skip to navigation skip to content


Treasury economic modelling is flawed say economists from the Centre for Business Research (CBR)

Her Majesty's Treasury

New research from the Centre for Business Research (CBR) at the University of Cambridge Judge Business School questions the accuracy of measurements of the impact of Brexit on the UK economy undertaken by the Treasury and by other official and academic bodies, during and after the Referendum.

The new paper, “How the economics profession got it wrong on Brexit”, predicts that the decision to take the UK out of the European Union will only have a small negative impact on economic growth over the coming years, and is likely to have a minor impact on living standards. These conclusions contrast with the large negative impacts predicted by others and particularly by the Treasury.

The paper is written by Ken Coutts and Graham Gudgin, who are both Honorary Research Associates at the CBR, and Jordan Buchanan of the Ulster University Economic Policy Centre. Graham Gudgin is also Chief Economic Advisor at Policy Exchange, London, visiting Professor at the University of Ulster and Chairman of the Advisory Board of the Ulster University Economic Policy Centre. Ken Coutts is also Emeritus Assistant Director of Research in the Faculty of Economics, and Life Fellow in Economics, Selwyn College, at the University of Cambridge.

The paper examines in detail the predictions of a range of official and academic reports on the economic impact of Brexit issued during and after the Brexit referendum. The paper concludes that most of these, and especially the Treasury reports, were “flawed”.
Gudgin, Coutts and Buchanan argue that:

Most estimates of the economic impact of Brexit are based on inappropriate applications of “gravity” models of trade and on a “knock-on” impact on productivity which probably does not exist.

In each case the effect is to exaggerate the negative economic impact of Brexit.

This is likely to be a result of a mix of unconscious bias and political expectations and perhaps a degree of group-think.

HMT should not have been asked to conduct what is essentially an academic exercise. Their subsequent refusal to discuss their work with outside economists is unacceptable.

Coutts, Gudgin and Buchanan’s own prediction using the CBR macro-economic model is that the outcome of Brexit, based on a two-year transition period and a free-trade agreement, will involve minimal loss of per capita GDP by 2030 (compared with the Treasury prediction of a loss of over 6%)

Graham Gudgin said: “Our analysis does not support the frequently repeated claim that membership of the EEC/EU has been good for economic growth in the UK.” Growth in UK per capita GDP has been slower since the UK joined the EEC/EU than it was in previous decades. The fact that the UK’s growth appeared to improve relative to the major EU economies was wholly due to the dramatic slowdown in growth in these EU economies. Taking US growth as a benchmark, this shows that there is no evidence that UK economic growth improved as a result of EEC/EU membership.

The new paper shows that the now obvious errors of forecasting in the Treasury April 2016 report on the short-term impact of Brexit was due to an arbitrary( and wrong) assumption on the degree of uncertainty generated by the referendum result. Publicity about the forecast of an immediate year-long recession never made it clear that was simply based on an assumption.

A major exercise to replicate the gravity model of trade, which forms the core of the May 2016 Treasury report on the long-term impact of Brexit on the UK economy, shows that the model was inappropriately used by the Treasury and exaggerates the negative impact of Brexit on UK exports. The knock-on impact on productivity, which forms a large part of the Treasury’s estimate of the impact of Brexit, is much too large and may not exist at all.

Other official reports, from the OECD and IMF suffer from similar flaws. Although the flaws are less pronounced, the impact is still in the direction of exaggerating the negative impact of Brexit. All of this work predicts that the volume of UK trade with the EU will fall substantially after Brexit, with no offsetting gains in trade with non-EU countries. “While the Centre for Economic Performance at the LSE uses a different approach, we also believe it is flawed”, Ken Coutts added. “Around half of the predicted declines in GDP, come from a calculation that trade losses will have a major negative knock-on impact on productivity. Our update of their evidence suggests that for advanced economies no such link exists between trade and productivity.”

The different approach used by the commercial forecasting group Cambridge Econometrics in its report for the Mayor of London is judged to be based on more appropriate techniques. This predicts relatively small negative impacts on per capita GDP and hence living standards. However, media coverage of this report focussed only on the most negative aspects and missed the optimistic conclusion that living standards would be little affected.

Graham Gudgin concludes: “The short-term forecasts of the Treasury and OECD, which have turned out to be wrong, have further damaged the already weak public confidence in economists’ contributions to public debate. Our paper is not necessarily an argument in favour of Brexit. But it will cast doubt on traditional economic modelling and it does question the ability of the economics profession to provide high quality policy analysis on issues of national importance.”

Ken Coutts concludes: “We expect that econometric models used by commercial forecasters like Cambridge Econometrics, will prove to be most accurate in the long-run. If so, the academic profession needs to reconsider both the relevance of its current attachment to theory based on unrealistic assumptions, and to the general quality of policy-relevant applied work. The UK civil service should also consider, when undertaking technical economic research, that it has a duty to be transparent in the methods it uses and to discuss with outside economists the strengths and weaknesses of the analysis as is customary in academic debate.”

About the authors of the report

Graham Gudgin was a Senior Economic Advisor at Oxford Economics from 2007 to 2015 and was director of the Northern Ireland Economic Research Centre from 1985 to 1998, after which he became Special Adviser to the First Minister in the Northern Ireland Assembly until 2002. Prior to this he was economics fellow at Selwyn College, Cambridge and a member of the Cambridge Economic Policy Group under Wynne Godley. He is the author of a large number of books, reports and journal articles on regional economic growth in the UK, the growth of small firms and electoral systems.

Ken Coutts was a member of the Cambridge Economic Policy Group, led by the late Wynne Godley, in his early career. His main interests are in macroeconomics, monetary and fiscal policy, trade, capital flows and balance of payments. He has published widely in these areas. He has also written extensively on the pricing behaviour of manufacturing industries in the UK and Australia.

Jordan Buchanan is research assistant at the Ulster University Economic Policy Centre. Together with Professor Neil Gibson at UUEPC, he has contributed to the development and operation of the CBR UKMOD macro-economic model. He has also co-developed the regional sectoral model of Northern Ireland and the local government forecasting model. He has worked on a range of public policy research projects most recently including decomposing the productivity gap in Northern Ireland.