Brexit free trade critics are wrong to dismiss the classical model of UK trade

Brexit free trade critics are wrong to dismiss the classical model of UK trade

Last week, Economists for Free Trade – a 16-strong group of experienced UK applied economists, which I chair – issued the opening parts of its report on the economic future after Brexit. My introduction to the report was greeted with warm and sometimes abusive disagreement from some quarters. This deserves a response.

In a nutshell, our proposals were for free trade, delivered by either a lot of trade agreements or, where these cannot be achieved, by unilaterally eliminating associated trade barriers; the replacement of EU regulation of our whole economy under the Single Market rules by pragmatic UK-orientated regulation; the ending of taxpayer-subsidised unskilled EU immigration; and, of course, the return of our EU net budget contribution. In fact, these are all current government policies but the government has been quite reticent in discussing the gains that come from this Brexit programme. Our work – using our research – has been designed to fill that gap. Our estimates of the net gain to the economy are around 7% of GDP, £135 billion.

Our critics made three main points:

  • That our free trade proposals involve dismantling important consumer standards by the elimination of non-tariff barriers (e.g. Prof Alan Winters, Sussex University)
  • That free trade and its implied competition with world producers would create mass unemployment or ‘economic suicide’ (e.g. Open Britain)
  • That our trade model was a classical one going back to Ricardo and not the ‘gravity model’ favoured by many of today’s trade economists (this is probably the most important of the criticisms)

These points are not new but are essentially a recycling of points made during the referendum and immediately following, to which I did reply (see the EFT webpage section of Responses to Critics.) Thus, rather than replying to detailed points again, I will take this opportunity of responding broadly to the three big themes listed above.

Consumer Standards

No, we do not propose the dilution of good consumer standards. There are by now, in most product areas, international bodies that monitor and agree standards. These are generally applied in all developed countries, which are members of these bodies. They may not all be identical but usually there is mutual recognition where differences exist. One of the benefits of Brexit is that the UK will be able to again take up her influential place on these bodies.

The definition of a non-tariff trade barrier is that there is discrimination against particular suppliers or even against all foreign suppliers in favour of domestic suppliers. This discrimination can take the form of standards that are, in effect, set in order to exclude suppliers; or of anti-dumping duties or the threat of them; or simply plain quotas on products that can be supplied. Often this
discrimination is subtle and concealed by arguments for the ‘greater good’. It is therefore quite hard to measure.

Nevertheless, in the case of the EU, efforts have been made, either (as we have done) by using detailed and comparable prices for OECD countries (averaged across broad product categories after allowing for transport and other margins) or by direct evaluation of measures by product as done by Berden et al, 2009*. It so happens that both we and Berden et al come up with similar figures on manufactures, implying a total trade barrier of around 20%, of which about three quarters is the non-tariff element. This is what we propose should be eliminated by trade agreements, if possible.

Mass Unemployment due to Free Trade Competition

Here our critics, led by Open Britain, say that free trade and the competition it brings to the currently protected farming and manufactures sectors will destroy the jobs in these sectors, so creating mass unemployment and ‘economic suicide’.

Truly, these critics speak with forked tongue. On the one hand, we have the Treasury emphasising the big gains to be made from free trade and competition, but bizarrely forgetting to include free trade with non-EU markets. Indeed, it is commonplace to hear politicians of all hues praising the virtues of competition because it raises productivity, living standards and wages. They know well that this competition will also create more jobs, in place of less productive existing jobs, because the higher wages will make work more attractive relative to benefits and the minimum wage.

But suddenly our critics have forgotten what they said when in pro-competition mode, and assert that protection and the avoidance of competition is needed to protect jobs and prevent unemployment. All such a policy does, however, is to preserve existing jobs, even though unproductive, and arrest the process of economic improvement via higher productivity. In short, to continue with protection is to prevent the competitive process by which new and better jobs will be created raising the employment rate even further.

Maybe, in the back of their minds, they worry about a transitional period when the protected industries lose their protection before the new jobs can be created. Yes, it is well-known that, when productivity is rising, it is good policy to support the economy with strong demand to help the creation of new jobs. This was the case in the 1980s and it is also the case today: we have had a massive Brexit devaluation – some 15% – which is stimulating our traded sectors, including farming and manufacturing. Recent CBI manufacturing surveys are the strongest on record for 20 years.

We have calculated that, as long as this devaluation stays in place, manufacturing will be better off after a full Brexit than before (even if the EU – against its own interests – elects to raise tariffs against us). In the longer term, as the devaluation may be whittled away, manufacturers will need to raise productivity incrementally by less than 1% a year to be as profitable as before. This should not prove daunting as manufacturing has routinely achieved 3% growth in productivity for three decades while under restrictive regulation of the EU.

Farmers are also gaining from the devaluation and they too can raise productivity over the longer term. In addition, the government has pledged to help directly those that are less efficient but contribute vitally to the rural environment.

So what we will see is that free trade will – through the pressures of competition – move resources and jobs into more productive use, raising total GDP, real wages and employment in the process.

