Treasury adopts Newton’s theory
Exaggerated forecasts of UK downturn
by Brian Reading in London
Thu 26 May 2016
The UK Treasury this week has produced a report saying a vote to leave the European Union on 23 June would produce ‘an immediate and profound economic shock creating instability and uncertainty which would be compounded by the complex and interdependent negotiations that would follow’. The document says this would increase unemployment by around 500,000 and reduce GDP by at least 3.6%. Average real wages would be lower, inflation higher, sterling weaker, house prices would be hit and public borrowing would rise, it says.
Having looked at the methodology, I could hardly stop myself laughing. It is as bad as or worse than using Newton’s gravity theory as an instrument of economic forecasting. I am not surprised that the estimable Charlie Bean, the former Bank of England deputy governor called in by the Treasury to review the assessment, put his name to it. He has used and contributed to forecasts based on the same methodology.
Undoubtedly the referendum adds to uncertainty. Yet inspecting the chart for growth and uncertainty – based on a Treasury-constructed ‘composite index of uncertainty’ – I find the Treasury’s conclusions of a causal link between heightened uncertainty and lower growth deeply flawed.
In the two cases studied – the early 1990s recession and the steep downturn in 2008-09 associated with the financial crisis – either there is no correlation or, more likely, the causality runs in the opposite direction, with falling growth generating uncertainty rather than the other way around. Following the 2000 bursting of the dotcom bubble, uncertainty peaked and UK GDP rose 3%.
To suggest a two-year period in which GDP could fall by either 3.6% (in a ‘shock scenario’) or 6% (‘severe shock scenario’) must be a gross exaggeration.
The scale and impact of the uncertainty have been analysed using a vector autoregression (VAR) model to identify the effect of increased uncertainty on economic activity. However the VAR model, and the application of the NiGEM global economic model developed by the National Institute of Economic and Social Research, are both suspect.
The period analysed using VAR starts in 1998 and covers the 2008-09 recession. The Treasury’s 'composite index of uncertainty’, based on survey evidence including the number of references to uncertainty in the media, implied volatility for the FTSE 100 and of sterling and so on. I have no objection to this.
But all current uncertainty is attributed to the referendum. A future level of uncertainty is simply assumed. Moreover, the Treasury is basing its findings on acceptance of its own flawed predictions for the long-term effects of Brexit to 2030. Surely 23 June cannot be the only factor the Treasury believes affects uncertainty? Or that there is no uncertainty about its own forecasting record?
The effects of uncertainty on growth is based on another model which has seven other variables – including VIX for the FTSE 100 and S&P 500. Again I would question the validity of such criteria, since we all know the stock market can amplify and distort expectations about actual economic developments.
British membership of the exchange rate mechanism was significantly responsible for the UK’s 1990 recession, while the global financial crisis led to the 2008-09 recession.
It is mystifying why the Treasury has allowed itself to be misused in what appears a blatantly political way. The forecasts seem about as accurate as the predictions by John Major, then prime minister (relying on Treasury advice), in early September 1992 on the harmful consequences of sterling devaluing or leaving the ERM.
Just over a week afterwards, sterling was outside the ERM. The consequences were almost entirely benign. I suggest that, if the UK leaves the EU after 23 June, history will repeat itself.
Brian Reading was an Economic Adviser to Prime Minister Edward Heath and is a Member of OMFIF's Advisory Board. This is No.70 in the series – the 100th article will appear on 23 June.
OMFIF’s series on the UK EU referendum presents a wide variety of perspectives from Britain and around the world ahead of the 23 June poll. We are assuring a balance between many different points of view, in line with OMFIF’s overall neutral stance on the issue.
Tell a friend