including the sensitivity of the results to the choice of countries in the control group and the weights assigned to them. Successive iterations of each model have often required some large changes in order to ensure a ‘good fit’.

What’s more, even if the 15pc figure for the hit to trade turns out to be correct, there is still a big leap between this and a 4pc fall in GDP. To get this number you have to make a strong assumption about the link between trade openness and productivity.

Again, basic economics would suggest that a reduction in trade is bad for economic growth.

But the hard evidence for the magnitude of this effect is still pretty thin, especially for economies like the UK that are already relatively open and developed.

Even the historical studies typically relied on by Brexit pessimists show results that are imprecise and only marginally significant in statistical terms. Others have found no link at all, or that productivity appears to drive trade, rather than the other way around.

Indeed, productivity is notoriously hard to explain or predict – as the OBR itself has repeatedly found. While it is plausible that a reduction in trade intensity alone could reduce productivity, Brexit might have positive effects on productivity too. For example, better management of immigration could help to shake the UK out of a low pay and low productivity trap.

The upshot is that we still only have assumptions about the long-term economic impacts of Brexit, not hard facts. There is at least some evidence on the impact of Brexit on trade, but none at all on the impact on productivity, which is just as important for the 4pc figure for the hit to GDP. The OBR’s latest analysis is, therefore, not the “final proof” that many would have us believe.