Brexit is the first case in which the unwinding of deep economic integration can be directly observed. The shock is most pronounced and most precisely dated in goods trade (Freeman, Manova, Prayer and Sampson, 2022). Drawing on UK customs data, these authors show that the UK's effective exit from the EU single market and customs union, once the Trade and Cooperation Agreement (TCA) entered into force, reduced British imports from the EU by 25% relative to the rest of the world. The decline is sudden and persistent. According to them, the effect on exports is, by contrast, more modest and temporary. Their analysis thus suggests that Brexit has primarily affected the UK's ability to source goods from the European Union, rather than its ability to sell goods into the EU market.
These results nonetheless depend on the benchmark chosen to measure Brexit's effects. Freeman et al. compare the evolution of EU–UK trade to that of UK trade with third countries using a difference-in-differences approach, which neutralises shocks common to all trading partners, such as the pandemic and the 2022 energy price surge. In other words, they ask whether trade with the European Union evolved differently from trade with the rest of the world after Brexit. An alternative approach consists in comparing the EU–UK trade flows actually observed against the trajectory they would most likely have followed in the absence of Brexit, extrapolating from pre-2016 trends. This second approach leads to a somewhat different interpretation of the data.
When EU–UK trade flows are compared against their pre-Brexit trajectory rather than against UK trade with the rest of the world, the picture appears more nuanced.
Until 2020, flows track their reference scenarios — that is, the level of trade that would have been expected had pre-Brexit trends continued unchanged. Indexed on their 2017–2019 average (=100), UK exports to the EU remain between 95 and 100. Then, with the end of the transition period and the entry into force of the TCA, they fall to 76 from 2021 onwards — nearly forty points below reference levels that continue to rise. The data show a rebound in 2022, to around 112, but an analysis of British export categories for that year reveals that this rebound stems primarily from the energy price spike. UK fuel exports surged by roughly 82% in value that year, driven by higher prices rather than greater volumes. The ONS volume series, which strip out inflation, thus provide a more accurate picture of the situation. Once the price effect is removed, the 2022 rebound disappears and reveals no genuine trade recovery. The series falls back to 82 in 2025 — 18% below the 2017–2019 average and nearly sixty points below reference levels that have by then reached around 140. Imports from the EU prove more resilient, reaching 108 in 2025, yet they remain roughly thirty points below their reference trajectory.
This divergence from Freeman et al. should not be read as a contradiction. The two approaches rest on different counterfactuals. Freeman et al. measure EU–UK trade performance relative to UK trade with the rest of the world, whereas Figure 1 compares observed flows to an extrapolation of pre-Brexit trends. Several additional factors further complicate the interpretation. The UK delayed until 2024–2025 the implementation of certain import controls on EU goods, which temporarily softened the impact of the new customs barriers on imports from the European Union. Furthermore, changes introduced by the ONS to its customs data collection methodology in January 2021 created a break in the statistical series.
The overall conclusion nonetheless remains unchanged. Regardless of the methodology employed or the counterfactual chosen, both approaches indicate that EU–UK goods trade remains significantly below its estimated non-Brexit trajectory. The debate is therefore less about whether a Brexit effect exists than about its precise magnitude and its distribution between imports and exports. More than five years after the end of the transition period, the available evidence suggests that the costs associated with the new customs formalities, regulatory divergence, and rules-of-origin requirements continue to weigh on trade flows. Far from fading over time, the gap between observed trade and its estimated counterfactual trajectory appears to be largely persistent.
The persistence of a trade deficit relative to the scenario in which Brexit had not occurred is nonetheless insufficient to account for the full range of observed changes. Aggregate trade data often imperfectly capture the underlying adjustments at work in the economy. Brexit may have altered the number of firms engaged in international trade, the variety of products exported, and the distribution of trade across firms. Examining these different margins of adjustment provides a better understanding of the microeconomic effects of the new trade relationship between the United Kingdom and the European Union.
