Remember the Treasury’s Project Long Term Fear? The long report during the 2016 referendum estimated that if we left the EU on ‘WTO terms’ with no trade deal with the EU it would cost us 7 per cent off GDP per annum in 15 years’ time compared with no Brexit.In the past week or so the ‘Civil Service’ or whoever in it (I will assume Dexeu though here refer to civil servants) has taken over Brexit assessment from the Treasury, leaked a report in which the same policy would cost us 8 per cent of GDP. Of course accompanying that are all sorts of detailed implications for areas of the country and industries, which are equally gloomy and have since also been leaked.
I must say I look back on the Treasury’s original 2016 report with nostalgia: at least it was published and downloadable and still is. The leaked ‘report’ is insubstantial and certainly unavailable.
The leaking has luckily told us some revealing things about it. First of all it is done with ‘quite different methods’ from the Treasury’s ‘gravity approach’. We are led to understand it uses a reputable Computable General Equilibrium Model- my guess is that this is GTAP, the workhorse created at Purdue University in Indiana that has been developed by multiple universities, government and international agencies since 1992. It is a very large model of the whole world and was used by Open Europe in 2016 via the Ciuriak Consultancy; the one they used which the Civil Service may well have used too had 57 sectors and 28 country groupings.
So far so good: These civil servants have dumped the Treasury’s gravity approach which was riddled with problems, essentially because it was just a set of correlations and could not properly be used to predict Brexit outcomes as if it was a ‘causal model’. Imagine using the correlations between crime and unemployment to predict unemployment!
However there the good news ends.
There are two main problems with the Dexeu/Civil Service analysis. The main one is the assumptions they have fed into the GTAP model. The report’s civil servant authors, like those in the Treasury and across much of the Civil Service, are obsessed with the EU’s importance in our trade. They see this in terms of the EU’s ‘closeness’ versus our ‘distance’ from the rest of the world; and especially in the avoidance of border costs. Both of these ideas are part of the gravity ideas in trade, that distance is costly and that border costs are large because of time to get paperwork agreed before ships allowed to unload.
However these ideas have been totally bypassed by the progress of technology. First, containerisation has reduced ocean transport costs to trivial amounts. Secondly, computerisation has more or less eliminated border costs among developed countries, since almost all ships are cleared before reaching port, with some 2 per cent or so physically inspected and even this taking only around a day typically. One of the great EU ideas was ‘integration’, to be created by having uniform EU regulation and no borders: today markets and technology mean there is no need at all for the EU ‘state’ to bring out its uniformity stamp any more.
Another of the civil servants’ assumptions is that we would not abolish EU non-tariff protection through our Brexit policies of free trade. But, we would be mad not to. That amounts to around 16 per cent on manufactures, a big difference form merely counting tariff protection which on manufactures is less than 4 per cent. Total protection on food and manufactures by the EU is around 20 per cent on both.
Having redone the GTAP trade calculations reported for Open Europe with the right assumptions there is a ‘WTO option plus EU Canada-plus trade deal’ gain of 2 per cent in place of the 5 per cent loss for this in the Civil Service Buzzfeed-leaked report- a difference of 7 per cent!
The other main problem in the calculations is with the GTAP model itself. It is a decent CGE model, much in principle like the much smaller one that my Cardiff research team uses. However it is vast and it may not model UK trade and the economy in particular very well. No one has tried to test it on the UK which is our concern in these calculations. However our smaller model does fit the UK well and our tests show it must be pretty accurate. On its assessment this WTO plus EU Canada+ deal generates a gain of 4 per cent. So with our model that fits the UK experience the gain is higher still than the one on the adjusted GTAP model: a gap with the Dexeu report of 9 per cent.
So there you have it in a nutshell. The old Treasury trade report with all its gravity correlations is dead; but its Buzzfeed-leaked successor is alive and as misleading as ever, 8 per cent of GDP misleading on the central deal we are concerned with for the UK, namely free trade under the WTO together with a sensible EU free trade deal. Since it gets the whole economy wrong, it also gets all its parts equally wrong in the same pessimistic direction, from North to South and from Wales to Scotland and Northern Ireland. Garbage in, garbage out, lots of it.