The Brexit policy shock: Were UK services exports affected, and when?

Abstract

We analyse the impact of the Brexit announcement shock on UK exports of commercial services, using a synthetic control method (SCM) to create a counterfactual based upon other countries’ exports. Our analysis considers two alternative scenarios: in the first, the shock was based upon the referendum date of 23 June 2016. We do indeed find significant evidence of a treatment shock based upon this date, as exports after this date are approximately 7% below the synthetic counterfactual. This is driven by ‘other commercial services’: the smaller tourist sector actually had a positive shock. If we repeat the analysis with an earlier treatment, it becomes clear that divergence was already starting in the first half of 2015, around the time of the general election campaign, although services exports had a clearer shock after June 2016, roughly doubling the discrepancy. This finding supports the suggestion that there was already some fear of Brexit before the referendum result, and indicates the importance of considering an earlier start date in studies of the Brexit shock.