The debate about economic forecasts

The statement by the Bank of England that its forecasts of the effects of Brexit have proved wrong has been greeted with glee by Brexiters and led to economic forecasting in general being denounced. There is no doubt that economic forecasting is a problematic activity, but it is important to make several points about the Brexit forecasts.

First, they were an attempt to respond to the Leave campaign’s claims that the UK would flourish or at least come to no harm if people voted to leave. Those were every bit as much forecasts as those of the Remain campaign, even though they were not backed up with any analysis. Similarly, claims about the £350M a week for the NHS if we left, the fall in immigration if we left, and the prospect of Turkey and other countries joining the EU if we didn’t leave were all forecasts, and they have all now been disowned by leading leavers.

Second, like all economic forecasts, those made during the Referendum campaign were based upon assumptions and simplifications. Most importantly, many of them, including that made by the Treasury, assumed what was then believed to be inevitable namely that Article 50 would be triggered immediately after a vote to leave (see paragraph 1.42 of the Treasury short-term forecast). It wasn’t, and instead we have been in a limbo since. So far as the BoE forecast is concerned, it did not (bizarrely, perhaps) factor in the actions that it, itself, would take if the vote was to leave: quantitative easing and a cut in interest rates. Both those things happened, and mitigated the effects of the vote, but they were not cost-free and the price of them is being paid by pension funds and savers. So whilst some of the adverse effects the BoE predicted have not so far occurred, it is because of other adverse effects created by the BoE’s attempts to prevent their predictions coming true: the cost of the vote has been displaced and deferred, not avoided.

Third, it is simply not the case that the vote has passed without very severe economic consequences. The rapid fall of sterling is the most obvious, and would in any other circumstances have been understood as an economic catastrophe. The knock on effects of inflation are beginning to be felt. And there have been significant deferrals or abandonments of investment. Most economists continue to believe that the long-term effects of Brexit will be bad. At the very least, the jury is still out and some things can be said with certainty: whenever the government indicates that a hard Brexit is in prospect, the pound falls; and the government’s fiscal position has deteriorated to the extent of £59 billion over five years as a result of the vote. This isn’t just ‘forecasting’, it means real additional cuts in public services, which will affect real people’s lives, which wouldn’t have occurred if the Brexit vote hadn’t happened. Again, in any other circumstances this huge collapse in public finances would have been seen as a major economic crisis. Post-Brexit it is all but ignored.

Fourth, and it is a version of the first point, although Brexiters are now decrying economic forecasting, they are happy enough to take as gospel those that have recently been provided by lobby group Change Britain saying that hard Brexit would make the UK £450M a year better off and create 400,000 new jobs. If forecasting is so discredited then why should we believe these figures?

Similarly, much Brexiter attention has been given to a report by a group of Cambridge economists saying that the official forecasts were wrong and that there would be little economic effect from Brexit. Seldom can an academic working paper have been accorded such respect! What is striking about that is, first, that we can be sure that had reached different conclusions it would have been derided as the work of the elite (or ignored altogether) and, second, that what is being lauded is the application of a new and as yet unproven model of economic forecasting. In short, there is much confirmation bias in play here.

With all that said, I have always been reluctant to try to quantify the effects of Brexit and in various talks I gave before the Referendum refused to do so. I have adopted the same approach on this blog (for example on last September’s OECD statement, and in the very first post I said that there was no point in following each and every economic indicator whilst we were still in the EU).  Instead, what should command out attention is the logic of different institutional arrangements. What I mean by that is that leaving the EU is a massive institutional change for the UK. That ought be to common ground between Brexiters (or else – why leave?) and Remainers (or else – why care that we are leaving?).

If that is so, then [given that 50% of our trade is with the EU and another 16% via EU trade deals – so – 66% of our trade – and given that some, at least of our foreign direct investment relates to single market membership, and given that many of our industries from agricultural harvesting to the most advanced science are dependent on free movement] how could it be anything other than true that it will have massive economic effects? No country has ever tried to simultaneously detach and re-attach itself to the global economy, with no clear process or timescale or terms for doing so, and that in itself makes predictions all but impossible, except to say that the consequences will be hugely destabilizing. For sure Brexiters might say, as some do, that any price is worth it. For them, no economic forecast matters. Or that in the long run the effects will be positive - and what is that other than an economic forecast, albeit based not on data modelling but on blind faith?