The Gravity Model

It is for our modelling efforts that our critics have reserved their most haughty displeasure, rather like a grumpy trade union, suggesting that only they are ‘experts’ in trade matters. Trade economists nowadays largely espouse the gravity model and they dismiss our classical model going back to the work of Ricardo, via such great economists as Samuelson, Heckscher and Ohlin, as ‘old’. It is not obvious that being an old model is a bad thing. Rather it suggests that it has survived the rough and tumble of two centuries of trade experience without being seen as irrelevant and misleading: as a result, the key welfare result obtained from the classical model, that free trade raises welfare and GDP, is almost universally accepted by economists and politicians.

Also, old models can have the last laugh empirically. Think how the Keynesian model became the fashion among macro-economists after WWII; and how it has been ignominiously dethroned by now in favour of the ‘New Keynesian’ model, which is essentially a classical model with some short-term price rigidity.

The gravity model is widely used to support the case for UK integration into the EU as our closest, largest neighbour. It also is thought to suggest lesser gains from lowering barriers against the rest of the world as compared to having less streamlined relations with the EU – though gravity modellers have been rather coy about assessing the impact of general free trade policies.

In fact – on the face of it – the model does not fit the UK’s trade history at all well. It cannot account for the surge in our services exports all around the world, or our huge trade with Anglosphere countries, or indeed the steady decline in the share of our trade with the EU. The classical model, based on comparative advantage, by contrast seems to fit these facts well. As for those ‘gravity associations’ found in the data on trade and GDP; actually, virtually all models can fit them well enough, as has been well known for the decades since Jan Tinbergen discovered them.

The irony of the situation is that the gravity modellers have produced no empirical justification for their claims beyond saying that the model fits these associations – which is plainly inadequate. It just seems that they have become convinced somehow that the central gravity ideas of imperfect competition and the direct impact of trade on productivity via an FDI channel are correct. However, how precisely in practice do these ideas fit the UK economy?

To answer these questions and to try to establish what model does make sense for the UK economy, my research group embarked a year or so ago on a project to determine which of these models fit UK data best, and how much impact the differences make to policy calculations. To do this, we have drawn not just on trade theory but crucially also on recent developments in the contiguous areas of computer methods, econometrics and applied macro modelling; as a science, we economists need to remember that cross-fertilisation is important in economics and we should not be narrow-minded.

The results of this exercise are rather instructive and can be found in a working paper just posted online here as ‘Classical or Gravity? Which trade model best matches the UK facts?’ (Cardiff Economics working paper no 2017/10 by Patrick Minford and Yongdeng Xu).

The bottom line is that both models fit the UK post-war facts, though the classical model has a rather higher probability of being correct. Nevertheless, the gravity model makes little difference to the policy implications of free trade. That is, we ran the same free trade policies through the gravity model and got the same welfare gain as for the classical model.

Why are the results so similar? The gravity model says there is ‘imperfect competition’ but in practice this makes little difference to price behaviour, as prices instead of being equal to costs are simply equal to costs times a constant mark-up. Furthermore, while this imperfection makes it more difficult to penetrate foreign markets and balance the current account, the floating exchange rate acts as a policy instrument to achieve this; meanwhile the supply-side of the economy behaves as in the classical model. Thus, while one can indeed lose trade with the EU due to additional barriers, one can also gain trade with the rest of the world as barriers come down; the elasticities in both cases are similar. As a result, total trade is not much altered. Consequently, the link to worse productivity via lower FDI does not much kick in, if at all.

This result rather confirms what we thought during the referendum about the many model assessments that damned Brexit: these damning verdicts came not so much from the models, whether gravity or not, but dominantly from the policy assumptions made about Brexit.

I made this point in The Brexit Consensus Bug (posted on our webpage in the section The Critique of Project Fear). Brexit policy was assumed by most modellers to be inward-looking, tough on all immigration, protectionist and interventionist in regulation. When I asked some of these model representatives at an NIESR conference on May 27th 2016 why they assumed these poor policies, they shrugged their shoulders and said something like ‘nothing else is politically possible’.

The role of economists is surely to explain the consequences of alternative policy options – not to foreclose them. Now furthermore that we have a government that is committed to free trade, a normal skill-based immigration policy, and sensible domestic regulation, should not these same modellers reconsider their assessments of what Brexit will bring?


Our critics have had some harsh things to say about the analysis and expertise of Economists for Free Trade in our recent Report. However, we are simply enunciating what good economics has always emphasised: that competition and free trade are good for the economy. It turns out, as in so many other areas of policy where competition is a key element, that the UK evidence on trade
favours our position.

It is in all our interests that the best economic policies are pursued in the coming years to make Brexit a success. We do not claim to have got everything perfectly right and though what we suggest is we think in the right ballpark, we welcome dialogue with our fellow economists to refine and improve our analysis. What we feel however is that cheap and childish abuse should cease in favour of such an adult dialogue.

*Berden, K., Francois, J., Tamminen, S., Thelle, M., and Wymenga, P. (2009): “Non-Tariff Measures in EU-US Trade and Investment: An Economic Analysis,” Final report, Ecorys; cited in Brexit 2016- Policy analysis from the Centre for Economic Performance, 2016, LSE; Table of ntbs on p 123.