Janez Kren and Martina Lawless (Economic and Social Research Institute, Ireland)
The UK exited the EU Single Market and Customs Union on January 1, 2021, arguably the largest incidence of trade disintegration in over fifty years. Between the referendum and exit date, a wide literature on the potential impacts of Brexit emerged covering many potential scenarios depending on the degree of continued trade links, almost all suggesting trade was likely to decline. Despite these negative prior expectations, research by Freeman, Manova, Prayer and Sampson (2022) found that the first year of Brexit showed a positive impact on UK to EU trade in their analysis using UK data. However, if we apply a similar approach using EU trade data, we find a one-third decline in trade from the UK to the EU and significant but smaller reductions in trade from the EU to the UK. So, how did Brexit really impact trade?
In our research we identify two key issues contributing to the contrast in estimates. First, an important question in the estimation of how much Brexit affected EU-UK trade is “compared to what?” Second, Brexit also changed some reporting requirements leading to breaks in the measurement of trade flows, particularly in the EU data.
To resolve these issues and get the most accurate possible estimate of the impact of Brexit, we combine UK-reported data for its trade with the EU (to minimise data breaks) and EU data for its trade with the rest of the world (as the more appropriate benchmark for comparison). From this best of both worlds “hybrid” dataset, we find that Brexit reduced UK to EU trade by 16% and trade from the EU to UK by 20%.
The importance of a comparison group
When considering Brexit as a “treatment” applied to EU-UK trade, the choice of the “untreated” control group is an important factor in the estimation process. Freeman et al. (2022) used the UK’s exports to the rest of the world as a comparison group. These grew slowly in this period and may have been impacted by Brexit spillover effects. Therefore, they are not particularly suitable as a no-Brexit counterfactual. EU trade with the rest of the world was less impacted by Brexit and hence provides a cleaner benchmark.
Data collection changes
Unhelpfully, the statistical agencies in both the UK and EU made changes to data collection when the UK left the EU Single Market. In both cases, this arose because the UK’s exit from the EU Single Market also resulted in it leaving the EU’s internal trade reporting system (Intrastat). The UK’s exit from this system meant trade flows from the EU side moved to being collected through the Extrastat system used for non-EU member states. The Extrastat system applies a lower threshold for trade declarations on non-EU imports than the Intrastat system applies on internal EU trade. More importantly, import trade within the EU is collected on a country of consignment basis but non-EU imports are defined based on their origin. There were also some changes in the mode of data collection in the UK but without the change in definition applied on the EU side.
A “hybrid” solution
If both the choice of comparator group and data differences can impact estimates of the impact of Brexit on trade, how do we answer the key question of how large the shock to trade really was? We believe the most accurate measure comes from combining the data sources. We take the HMRC data as the better choice for UK-EU trade flows, given that they appear to have been subject to less methodological change than the Eurostat data. As discussed above, we believe there is a strong case that EU trade with the rest of the world is the preferable comparison group so we use Eurostat data for EU flows with the rest of the world. Using this combined data, we find a substantial negative impact of Brexit for trade in both directions, with UK to EU trade reduced by 16% and EU to UK trade reduced by 20%.
While our results highlight the importance of the appropriate comparison group in assessing the impact of Brexit, this leaves open for future research the question of whether the patterns of relatively slow UK export growth to the rest of the world and more rapid import growth can themselves be attributed to Brexit.
This blog is a summary of Janez Krena and Martina Lawless, “How has Brexit changed EU-UK trade flows?” ESRI Working Paper No. 735 (October 2022). The paper can be downloaded here.
 Measuring imports on a consignment basis means that the import source is recorded as being from the last country where a change was made to the goods, whether that was a change in ownership, repackaging or further processing. Origin basis attributes imports to the country in which the goods originate.
The views expressed in this blog reflect the position of the author and not necessarily that of the Brexit Institute Blog.