Probably right, definitely not accurate. The Treasury on the impact on Brexit

Probably right, definitely not accurate. The Treasury on the impact on Brexit

Economics is an art masquerading as a science. Always has been, always will be.

An economic model looks complicated, full of factors and equations, but it is invariably balanced on a bunch of assumptions that are, if not plucked out of the air, then at the very least guesstimates. Art plus science equals economics.

The Treasury’s analysis of the impact of Brexit is a case in point. It is a well-researched, chunky piece of work. It has some of the best inverse gravity modelling known to man. It has many equations which look impenetrable and intimidating. It is shrouded in a cloak of science.


And yet underneath its surface there are a whole load of assumptions and guesstimates.


To understand how and why, consider how those big numbers above (for instance, the 6.2% hit to UK economic growth, £4,300 for each household, by 2030) were forged. It was a long-winded six-step process. By way of explanation, allow me to paraphrase the bottled Treasury argument.

  1. First, trade matters, really matters, to the economy: UK trade flows are now over 60% of gross domestic product. So any fall in trade would naturally impact UK economic growth. This is pretty uncontroversial.
  2. Under any of the three Brexit scenarios they’ve looked at (the UK becoming a. more like Norway, b. more like Canada, or c. more like Brazil/Russia in terms of its economic relationship with the EU) trade and foreign direct investment would fall. This is an assumption. A reasonable one, but an assumption all the same.
  3. If trade falls, so does productivity and with it the performance of the wider economy. Again, this is a reasonable assumption, but there is a question mark over how you model the extent of the impact. The Treasury used a load of academic research based on previous episodes of trade trauma, including the Suez crisis of the 1950s, when the canal was briefly closed, to model this.
  4. Then they plug that resulting number into a broader model of the world economy, supplied by the National Institute for Economic and Social Research.
  5. After that they lop an extra bit of growth off the UK economy to account for what they call the “persistence effect” – the permanent chunk of growth that goes as a result of business firms putting off investment, households cutting spending, because of the fear and shock of the departure. The “persistence effect” means in each of the scenarios modelled in the HMT paper the economy is 1% weaker by 2030. Quite why the effect is the same in every scenario is unclear. It is yet another guesstimate.
  6. Once the Treasury has this growth number, it puts it into its public finance models and calculates the impact on tax revenues and spending. Unsurprisingly, because the economy will be weaker, so will the public finances (to the tune of £36bn in the Canada scenario). Provided you believe the headline numbers, this is totally uncontroversial. What is a little more odd is that in each case, the Treasury has lopped off £7bn to account for the fact that the UK would no longer have to contribute to the EU budget. This is oddly imprecise, given that these scenarios imply totally different grades of relationship with the EU.

You get the idea. There is plenty of detailed, forensic analysis throughout this report. But there are also a fair few unscientific assumptions thrown in to boot.

There’s the presumption that migration won’t be affected at all in any of the scenarios. There’s the fact that the report gives short shrift to the idea that regulations might be cut if the UK leaves. There are those slightly unscientific touches, like assuming the change in contributions to the EU is the same in every scenario, and lopping a seemingly arbitrary extra 1% off the growth impact for each scenario as a result of the “persistence effect”.

Now, economists would reason that all of the above is simply what happens in economic exercises like this. The more specific you get in each case, the more hypothetical your modelling looks, and the easier it is for critics to pick apart, so perhaps better to apply the same numbers and assumptions to each scenario.

However, it underlines that this is not a purely scientific exercise. That £4,300 figure is the product of a cocktail of equations and assumptions, some precise, some unexpectedly crude.

But ‘twas ever thus. Any economic assessment of the impact of some unknown future event will have to incorporate many imponderables. And in broad terms, the Treasury has been relatively conservative. Its forecasts are more pessimistic than those from Oxford Economics, but less so than those from the London School of Economics. It has attempted to measure not just the impact on trade but also on the broader economy (a dynamic rather than a static projection). You can quibble to your heart’s content, but at least the Treasury has been open enough about its assumptions.

And if one takes the mass of evidence from economic analysts over the past few months, one overarching lesson is emerging from the statistical clouds. In most scenarios, the UK would probably be worse off in the event of Brexit. Whether that loss is a price worth paying for extra sovereignty is something the Leave camp must  confront. It seems increasingly difficult to argue that people would be better off in the event of the UK’s departure.

But to suppose that the Treasury is able to put a two-decimal place precision on the impact of departure is frankly a little silly. Not that that will stop them.