Treasury’s Brexit analysis: what it says — and what it doesn’t
The Treasury’s long-term economic analysis of a British exit from the EU arrives at a central estimate that it would ultimately lower UK economic output 6.2 per cent, amounting to an annual cost to British households of £4,300 a year.
This central estimate is based on securing a free-trade-in-goods relationship with the EU such as the one that has almost been agreed with Canada. A closer, Norwegian-style relationship would impart a 3.8 per cent hit to gross domestic product while a looser trade relationship under World Trade Organisation rules would lower GDP by 7.5 per cent.
Each of these estimates has a margin for error around it, suggesting reasonable Treasury confidence that the estimates of the economic damage caused by leaving the EU would range from 3.4 per cent to 9.5 per cent.
Would households really be £4,300 a year worse off?
The Treasury is not saying anyone would be £4,300 worse off than they are today. This is not an economic forecast for 2030; it is a comparison of two possible futures that contain two possible trading relationships — but one of them is more prosperous than the other.
What that figure says, therefore is that at today’s prices, households would be £4,300 worse off a year by 2030 if Britain secured a Canada-style trading relationship than if it remained an EU member.
Dividing the hit to GDP by the number of households to get £4,300 is not strictly the hit to household incomes. It is, however, a standard way economists attempt to translate complex numbers into something understandable.
Once the effects of leaving the EU are fully in place, the Treasury does not assume growth rates would be different. This is an estimate of the change in the level of gross domestic product, not of GDP growth.
Isn't a forecast for 2030 absurd, since economic forecasts for even two years ahead fail?
This is to misunderstand the exercise the Treasury has carried out. As stated, the Treasury is estimating the difference between two possible futures on the basis of a decision regarding Britain’s trading relationship. That does not require accurate forecasts, just a good understanding of the effect of trade on prosperity.
It is the same as saying, “We do not know how heavy you will be in 15 years, but if you drink a bottle of cola a day, we are pretty sure you will be fatter than if you keep off the sugar.”
Alternatively: “We do not know when you will die, but are pretty sure it will be earlier if you maintain your 20-a-day cigarette habit.”
How does the Treasury reach its conclusions?
It uses a three-stage approach. First, it uses gravity models to estimate the effect of different trade relationships on the quantity of trade and foreign direct investment. Gravity models take into account how close countries are to each other geographically, as well as their historic links, rather than assuming trade flows to wherever the lowest tariffs are.
Second, it uses external academic results to estimate the consequences for productivity — the efficiency of the UK economy — from different levels of trade and foreign direct investment.
Third, it plugs the productivity numbers into a global economic model run by the National Institute of Economic and Social Research to estimate the long-run differences in national income and prosperity.
UK’s EU Referendum: How people would vote
For a more detailed summary of opinion polling visit the FT’s Brexit poll tracker page
Are the Treasury’s numbers realistic?
Realistic is a difficult word to apply to economic models. The Treasury has decided against using one model to give all the answers because these models — called structural models, or computable general equilibrium models — cannot estimate many of the important and likely consequences of a British exit.
Such models tend to give weaker results, so one can certainly say the Treasury’s modelling strategy is more comprehensive. A better word to describe its approach would be reasonable, rather than realistic.
What are the model’s main weaknesses?
Uncertainty. While the understanding of the importance of trade is as good as the medical profession’s understanding of the harms of smoking, precision is difficult.
A more technical valid criticism is that the three models that the Treasury uses are not estimated simultaneously, so the end results are prone to inconsistencies. It is more tenuous to prove causality of the relationships in a series of models.
In addition, the Treasury assumes migration and regulation will not affect the calculations. It is not clear whether including these effects would have increased or decreased the overall hit to GDP. The majority of Britain’s most onerous regulations — for example, planning restrictions — have nothing to do with the EU.
What are the main strengths of the analysis?
The Treasury’s model seeks to include everything. It links known effects of trade to a holistic assessment of the effects on output and prosperity. Narrower models might have less uncertainty of estimates inside the modelled world, but only because they miss out the most important aspects. It was the use of narrow models — excluding a banking sector, for example — that led some economists to be overly complacent about the world’s economic strength before the financial crisis.
The Treasury estimates that the effects of a British exit would be worse than under models produced by the CBI employers’ organisation, Oxford Economics and Open Europe because its model is more comprehensive.
Its results are slightly more optimistic than results from the London School of Economics (which adopts the same modelling strategy) because it assumes that weaker EU trade has a smaller negative effect on productivity than the LSE model. In that sense, the Treasury is adding caution to its results.
So is it being overcautious?
Officials have missed out one potentially big negative consequence of an exit: the shock of leaving. It assumes that by 2030 the short-term consequences for confidence and spending, apart from a small hit to investment, will have been regained.
By contrast, the 2007-08 financial crisis has left a permanent hangover. Were Brexit to create a short-term shock with lasting effects, the financial hit would be greater than the Treasury expects. However, that would have involved adding a fourth model to the sequence.
Would the public finances benefit from leaving the EU?
No. What the Treasury’s analysis makes perfectly clear is that any lasting economic damage is far more important for the public finances than Britain's annual net contribution to the EU budget.
On very conservative assumptions, the Treasury estimates that each 1 per cent hit to GDP lowers revenues by roughly £6bn a year — a little less than Britain’s net contribution to the EU.
The central estimate of a 6 per cent cost to GDP lowers revenues by £36bn a year, far more than the annual net contribution. Again, the numbers are uncertain, but models are useful for giving a sense of scale. It shows that fiscal savings from Brexit are likely to be a false economy.
But the Treasury supported joining the euro — why should we listen now?
No, it didn’t. The lead official on this EU exercise is Dave Ramsden, the Treasury’s chief economist. As it happens, Sir Dave was also the lead official on the 2003 Treasury assessment of the economics of joining the euro. That analysis actually opposed euro membership even though some in government, including Tony Blair, then the prime minister, supported adopting the single currency.
Brexit? In or Out
What a British divorce from the EU would look like
How any break-up is carried out will have a huge impact on Britain for generations
The economic consequences of Brexit
Three very different outcomes of a British vote to leave the EU
What would Brexit mean for the City of London?
There is a clear split over how a vote to leave would shape the capital’s future as a financial centre
What the City stands to lose and gain from Brexit
Sectors such as foreign exchange trading have boomed during EU years
What has the EU done for the UK?
The long-running debate over the economic benefits of membership remains unresolved