Bank of England governor Mark Carney’s Brexit report is interesting
mainly because of the methodology it employs.
The Bank of England’s report about the impact on the UK economy of leaving the EU issued on 28 November has attracted huge attention, but mainly for the wrong reasons.
The headlines it’s generated are indeed compelling.
My favourite was in Scotland’s The Herald which reads thus: “Bank of England warns that crashing out of EU would lead to worst slump since World War II.”
For economists, the report’s value is in exposing the methodology Bank of England economists used to produce it.
Problem one: Its projections are not forecasts. They are scenarios.
This is an approach made famous by Royal Dutch Shell and involves imagining two or more alternative futures. These are then used to develop data-based projections for each.
Scenario projections are designed to aid discussion and shape decision-making.
They are very useful. But they are also artificial.
The methodology is based on the idea that there are definable alternative futures and that each is discrete.
But we know the future is infinitely confusing. Who three years ago would have thought Donald Trump would become US president?
A disruptive no deal scenario leads to lower growth and higher inflation and unemployment. And yes, it could be the worst economic setback since 1945.
But it’s one that no rational person would choose. It’s a dummy scenario. Or perhaps a scenario for dummies.
The choice of scenarios is critical.
If you want to manipulate a decision you offer several scenarios that are plainly unlikely or unpalatable. The cunning narrows the real choice to two.
This is the case with the report, which offers four, including a “disruptive” scenario where there is a no-deal Brexit on 29 March, it’s unexpected and the EU takes immediate punitive actions.
It’s this scenario which attracted, at least initially, the most attention and for obvious reasons.
Scenario planning encourages people to travel in their minds to the future and in their imaginations compare where they are with other future situations.
That’s enormously powerful, and fun.
But it tends to discourage thinking about comparing where we are now with where we will be. That involves bringing the future to the present and is more in line with conventional forecasting. This too can be both powerful and misleading.
The Bank of England’s projection shows there will be a large gap between two of its imagined futures. One (which is already impossible) is where the UK goes back to June 2016, votes to Remain and then travels on to 2024. The other is the unlikely, disorderly Brexit.
But its figures also suggest the UK economy in real terms in the Doomsday scenario will actually be not much smaller than it is now. The disorderly scenario projection is scary. The forecast that it implies is merely depressing.
In all of the three other scenarios – which range from disruptive to two forms of close partnership with the EU – the UK economy in 2024 will be smaller than it would have been if there’d been a Remain vote in 2016, but not by that much.
Problem two: The starting point.
The Bank of England’s is May 2016, the month before the UK referendum on EU membership.
If it had used the autumn of 2018 as its start point, its projection would show that growth under the two close partnership Brexit scenarios will be higher between now and 2024 than it would have been in the 2016 projection. That’s because there will be a robust rebound in growth under the close partnership scenarios from 2019.
This brings you to a further issue.
Problem three: You’ll never know what might have been.
It is possible that that the UK economy would have gone the way the Bank of England expected after 2016 if there’d been a Remain vote. But it might not have done. No one knows.
Problem four: The end date of the projection.
Conventionally, scenario-based forecasts deal with the long-term future. The Bank’s report covers a comparatively short five-year period. That’s because even now the possible variations of Brexit are numerous. But what would a 10-year projection shown if it had been attempted?
Problem five: Compounding the impact.
Its projection of the impact on UK GDP in a close economic partnership arrangement, presumably the one closest to the deal UK Prime Minister has agreed with the EU, is that it will be at least 1.25 per cent lower in 2024 than it would have been if trends seen before May 2016 had continued.
That’s an annual rate of 0.2 per cent lower than the trend over an eight-year period.
For many, this is well within the range of possible statistical error.
Problem six: Quantifying the effect of tariff and non-tariff barriers.
How has the Bank of England incorporated key factors into its econometric model?
The easiest thing to deal with is tariffs.
The average EU tariff is 5.7 per cent. You can therefore use this in forecasts by eliminating that from trade between the UK and non-EU nations and adding it to trade between the EU and the UK for the period between March 2019 and 2024.
But the tariff rates the UK might apply to imports from EU and non-EU countries will depend upon whether it opts for a WTO system, a temporary no-tariff deal with the EU or something else.
That’s not yet fixed.