Aggregate trade values do indeed conceal the most revealing adjustment: the decline in the number of firms exporting and the narrowing of the range of products traded. According to Freeman et al. (2022), Brexit reduced by approximately 30% the number of product-destination pairs exported to the EU each quarter, meaning that many products ceased to be exported to certain EU markets — primarily through the elimination of the smallest trade flows. The relative resilience of aggregate trade values should therefore not be interpreted as evidence that Brexit had only a limited effect on trade.
The disappearance of many small-scale trade flows is not a mere statistical phenomenon. Firm-level data confirm that Brexit altered the composition of EU–UK trade by affecting smaller firms disproportionately. Freeman et al. (2025) show that the largest exporters and importers largely weathered the shock, while smaller firms have struggled to maintain their activities in the European market. Our analysis of ONS data on trade by firm characteristics leads to the same finding: the number of micro-enterprises and small firms exporting to the European Union fell sharply after 2021, while large firms broadly maintained their presence in these markets. These results corroborate, using new data, the existence of a differentiated Brexit effect by firm size.
In total, UK goods exports to the world fell by 6.4% and imports by 3.1%, with buyers partially switching to non-EU suppliers (Freeman et al., 2025). But applying this filter of size-differentiated impact, the number of UK firms exporting to the EU declined by approximately 31% for micro-enterprises and 22% for small firms between 2019 and 2024, by 14% for medium-sized firms, while remaining stable for large firms (≈0%). The smallest firms account for the vast majority of exporters by number, but only a small share of export value; aggregate value thus held up even as the base of active exporters contracted sharply. Trade has therefore not so much contracted as concentrated among large firms. Crowley, Exton and Han (2018) had anticipated this mechanism: the mere threat of higher barriers deters exporters from entering and weakens those who remain, starting with the smallest.
Brexit thus appears to have weighed most heavily on the firms least able to absorb the fixed costs associated with new customs procedures, new rules-of-origin requirements, and new regulatory formalities. Taken together, these new procedures have increased the cost of EU–UK trade. While the TCA cushioned part of the shock by maintaining a tariff-free, quota-free goods trade regime, a standard free trade agreement such as the TCA can only partially offset the broader rise in trade costs resulting from non-tariff barriers (behind-the-border barriers).
Beyond firms, the effects of Brexit have also varied considerably across sectors. The nature and intensity of the shock depend on the barriers specific to each industry, as well as on rules-of-origin and regulatory requirements.
The sectoral breakdown does indeed reveal a highly uneven impact. Textiles and clothing have declined in both directions, by approximately 37% on the export side and 30% on the import side — a sector particularly exposed to rules-of-origin requirements. Motor vehicles and machinery display a form of asymmetry: UK imports have continued to grow by 7% while exports to the EU have fallen by 21%, a gap that points to the TCA's rules of origin, whose requirements are particularly onerous in the automotive sector. Food and agriculture presents the reverse picture: UK exports have declined by 3%, having been subject to EU sanitary and phytosanitary controls since 2021, while imports from the EU remain elevated and have increased by 28% — notably because the equivalent UK controls were only introduced gradually under the Border Target Operating Model (BTOM), with the first checks entering into force in January 2024 and the full regime being rolled out in stages throughout 2024.
Two categories call for particular caution in interpreting the data: precious stones, where gold and re-exports dominate, and energy, whose trade values are strongly influenced by price fluctuations. Taken in isolation, these indicators are of limited relevance for assessing the structural changes that have occurred in trade between the United Kingdom and the European Union.
The intra-EU reference scenario rests on the assumption that UK–EU trade would have grown at the same pace as intra-EU trade (+36.8%). This comparison point must nonetheless be interpreted with caution, as it is expressed in current values and thus conflates three distinct factors: the underlying growth trend of European trade, the inflationary surge of 2021–2023, and the Brexit-specific effect that the analysis is precisely seeking to isolate. A negative gap for a given sector therefore signals underperformance relative to this composite benchmark, and not a precise measure of the Brexit shock alone.
