Economic forecasting came in for much criticism at the time of the financial crisis and credit crunch. Few economists had predicted the crisis and its consequences. Even Queen Elizabeth II, on a visit to the London School of Economics in November 2008, asked why economists had got it so wrong. Similar criticisms have emerged since the Brexit vote, with economic forecasters being accused of being excessively pessimistic about the outcome.
The accuracy of economic forecasts was one of the topics discussed by Andy Haldane, Chief Economist at the Bank of England. Speaking at the Institute for Government in London, he compared economic forecasting to weather forecasting (see section from 15’20” in the webcast):
“Remember that? Michael Fish getting up: ‘There’s no hurricane coming but it will be very windy in Spain.’ Very similar to the sort of reports central banks – naming no names – issued pre-crisis, ‘There is no hurricane coming but it might be very windy in the sub-prime sector.” (18’40”)
The problem with the standard economic models which were used for forecasting is that they were essentially equilibrium models which work reasonably well in ‘normal’ times. But when there is a large shock to the economic system, they work much less well. First, the shocks themselves are hard to predict. For example, the sub-prime crisis in 2007/8 was not foreseen by most economists.
Then there is the effect of the shocks. Large shocks are much harder to model as they can trigger strong reactions by consumers and firms, and governments too. These reactions are often hugely affected by sentiment. Bouts of pessimism or even panic can grip markets, as happened in late 2008 with the collapse of Lehman Brothers. Markets can tumble way beyond what would be expected by a calm adjustment to a shock.
It can work the other way too. Economists generally predicted that the Brexit vote would lead to a fall in GDP. However, despite a large depreciation of sterling, consumer sentiment held up better than was expected and the economy kept growing.
But is it fair to compare economic forecasting with weather forecasting? Weather forecasting is concerned with natural phenomena and only seeks to forecast with any accuracy a few days ahead. Economic forecasting, if used correctly, highlights the drivers of economic change, such as government policy or the Brexit vote, and their likely consequences, other things being equal. Given that economies are constantly being affected by economic shocks, including government or central bank actions, it is impossible to forecast the state of the macroeconomy with any accuracy.
This does not mean that forecasting is useless, as it can highlight the likely effects of policies and take into account the latest surveys of, say, consumer and business confidence. It can also give the most likely central forecast of the economy and the likely probabilities of variance from this central forecast. This is why many forecasts use ‘fan charts’: see, for example, Bank of England forecasts.
What economic forecasts cannot do is to predict the precise state of the economy in the future. However, they can be refined to take into account more realistic modelling, including the modelling of human behaviour, and more accurate data, including survey data. But, however refined they become, they can only ever give likely values for various economic variables or likely effects of policy measures.
Andy Haldane in Conversation Institute for Government (5/1/17)
‘Michael Fish’ Comments From Andy Haldane Pounced Upon By Brexit Supporters Huffington Post, Chris York (6/1/17)
Crash was economists’ ‘Michael Fish’ moment, says Andy Haldane BBC News (6/1/17)
The Bank’s ‘Michael Fish’ moment BBC News, Kamal Ahmed (6/1/17)
Bank of England’s Haldane admits crisis in economic forecasting Financial Times, Chris Giles (6/1/17)
Chief economist of Bank of England admits errors in Brexit forecasting BBC News, Phillip Inman (5/1/17)
Economists have completely failed us. They’re no better than Mystic Meg The Guardian, Simon Jenkins (6/1/17)
Five things economists can do to regain trust The Guardian, Katie Allen and Phillip Inman (6/1/17)
Andy Haldane: Bank of England has not changed view on negative impact of Brexit Independent, Ben Chu (5/1/17)
Big data could help economists avoid any more embarrassing Michael Fish moments Independent, Hamish McRae (7/1/17)
- In what ways does economic forecasting differ from weather forecasting?
- How might economic forecasting be improved?
- To what extent were the warnings of the Bank of England made before the Brexit vote justified? Did such warnings take into account actions that the Bank of England was likely to take?
- How is the UK economy likely to perform over the coming months? What assumptions are you making here?
- Brexit hasn’t happened yet. Why is it extremely difficult to forecast today what the effects of actually leaving the EU will be on the UK economy once it has happened?
- If economic forecasting is difficult and often inaccurate, should it be abandoned?
- The Bank of England is forecasting that inflation will rise in the coming months. Discuss reasons why this forecast is likely to prove correct and reasons why it may prove incorrect.
- How could economic forecasters take the possibility of a Trump victory into account when making forecasts six months ago of the state of the global economy a year or two ahead?
- How might the use of big data transform economic forecasting?