Flaky economics? Brexit, trade and the Treasury estimates

By Graham Gudgin, Ken Coutts & Neil Gibson

One of the more depressing aspects of the Brexit debate has been that once again the economics profession has not covered itself in glory. The predictions of economists were overwhelmingly pro-remain and not as soundly based as many assumed. Evidence viewed as pro-Brexit could be met by rudeness and we experienced this ourselves. A few economists supported the Brexit case but the middle ground seems to be under-populated, indeed hardly populated at all. This is a real problem since the general public needs neutral referees in such a difficult and fraught issue.

The great majority of the economics profession famously failed to foresee the economic crisis of 2008-9. In the UK they also greatly overestimated the scale of post-crash recovery especially in productivity (which remains a ’puzzle’ to most economists) and now have egg on their faces over the post-referendum performance of the UK economy. The Economists for Remain letter signed by 12 Nobel Laureates, three peers, seven knights of the realm, 88 professors of economics and many other academic economists predicted three main things: 

• A recession causing job losses will become significantly more likely due to the shock and uncertainty of Brexit
• A drop in the pound and increased tariffs on imports will cause… higher inflation
• Investment in the UK will drop harming innovation and future job growth

The precise nature, and especially timing, of these predictions was left rather vague but the recession which the letter implied would follow the referendum has clearly not occurred. The fall in the pound did happen immediately after the referendum, and of course will indeed lead to rising inflation. One out three correct predictions is something, but is well wide of a bulls-eye.

The tiny minority of economists which predicted economic gains from Brexit, was mostly within the ‘economists for Brexit’ group. These are strong free-marketeers who appeared to rely on a view that UK exports have the characteristics of primary commodities which can always be sold at existing world prices even if frozen out of protected EU markets.

The example of our own work which caused offence was a working paper published by the Centre for Business Research at the University of Cambridge [1]. Because we needed an estimate of the long-term impact of Brexit for our macro-economic forecasting model, and judged the Treasury’s estimates of trade losses from Brexit to be implausible on economic grounds, we set out to generate our own estimates. This was done by examining the time-series of UK exports to the EU28 countries since 1950, and also by reproducing the Treasury’s gravity model equation, albeit for a limited number of countries (the main UK export markets) and a short recent period. Our conclusion was that the Treasury estimates of the benefits of EU membership to UK trade were considerably exaggerated.

Two attacks on this work came in the widely-read blog of Simon Wren-Lewis, Professor of Economics at Oxford, and signatory of the Economists for Remain letter mentioned above. In his blog of January 18th 2017 he said: 

“But didn’t the CBR report say that the benefits of the Single Market had been exaggerated by the Treasury? Yes it did. Here is some of its reasoning. The growth of the UK market share after the single market is not as impressive as it looks, because there is an underlying 6% positive trend in the share, which you can detect before we joined the EU. That looks pretty on a picture, until you realise it is nonsense. A 6% trend in the export share will imply that at some point not too far away UK exports to the EU will be as high as total EU GDP”.

This would be fairly devastating if it were true, but it is not. Wren-Lewis had confused two of our charts. The 6% trend was on a chart of the growth of UK exports to the EU (chart 4) not the growth of the export share (chart 6). The volume of UK exports to the EU did indeed grow at 6% per annum from 1950-2000 as the chart (chart 4) clearly shows, and this was lower than the 7% growth in world trade over the same period.  The growth in the EU share of UK exports has grown much more sedately as the paper made clear. Importantly that share has been falling since the early 1990s and since the formation of the Eurozone at the end of 1999 UK exports to the EU have hardly risen at all. Wren-Lewis makes no comment in this blog on the implication of the falling share of UK exports going to the EU.

In his earlier blog ‘Fake Economics and the Media’ he had been even ruder. He defines ‘fake economics’ as “analysis or research that is obviously flawed whose purpose is to support a particular policy”. He describes a recent study by the organisation’ Change Britain’ as fitting this definition and then goes on to say, “The CBR study is less obviously fake”.  He made no attempt to support his implication that our report was written to support the Brexit policy. In fact, two of the three authors of the report oppose Brexit, but all three insist that evidence should be accurate.  The burden of his objection to the study is that several trade experts, contacted by the Independent Newspaper and including the American Richard Baldwin and our former Cambridge colleague, Alan Winters, “pour some very cold water over the study”. It is unusual for Wren-Lewis to rely uncritically on mainstream economists, but he was willing to do so in this case.  With many of Wren-Lewis’ articles being used by one of us in economics teaching to encourage students to query and to test what is considered ‘mainstream’ it seems more than a little surprising to be discredited for daring to do so.

The context here is the evolution of the economics of trade. After decades of highly theoretical but not very realistic work, trade economists started to do more empirical analysis. They discovered that simple gravity models fitted the trade data well and also allowed estimates to be made of the impact of membership of free-trade agreements and currency unions. Gravity models simply relate the amount of trade between two countries directly to the size of those countries’ economies and inversely to the distance between them. The impact of FTAs is estimated simply by adding dummy variables which identify the members of the FTAs. The technique is simple and empirical but the practitioners have evolved a conventional approach to using it, and are intolerant of any deviation from what they invariably call ‘best practice’.

Best practice involves including over one hundred countries and long periods of half a century or more when estimating a gravity model. Our model with only 28 of the UKs main trade partners and only three recent years broke with ‘best practice’, and was roundly denounced by trade economists contacted for an article in the Independent newspaper written by Ben Chu.

We contacted the experts named in the article but none would tell us what their objections were in detail. We have subsequently extended our database to include 120 countries and over one million observations, but still conclude that the Treasury estimates are exaggerated. This exaggeration should have been pointed out by the trade experts themselves, who knew or should have known, that the meta-analysis of 159 papers reported by Head and Mayer in the Handbook of International Economics [2] reported an average uplift in trade due to EU membership of 16%. This is much lower than the Treasury’s 118%, but as far as we know no trade experts have commented on this disparity.

The question of how large the impact of EU membership is on trade flows remains a paramount policy issue, affecting not only the Brexit negotiations but also the question of Scottish independence and even Irish unity. What is needed is an informed debate, but the suspicion lurks that many economists have been influenced by their political preferences and a tendency to rely on theory rather than careful assessment of the data when writing on questions of practical policy.  

Drs Graham Gudgin and Ken Coutts are Honorary Research Associates at the Centre for Business Research (CBR) in the Judge Business School at the University of Cambridge.  Neil Gibson is Professor of Economic Policy and Director of the Economic Policy Centre, Ulster Business School at the Ulster University;

Footnotes

[1] The Macro-economic Impact of Brexit. Centre for Business Research, University of Cambridge. Working Paper No. 483. www.cbr.cam.ac.uk/publications/working-papers/2016 

[2] Handbooks in Economics. Vol 4 International Economics. Chapter 3, table 3.4. 2015