Beyond their effects on trade flows, the additional costs associated with the new post-Brexit trade relationship are ultimately borne, at least in part, by consumers. Assessing this impact is therefore an important component of any evaluation of Brexit. According to Bakker et al. (2022), the rise in non-tariff barriers pushed UK food prices up by 6% over two years — equivalent to £210 per household and £5.84 billion in total — with retail price pass-through ranging between 50% and 80%. The distributional consequences are also significant: the poorest decile experienced a cost-of-living increase roughly 52% higher than that of wealthier households.
This order of magnitude is consistent with estimates placing the cost of Brexit-related non-tariff barriers at approximately 8% on average (under the Trade and Cooperation Agreement), rising to 12–13% once customs formalities and rules of origin are factored in, with food products among the most affected sectors. Several elements suggest that Brexit played a decisive role in these price increases. The timing coincides closely with the end of the transition period in January 2021. No comparable effect is observed in trade with third countries, and the increases are concentrated in the product categories for which the UK is most dependent on imports from the EU — findings that cannot be explained by common macroeconomic shocks (Bakker et al. 2022). Post-pandemic inflation and sterling depreciation have of course contributed, the latter having its own effect on living standards (Breinlich et al. 2022). But neither of these explanations accounts for the fact that the price increase coincides precisely with the end of the transition period, nor that its most pronounced effects are concentrated in the sectors most exposed to the trade frictions generated by Brexit.
01 | Services: the earliest rupture
The picture appears different for services, which are particularly important for the United Kingdom given that they constitute the bulk of its external trade surplus. Here, the Brexit shock appears both earlier and, in certain respects, more pronounced than in goods trade.
Du and Shepotylo (2022, 2024) find that the deterioration of EU–UK services trade predates Brexit. Comparing the UK to a set of control countries, they estimate that UK services exports were already declining by approximately 5.7% per year between 2016 and 2019 — representing a shortfall of around £18.5 billion per year — and by 6.2% for trade with the EU alone.
Over the same period, Ireland experienced growth in its services exports of close to 15% per year. This divergence suggests that a portion of the activity previously conducted from the UK may have been relocated to other business centres established within the EU, particularly Dublin. Crucially, these effects emerged before any formal change to the trade relationship. The data therefore point to anticipation and uncertainty as important drivers of adjustment, with firms responding to the prospect of future barriers and regulatory divergence. This mechanism had been anticipated by Crowley, Exton and Han (2018), who argued that the anticipation of higher trade costs could in itself affect firm behaviour well before new rules entered into force.
Behind the relative stability of aggregate services exports lies a profound sectoral recomposition. The categories that have grown the most are those that can be provided remotely, without cross-border movement of goods or people, and that are therefore least exposed to the frictions created by Brexit: intellectual property services exports have almost quadrupled (index 399 in 2024, base 2016=100), other business services exports have more than doubled (216), and information and communication technology services have reached 221. Their strong performance reflects both the global expansion of digital services and the UK's comparative advantage in these activities.
Conversely, the service categories involving physical presence or on-site activity are the only ones to have declined: exports of maintenance and repair services have fallen to 78 and construction services to 54, well below their 2016 levels. These are typically services that depend on the temporary cross-border movement of workers — a mode of delivery that Brexit has fundamentally altered. Exports of travel and transport services, for their part, collapsed in 2020 and had only recovered to around 116 in 2024, a dynamic attributable primarily to the Covid-19 pandemic rather than to changes in the EU–UK trade relationship.
Relative to their control scenario, Du and Shepotylo identify transport, travel, insurance, telecommunications, and intellectual property services among the sectors most affected by Brexit. This ranking only partially aligns with the picture emerging from our index levels, in which telecommunications services feature among the best-performing sectors. This apparent divergence reflects methodological differences rather than any contradiction in the results. Whereas a control-group approach measures performance relative to an estimated non-Brexit counterfactual, an index level simply records observed outcomes. Moreover, since the estimated effect extends to UK exports to non-EU markets as well, the adjustment cannot be considered a purely bilateral EU–UK phenomenon.
