One of the most contested issues, before and since the referendum on UK membership of the EU, has been the potential impact of Brexit on the UK economy. The exercise is almost as difficult now as it was before the referendum, because we still don’t know what Brexit will mean for the UK’s trading relationships, or the regulatory environment in Britain, two issues that will have a significant bearing on the long-term economic consequences.
Almost 18 months on from the referendum, we are, however, in a better position to scale the medium-term consequences. One approach is to compare what independent economic forecasters were saying about the prospects for the UK economy in the first half of 2016, shortly before the referendum, with what they are saying now. Medium-term economic forecasts are typically made about five years ahead, allowing us to compare forecasts for the years 2016 to 2020.
It is more likely than not that nothing much will have changed in the UK’s relationship with the EU by the end of 2020. The UK will only leave the EU in March 2019. The UK and the EU both appear to want a transition period afterwards of around two years, taking us through to the beginning of 2021. And both sides say this should largely retain the rights and obligations of EU membership. By focusing on the period 2016 to 2020, we are therefore able to scale the effect of Brexit on expectations of future changes, before those changes even come into effect.
It’s important to understand the international context, before looking at what independent forecasters are saying about the UK economy. Figure 1 shows how the IMF’s average growth forecast for 2016 to 2020 changed from its pre-referendum estimate (in April 2016) to its most recent (in October 2017). For the world economy, not much has changed. But the aggregate picture masks some country variation. Among the G7, the outlook for Japan and Canada is a little better, while for the US, it is a little worse. Perhaps most significantly, the outlook for the euro area has also improved, with forecast growth up by 0.3pp on average each year.
Figure 1: percentage point change between April 2016 and October 2017 in the average annual growth forecast for 2016 to 2020
How have forecasts for the UK economy changed over this period? For this, we don’t just use the IMF forecast, but instead compare the average for the panel of more than 20 independent forecasters that the UK Treasury uses to inform its policy decisions. This panel includes forecasters in the City of London and other non-City forecasters, some of which, like the IMF, are based outside the UK.
Figure 2 shows how the average UK growth forecast for 2016 to 2020 changed between February 2016 to November 2017. The average forecast is for growth that is on average 0.6pp lower per annum, with a cumulative growth rate over the five-year period that is 3pp lower. This also turns out to be similar to the adjustment made by the independent Office of Budget Responsibility to its forecasts between March 2016 and November 2017, which is down by a cumulative 3.2pp over the five-year period.
Are these figures significant? The nominal value of the UK economy was £1.89 trillion in 2015, so this is roughly the equivalent of £60 billion in lost GDP by 2020, or around £2,200 on average for the UK’s 27 million households. With taxes at around 34% of GDP, this implies foregone receipts of around £20bn by 2020, which is not far short of the total amount spent on industry, agriculture and employment each year. These are not one-off costs, but a permanent loss in UK output and, by extension, potential tax revenue.
This calculation has its weaknesses. Plenty has changed since the first half of 2016 that has little to do with Brexit. But the prospects for global growth have not changed much, and the outlook for the rest of Europe is much better, which should boost the UK economy. So this figure – £60 billion by 2020 – is a reasonable estimate of the medium-term impact of Brexit, even before Brexit changes anything much at all.
Figure 2: average annual growth rate forecast for 2016 to 2020, by time of forecasts
The views expressed in this note can be attributed to the named author(s) only.