Region and district-level Brexit-vote cost estimates
Since the Brexit-vote on 23. June 2016, the overall UK economy has grown at a noticeably lower rate compared to before the EU referendum. There are several methods used to estimate the counterfactual capturing how the UK economy would have grown, had it not been for the UK’s vote to Leave the EU. Descriptive analysis suggests that since the EU referendum, Britain has fallen from the top to the bottom of the league of G7 leading economies in the year since the Brexit vote, highlighting significantly slower growth. Other analysis by academic economists suggest that the UK’s economy is at least 2% smaller, owing to the Brexit-vote. This project sets out to study the geographic breakdown of the cost of the Brexit-vote across UK regions and districts.
The visualizations presented below are based on the research paper of Fetzer and Wang (2020), Measuring the Regional Economic Costs of Brexit: Evidence up to 2019. This working paper describes in detail the synthetic control method as an econometric tool is applied to subnational economic activity data from the UK and other countries to both, estimate the overall economic impact of the Brexit-vote to date and to study how this cost is distributed across regions- and districts in the UK.Tweet
Synthetic control estimate
The gross-value added (GDP) in Scotland appears to be 2.82 percentage points lower relative to a statistically constructed synthetic control that best captures how GDP would have evolved had it not for the UK's vote to Leave the European Union. The estimate suggests that the Brexit-vote has cost Scotland 's economy 3954.89 million pounds in 2019. This amounts to a loss of economic activity of about -736.07 pounds per resident in 2019.
Region-level Brexit-vote cost estimates
Below table presents the estimate of the average estimated economic cost of the Brexit vote as of 2019 across UK regions. The numbers are expressed in absolute terms millions of pounds, in relative terms and in per capita terms. The headline estimate is the Ensemble estimator that is presented visually in the above region-by-region graphs.
|Region||Brexit-vote cost Ensemble estimate||Other estimates in £million|
|NUTS1||Name||£million||in %||£ per capita||Lower||Upper||Permutations||AAPE||RMSE||MAPE|
|UKE||Yorkshire and The Humber||-1932.41||-1.59||-358.50||-2580.29||-1287.94||-1903.18||-3152.68||-2830.70||-3436.57|
Synthetic control estimate
Lewisham appears to be a Brexit-vote Loser. In 2016, 30.14 percent of Lewisham's electorate voted to Leave the European Union. In 2018, Lewisham's gross-value added (GDP), appears to be 14.9 percentage points lower relative to a statistically constructed synthetic control that best captures how GDP would have evolved had it not for the UK's vote to Leave the European Union. The estimate suggests that the Brexit-vote has cost Lewisham 's economy -578.05 million pounds in 2018. This amounts to a loss of economic activity of about -1904.4 pounds per resident in 2018.
Results from Placebo tests
These placebo tests allocate “fake” Brexit-votes to districts or spatial units in the donor pool, estimating Brexit-vote effects fo each of them. The result for the UK district that was actually affected by the Brexit vote is presented as the red line. The density plot highlights the distribution of the 2018 value for the UK district relative to the distribution of “fake Brexit” effects in the donor pool.
Results from Donorpool Permutation tests
The results below highlight whether the path identified for the UK’s synthetic control is potentially biased or driven by the size of the donor pool to draw from. The figure presents the synthetic control paths that are constructed using 70 different donor pools with fixed sizes with donors being sampled at random.