As noted, two reasons may help explain the developments highlighted in the analysis above. First, the end of free movement has restricted the temporary cross-border mobility of professionals — a mode of services delivery designated in WTO terminology as "Mode 4." This appears to have weighed particularly heavily on UK exports in sectors such as construction and personal services, whose performance has lagged behind broader global trends. Second, part of the adjustment may reflect a shift from cross-border services provision to local establishment within the European Union through subsidiaries. Rather than serving their clients remotely from the UK, some firms may have chosen to relocate part of their operations within the single market in order to preserve their market access. This hypothesis is plausible and consistent with the evidence presented above, but it remains difficult to verify empirically. Doing so would require detailed data on foreign affiliate activities, which existing statistics on modes of services supply do not capture (Breinlich and Magli 2024).
Financial services, the country's flagship sector, appear to constitute a case apart. UK exports of financial services to the EU reached an index of 156 in 2024 (+56% on 2016). The question, however, is not one of decline but of counterfactual: does this growth correspond to what would have occurred with maintained access to the single market? Relative to a non-Brexit control scenario, Du and Shepotylo identify financial and insurance services among the sectors where UK exports underperform their estimated trajectory — a shortfall in relative terms that coexists with growth in absolute terms.
At the global level, however, the picture is one of remarkable resilience. The UK remains the world's leading net exporter of financial services, and its surplus, far from shrinking, has reached record levels, on the order of $98 billion in 2023 and close to $127 billion in 2024. This resilience highlights the enduring strengths of the City of London — notably the depth of its capital markets, its legal infrastructure, its concentration of expertise, and its global networks.
Brexit therefore appears less as a collapse of UK financial services exports than as a gradual geographical reorientation. The most significant change lies in the shifting relative weight of the European market. The share of UK financial services exports destined for the EU fell from around 40% in 2019 to approximately 30% in 2023. Over the same period, the United States overtook the European Union to become the UK's largest market for financial services exports.
These developments suggest that the City has, at least in part, offset slower growth in its European activity by strengthening its position in global markets. That said, this adjustment should not be interpreted as evidence that Brexit had no effect on the sector. It points instead to a redistribution of activity across markets. The EU has become relatively less important to the UK financial services industry, while non-European markets — and the United States in particular — have taken on a larger role in supporting the sector's growth.
Furthermore, the resilience of trade flows should not be conflated with institutional continuity. Beneath relatively stable export figures, a significant relocation of assets, legal entities, and business functions has taken place since Brexit. According to Hall (2022) and the New Financial tracker, approximately £900 billion in banking assets — representing 10% of the UK system — have been transferred or redirected to the EU. A further hundred billion pounds in insurance and asset management have also migrated, involving more than 440 firms.
The drivers of this evolution are well identified. By leaving the single market, financial institutions established in the UK lost the European financial passport that had previously allowed them to serve clients across the European Union from a London base. Unlike goods, financial services were largely excluded from the TCA. Access to the EU market for UK-based financial actors now depends on unilateral equivalence decisions granted at the Commission's discretion, which remain limited in scope and can be revoked at any time. A Memorandum of Understanding signed in June 2023 established a regulatory dialogue framework between the EU and the UK, but it has restored neither passporting rights nor genuine market access guarantees.
02 | The lens of financial flows
A granular, product-by-product analysis of financial services trade is also highly instructive. In the clearing of euro-denominated interest rate derivatives, the London-based clearing house LCH still held approximately 80% of the global market at end-2023. Likewise, London has remained the leading foreign exchange trading centre, processing close to 42% of euro transactions in 2022. In other segments, however, Brexit has led to a redistribution of activity. Amsterdam overtook London as Europe's leading equities trading venue as early as 2021, while a portion of derivatives activity migrated not to continental Europe but to New York.
