Transcript Tom Settle, Tonbridge School

Why do economists struggle to forecast
how the UK economy will perform?
Tom Settle, Tonbridge School
To see more of our products visit our website at www.anforme.co.uk
Background
• C Gordon Brown’s declared in March 2007 “the end of boom and
bust”.
• C This reflected the prevailing conventional wisdom that
economics had finally triumphed over the business cycle.
•C
Uninterrupted economic growth, low unemployment and stable prices
had come to be expected.
•C
In such tranquil times, economic forecasts, which seemed only to
reinforce the view that growth would continue at an unabated steady
pace, unsurprisingly failed to grab the public’s attention.
• C The implosion of the world’s financial markets and the onset of the
‘great recession’ in 2008 exposed the fragility of the economy and the
uncertainty that surrounds the performance of economic variables.
Are economic forecasts
inaccurate?
• C For as long as there have been economic forecasts, there has been a
certain degree of scepticism about their validily.
• C In the 1950s, J.K.Galbraith said: “the only function of
economic forecasts is to make astrology look respectable.”
• C However, the recent failure of forecasters to predict the
severity of the current downturn has had a pervasive effect
on the level of trust in such forecasts.
“A stopped clock
is right twice a day....”
• C In 2008, on the eve of the biggest downturn in economic activity since the
Great Depression, the Bank of England predicted that: “GDP will be
broadly flat for the next year or so.”
• C The Bank was not alone in thinking any slowdown would be small and
transitory.
• C The average forecast of 30 of the world’s most significant financial
institutions predicted UK growth in 2008 and 2009 to be 0.8% and -1.1%
respectively – when in fact the economy shrank by over 6%.
• C Forecasters not only failed to foresee the onset of recession, they have
also had little success in predicting the path to recovery.
• C The Office of Budget’ Responsibility (OBR) came only 30th in a league
table for the forecasts for 2011. The IMF was placed 33rd and the
European Commission 36th!
Are forecasts always
this inaccurate 1?
C
Long-term versus short-term forecasts:
• C Institutions like the Bank of England or the OBR make forecasts that look at
varying lengths into the future.
• CNot surprisingly, forecasts that focus on a short time
horizon (up to one year) are more accurate than those
that look further into the future.
• CResearch by the Adam Smith Institute shows that
whilst one-year forecasts are correlated with actual
outcomes, two and three year forecasts show little
correlation, and have very little predictive power.
Are forecasts always
this inaccurate 2?
C
Stability and volatility:
• C Today’s economic climate represents the most tumultuous episode in postwar economic history.
• C But the preceding decades from the mid 1980s to 2007 showed a
prolonged period of economic stability, which has been called “The Great
Moderation”.
• C But forecasts consistently fail to identify large ‘shocks’ to data and
significant changes ahead of time.
• C Unfortunately it is in these very periods that firms, households and
policymakers need the most accurate data to reduce the impact of any
changes as effectively as possible.
Why so inaccurate 1?
C
Rubbish in, rubbish out:
• C The predictive power of even the most powerful computer programme is
limited by the accuracy of the data fed into it.
• C The reality is that much of the data available is only an
estimate of the reality, subject to revision with hindsight.
Even past GDP figures are subject to constant change.
• C For forecasters it is the equivalent of aiming at a moving
target from a moving platform.
Why so inaccurate 2?
C
Fooled by randomness”:
• C Significant ‘turning points’ in unemployment, inflation and growth can be
created by changes which are hard to uncover and difficult for forecasters
to anticipate.
• C One potential cause of such rapid changes are external shocks.
• C In a globalised economy the impact of changes in international markets
has a much greater impact on UK performance, such as increases in oil
prices causing a surge in domestic inflation.
• C Even when forecasters can identify a potential shock on the horizon, it
may be difficult to estimate the precise impact on the UK economy.
Why so inaccurate 3?
• C Another cause of rapid change is changes in ‘animal spirits’.
• C Behavioural economists emphasise that turning points in economic
activity may be caused, or at least reinforced, by changes in sentiment
amongst firms and consumers.
• C People’s perception about the economic outlook changes their behaviour
in ways that make that outcome more likely to occur.
• C The result is that small changes in confidence, or as Keynes termed it
‘animal spirits’, may create significant changes in the final performance of
the economy.
• C The problem for forecasters is that although business surveys are apt at
indicating the level of confidence amongst investors, it is very difficult to
foresee changes in sentiment ahead of time.
Why so inaccurate 4?
C
Confusing risk with uncertainty:
• C A recent fundamental cause of forecasting error has been the failure of
modern forecasting models to consider the small, but significant, distinction
between risk and uncertainty.
• C Situations with risk can be defined as those where the outcomes are
unknown, but governed by probability distributions.
• C These differ from ‘uncertain’ situations, where the outcomes are also
random, but governed by an unknown probability distribution.
• C Many models observe the recent performance of key variables, expecting
them to behave in the same way in the future.
• C The result is that such models underestimate the chances of significant
shifts, simply because such large changes are observed so infrequently.
Conclusion: Are inaccurate
forecasts a problem 1?
• C Forecasts matter in an economy where many of today’s decisions must be
made on the basis of future economic conditions.
• C Firms must decide whether or not to invest in new capacity based on the
likelihood of increased returns.
• C Governments must set policies ahead of time if they are to head off
undesirable movements in inflation or unemployment.
• C Household spending decisions are at least partly predicated on the
anticipation of future earnings.
• C Forecasting errors lead to sub-optimal decision-making and inappropriate
policy solutions and should attract public scrutiny.
Conclusion: Are inaccurate
forecasts a problem 2?
• C However, forecasts cannot remove all the uncertainty that is inherent in the
future.
• C It is important to understand the limitations of any attempt to predict the
unknown.
• C Forecasts should be viewed as a tool to aid decision-making, a rough
guide to the direction of travel rather than an oracle of the future.
• C In this more befitting context, the errors of forecasters would have much
less pervasive effects.