More challenging is the likely response in demand for EU imports and non-EU imports and UK exports to EU and non-EU nations after Brexit and the introduction of new tariff system. That’s the elasticity of demand and supply for imports and exports classified according to whether the goods are made in the EU or not.
The best anyone can do here is provide an estimate based on historic trends.
The Bank has used a gravity-based analysis to project trends in UK-EU trade. That centres on the idea that large economies trade more with each other as do countries that are geographically close.
You’re more likely to trade with a big economy that’s nearby. For the UK, that’s the EU.
This approach makes sense and is empirically tested but still falls short of providing complete certainty.
And then of course, there’s the Bank of England’s entire model of what shapes imports, exports, consumption, saving, investment, productivity, foreign investment, foreign exchange flows and a host of other factors. It’s fiendishly complex and technical.
And there can always be a better one, as every non-conventional economist in earth agrees.
More substantial is getting to grip with non-tariff barriers: ways that countries encourage or discourage imports using administrative measures.
The Bank’s report shows that EU non-tariff barriers, including standards, are the equivalent to a tariff of around 9 per cent. In other words, non-tariff barriers are a bigger factor in trading with the EU than tariffs.
This is a critical issue for the report. EU tariffs are comparatively low on average. The trade-weighted mean is around 3 per cent.
That would lead to the conclusion that a country outside the EU would find tariffs a modest discouragement to trade with EU nations. Put simply, being in or outside the EU on balance has limited impact on trade flows on average.
But non-tariff barriers could seriously jeopardise trade. In some circumstances, goods could be barred from EU markets.
It makes sense to include this in any projection of the UK economy outside the EU.
What’s more debatable is the Bank’s attempt to quantify the impact. It’s guesswork and raises questions about the validity of its projections.
Problem seven: Services.
The projection becomes even more unhinged once services are considered.
By definition, services are indefinable and it’s impossible to locate exactly where they are created. This makes any forecast of the impact of leaving the EU on trade in services between the UK and the EU almost impossible.
At present, despite the Single Market’s call for complete regulation of all services, most are unregulated at the EU level. How a shop operates in Budapest depends upon rules in Hungary and the behaviour of shop owners and shoppers in that country.
On top of that, cross-border trade in services in advanced economies tends to be lower as a proportion of GDP than it does in tangible goods.
Plenty of people use electronic equipment and wear clothes made outside the UK . But how many cross borders for education or healthcare services? Almost no one.
In the UK, more than 90 per cent work in service industries. They are affected by EU labour and other regulations. But trade with non-UK customers is modest, except in banking and finance. In other words, most people aren’t affected much by EU trade rules.
There’s no doubt that restrictions on EU people working in the UK could have a serious impact in low-skill farm work and the NHS.
And nobody wants long passport queues when they go on holiday to EU countries.
But leaving the EU won’t have a direct impact on most people working in the UK.
The Bank of England has incorporated in its model projections of increasing restrictions on service providers supplying EU customers once the UK’s outside the EU. Financial services could be significantly affected. This is reasonable, but the impact of EU restrictions on non-finance service trade is a huge open question.
The idea of regulating every EU service sector in a dynamic and effective way is fanciful. The reality is that it’s beyond the EU’s capacity and won’t be attempted much outside finance.
Problem eight: Dynamics.
The outcome of Brexit will depend upon how people, business and government react. If the consensus is that Brexit is a disaster, it’s likely the economy will perform less well than if the majority believe it’s a good thing.
How the EU and its 27 members will react could be crucial. And then there’s the way the UK deals with the rest of the world.
There are no easy answers there.
So, on balance, the Bank of England’s report should be read with caution. It could to be right, though it’s impossible empirically to test scenarios ex post. Nobody can say what might have been.
There is no doubt that the no-deal Brexit scenario is unpalatable, but only because the assumptions it’s built on are both extreme and unlikely.
More interesting are the close partnership projections. These suggest the impact on the UK will be quite modest. It’s conceivable, but not for the Bank of England apparently, that it could be positive. It all depends how the UK, the EU and the rest of the world react to Brexit.
That’s common sense.
And despite the emotion Brexit stimulates, the reality is that the EU is neither as beneficially powerful nor as destructively malign as partisans suggest.
That’s the boring truth buried inside the Bank of England report.