In this sense, Brexit has not produced a straightforward transfer of financial activity from London to a single European rival. It has instead contributed to the fragmentation of activities across multiple centres. The City has retained many of its traditional strengths, but has lost part of its role as the undisputed hub for financial services directed at the EU. The principal long-term beneficiary of this reorganisation may ultimately prove to be New York rather than any EU financial centre.
The principal risk for London is therefore no longer that of a sudden rupture — the bulk of which has already occurred — but rather that of gradual erosion. Regulatory divergence, the continued development of financial capabilities within the EU, and the Union's broader pursuit of strategic and financial autonomy could progressively weaken London's position over time, even in the absence of any dramatic disruption.
03 | From investment to income: the cumulative channel
Foreign direct investment offers another important lens through which to assess the economic effects of Brexit, linking trade, productivity, and future income generation. Here again, the adjustment appears to have begun well before the end of the transition period.
Breinlich, Leromain, Novy and Sampson (2020) show, using transaction-level data, that the 2016 vote triggered a reorientation of investment flows. UK firms increased their investments within the European Union, while European investors reduced their investments in the UK — an early sign that firms on both sides were adapting to the prospect of future trade and regulatory frictions. Dhingra and Sampson (2022) attribute a large part of this adjustment to uncertainty. Before any formal change in market access conditions, firms were already responding to the possibility of future barriers by reconsidering the location of production, investment, and business activity.
It is this channel that Bloom et al. (2025) identify as the principal of the four mechanisms through which Brexit has reduced GDP.
Investment stocks recorded in the United Kingdom confirm this adjustment, while also underscoring its complexity.
At first glance, they do not point to a collapse of European investment in the UK. The EU's direct investment stock has continued to increase throughout the post-Brexit period, rising from approximately €747 billion in 2019 to €878 billion in 2023 and €900 billion in 2024. There has therefore been no aggregate decline: in absolute terms, European capital has continued to accumulate in the UK.
However, national trajectories are markedly more divergent.
Ireland's investment stock in the UK increased more than fivefold between 2015 and 2024, rising from approximately €17 billion to €91 billion, with most of the increase occurring in 2023 and 2024. The Dutch stock, which is highly volatile, has remained well above its pre-referendum level, at approximately €201 billion in 2015 and €236 billion in 2024, with a peak of close to €268 billion in 2022. Germany's stock, by contrast, has not grown. After peaking at close to €102 billion in 2018, it fell back to approximately €59 billion in 2024 — below its 2015 level of around €70 billion.
These elements must be interpreted with caution. A significant share of international investment flows passes through holding companies and financial vehicles that do not necessarily correspond to underlying productive activity.
The picture is sharper when one examines outward UK investment. According to provisional ONS data for 2024, the European Union appears as the only major region in which the UK's outward direct investment stock has declined — by approximately €166 billion since Brexit. Were this trend to persist in the years ahead, it would mark a structural withdrawal potentially consequent on Brexit.
The regional distribution of investment within the UK also offers a more nuanced picture than is sometimes assumed. London's share of inward investment has remained broadly stable, increasing only marginally from 48.9% to 49.7% over the period. But the apparent concentration of investment in the capital reflects not so much London's growing attractiveness as the relative weakening of other regions.
The investment income channel is both the slowest to materialise and the most difficult to assess. Unlike trade or investment flows, changes in income generated by foreign assets and liabilities can take many years to appear, reflecting past investment decisions rather than current economic conditions.
To date, the available evidence remains inconclusive. The investment income balance with the EU has stayed negative throughout the period, but unstable and without a clear direction. The deficit oscillates between €45 billion in 2014 and €5 billion in 2022, then €24 billion in 2024. These movements suggest considerable volatility but do not, at this stage, point to a systematic deterioration of the UK's income position vis-à-vis the European Union.
