12 January 2017
Economics – a profession in crisis or at a watershed?
We should be wary of regarding economics as a science.
by Frank O’Nomics
The American economist Paul Samuelson once joked that Wall Street indices had predicted 9 out of the last 5 recessions. The joke could just as easily be applied to the economics profession over the last few years. There is a tendency for economists to develop a view of the economy and to stick with it regardless of any confirmation from the data. Hence the prediction from so many, including Bank of England Governor Mark Carney, of a “technical recession” after the referendum vote, andthe continuing forcasts that the slowdown will still come in the medium term. There is merely an issue of timing they argue, as businesses will ultimately delay investment plans due to the uncertainty regarding the trade deals which will be struck by the government, and households will tighten their belts as weak sterling starts to push up prices. If economists keep predicting a slowdown they will surely be right eventually, but they risk being regarded as nothing better than a stopped watch, which will be right twice a day. There is a chance that they are wrong yet again; sterling may be having a bad run, but slow growth in Europe and lower-for-longer European rates could well mean that sterling does pretty well against the Euro, and that inflation fears prove unfounded. In the second half of last year UK growth was the strongest of any G7 nation based on last week’s PMI data. However, there may be something much more than economists having a bad run, and some argue that the whole process of economic forecasting is fundamentally flawed.
The Bank of England Chief Economist, Andy Haldane, last week described economics as a profession in crisis, with the failure to predict the 2008 financial crisis its “Michael Fish” moment (Mr Fish famously dismissed talk of a hurricane in 1987). Predictions at the end of 2007 were almost universally constructive, and this year economists got the impact of the Brexit vote horribly wrong – at least in the short-term. One of the major issues seems to be that economic forecasting has become based almost solely on mathematics. Indeed, this is an issue for the subject as a whole; students now need ‘A ‘ levels in both maths and further maths to study the discipline at our top universities. Pumping numbers into a model and turning a handle ignores the key impact that the behaviour of businesses and households has on the process, much of which will be regarded as irrational by economists. It is very hard to introduce irrational factors into a model. Andy Haldane suggests that current economic models are “fragile” and suffer from a methodological monoculture” that inhibits their ability to make sense of irrational behavior. He has encouraged a move towards “agency based” modeling that takes account of the environment and the behaviour of the agents that interact with it. If the models used in 2007 had factored in the importance, and vulnerability, of financial institutions their predictions might have been very different.
Similarly most models adopt a steady-state approach for factors such as productivity – expecting long-term trends to continue. One of the most difficult factors to explain in recent years has been the very low level of productivity growth in the UK, and this has undermined many a forecast. One feasible explanation has been that, because of the prevalence of zero hours contracts, many employers have maintained a higher labour force than they would otherwise have done because there are minimal cost implications.
Perhaps the over-riding problem with the approach of modern economists is that they are too aware of each other and get drawn into a kind of ‘group think’. On a visit to LSE the Queen famously asked why no-one saw the 2008 crisis coming and it seems that there is a tendency to look at official forecasts, decide whether you think they are too high or too low and then set your own levels slightly differently. If everyone is broadly wrong together they can all use the same excuses. Some economists go to the other extreme of trying to ensure a high media profile by being extreme. Being different for the sake of it has little intellectual justification and there is a risk of becoming a propagandist rather than a credible economist.
There is a certain irony in all of this navel gazing and self-examination by the economics profession, and that is that they actually had a good track record for 2016 as a whole. If one takes consensus forecasts for the major UK economic variables issued at the end of last year, there was remarkably little deviation against the actual out-turn, with growth a little lower than expected and inflation slightly higher. Some of this is clearly a happy coincidence, given that the timing of the referendum was unknown at the start of the year and even when it was announced, economists expected a remain vote to spark a bounce which would compensate for pre-referendum uncertainty. What they got wrong of course was their forecasts of a sharp slowdown on leave vote – if they had just kept quiet, their initial forecasts were pretty good.
It seems then that we may start to regard economics as nothing more than a science of wishful thinking, but in doing so there is a danger that we start to question the credibility of independent central banks and the pillars by which they operate. If we see little likelihood of official economists being right, why should we be allowing them to set interest rate policy? In this sense there would be an argument for putting the onus of interest rate setting back to the government and having the process more aligned to fiscal policy. To do so would be to return to the dark ages. Economics is not a science and the more positive approach would be to try to incorporate some of the suggestions of the likes of Andy Haldane into a more behavourially based modeling approach. If in doing so economists develop a more successful track record they may just regain some credibility.
If you enjoyed this article please share it using the buttons above.
Please click here if you would like a weekly email on publication of the ShawSheet