The picture is no clearer when one turns to the rest of the world, where investment income balances display even greater instability. The limitations of available data also complicate the interpretation of results. While the composition of investment income can be studied by investment category at the global level, no comparable public data are available for UK–EU relations specifically. As a result, any assessment of Brexit's impact on investment income must, for the time being, remain cautious.
04 | A cumulative effect that may yet unfold further
The net international investment position nonetheless offers a possible explanation of how Brexit's effects could become cumulative over time.
The net direct investment balance, which stood at a surplus of some €70–80 billion in the mid-2010s, turned negative from 2017 onwards, reaching a deficit of €346 billion in 2022, before recovering to €181 billion in 2024.
These fluctuations must be interpreted with caution, as they are influenced not only by investment flows but also by asset valuation changes and exchange rate movements. The sharp recovery recorded after 2022 does not therefore necessarily signal a reversal of the underlying trend.
The overall movement is nonetheless significant. A deterioration in the net direct investment stock implies a reduction in the UK's net external wealth. Over time, a lower stock of foreign assets relative to foreign liabilities can translate into lower future investment income, even if such an effect is not yet visible in current income flows.
It is this mechanism that underpins the idea of a cumulative Brexit shock — the notion that the consequences of lower investment accumulate gradually and may not become fully visible until a much more distant time horizon.
At present, this effect remains a hypothesis rather than an established fact. The available data neither confirm nor refute such an evolution, while the sensitivity of the net investment position to asset price fluctuations calls for particular caution in interpretation.
Bibliography
- Bakker, J. D., Datta, N., Davies, R. and De Lyon, J. (2022), « Non-tariff barriers and consumer prices, evidence from Brexit », CEP Discussion Paper 1888, LSE.
- Bloom, N., Bunn, P., Mizen, P., Smietanka, P. and Thwaites, G. (2025), « The Economic Impact of Brexit », NBER Working Paper 34459.
- Born, B., Müller, G. J., Schularick, M. and Sedláček, P. (2019), « 350 million a week, the output cost of the Brexit vote », VoxEU.
- Breinlich, H., Leromain, E., Novy, D. and Sampson, T. (2020), « Voting with their money, Brexit and outward investment by UK firms ».
- Breinlich, H., Leromain, E., Novy, D. and Sampson, T. (2022), « The Brexit Vote, Inflation and UK Living Standards », International Economic Review 63(1).
- Breinlich, H. and Magli, A. (2024), « Services trade and modes of supply after Brexit », CEPR Discussion Paper 19525.
- Crowley, M., Exton, O. and Han, L. (2018), « Renegotiation of trade agreements and firm exporting decisions », Bank of England.
- Dhingra, S. and Sampson, T. (2022), « Expecting Brexit », Annual Review of Economics 14.
- Du, J. and Shepotylo, O. (2022, 2024), « Brexit and UK services trade », in Portes (ed.), CEPR Press, and later longitudinal analysis.
- Freeman, R., Manova, K., Prayer, T. and Sampson, T. (2022), « Unravelling deep integration, UK trade in the wake of Brexit », CEP Discussion Paper 1847.
- Freeman, R., Garofalo, M., Longoni, E., Manova, K., Mari, R., Prayer, T. and Sampson, T. (2025), « Deep Integration and Trade, UK Firms in the Wake of Brexit », CEPR Discussion Paper 19869.
- Hall, S. (2022), « The economics of UK financial services post Brexit », in Portes (ed.), CEPR Press.
- New Financial (2021), « Brexit and the City, the impact so far ».
- Office for Budget Responsibility (2024), « Brexit analysis », Economic and Fiscal Outlook.
- Portes, J. (ed.) (2022), The Economics of Brexit, What Have We Learned, CEPR Press.
- Posen, A. and Rengifo-Keller, L. (2022), « Brexit Britain in a changing global economy », in Portes (ed.), CEPR Press.
- Springford, J. (2022), « The cost of Brexit to June 2022 », Centre for European Reform.
