Key takeaways

  • The OBR assumes Brexit reduces long-run UK productivity by about 4% and trade intensity by around 15%, and says the trade assumption is broadly on track.
  • The Centre for European Reform estimates goods exports to the EU are roughly 16% lower and services imports about 19% lower than they would have been.
  • Academic work led by King’s, Stanford, Nottingham and the Bank of England puts missing business investment at 12-18%, driven largely by uncertainty.
  • The end of free movement reshaped the labour market: EU worker numbers fell while non-EU migration surged, and total net migration hit a record before falling back.
  • The 2025 UK-EU “reset” delivered an SPS deal and fisheries agreement, but economists say it only modestly offsets the wider cost.

Ten years on from the referendum that split the country, the Financial Times returned to the question that has haunted British economic policy ever since: how much has leaving the European Union actually cost? In its FT Film “The Brexit effect,” the paper examined how the decision has rippled through trade, investment and the labour market a decade after the June 2016 vote, and how difficult it remains to disentangle the effect of Brexit from the pandemic and the energy crisis that followed.

That difficulty is precisely why the debate endures. Britain did not leave the EU in a laboratory. It left amid Covid lockdowns, the largest energy shock in a generation, and a global surge in inflation. Yet after a decade of data, a broad body of evidence has accumulated from the Office for Budget Responsibility (OBR), the Bank of England, the Office for National Statistics (ONS) and independent researchers. The picture that emerges is contested at the edges but consistent at its core: Brexit has imposed a measurable, ongoing drag on the UK economy, even if the precise figure will never command universal agreement.

This article sets out what the evidence shows, where the honest uncertainties lie, and what the counter-arguments from Brexit’s defenders actually claim.

What the evidence shows on trade, investment and GDP

The single most cited number in the entire debate belongs to the OBR. Since shortly after the 2016 vote, the fiscal watchdog has built its forecasts on the assumption that leaving the EU would reduce the UK’s long-run productivity by around 4% relative to remaining a member, and would cut both exports and imports by roughly 15% in the long run. Crucially, the OBR treats these not as predictions of a cliff-edge event but as a gradual adjustment playing out over 15 years.

The watchdog has repeatedly reviewed whether reality is tracking its assumptions. In its assessments it has concluded that the assumption Brexit would reduce overall trade intensity by 15% “appeared to be broadly on track,” while noting a striking divergence between goods and services since the pandemic. Goods trade, exposed to new customs paperwork and regulatory checks, has underperformed; services, much of it digital and less dependent on physical borders, has held up better than many feared.

The Centre for European Reform (CER), which has tracked the “cost of Brexit” year by year using gravity-model and counterfactual techniques, offers some of the most granular estimates. In its ten-year analysis of leaving the customs union and single market, the CER estimates that UK goods exports to the EU are around 16% lower, and goods imports around 14% lower, than they would otherwise have been. Services show a split: services exports to the EU down about 7%, but services imports down roughly 19%. Notably, the CER attributes the bulk of the damage not to tariffs or customs union exit but to leaving the single market, concluding that “the regulatory costs related to Brexit… have had a much more significant impact on UK-EU trade than customs-related barriers.”

On the wider economy, the range of GDP estimates is well known and much argued over. Studies collated by the CER and others typically place the long-run hit to GDP somewhere in the region of 4% to 8%, with the OBR’s productivity assumption sitting near the lower end of that band. A major academic study published in 2026, led by King’s Business School with Stanford University, the University of Nottingham and the Bank of England and drawing on the Decision Maker Panel survey of more than 7,000 firms, found business investment to be between 12% and 18% lower than it would have been without Brexit. The same research pointed to GDP several percentage points below its pre-Brexit trajectory and productivity down in the region of 3% to 4%.

Investment is arguably the clearest transmission channel. After the 2016 vote, UK business investment flatlined even as it rose across other advanced economies. The King’s-led work attributes much of this to uncertainty rather than to trade barriers alone: more than half of firms cited Brexit as a top-three concern, and the “large, broad and long-lasting” uncertainty discouraged spending on machinery, digital systems and research. As Professor Paul Mizen, who led the study, put it, “the evidence shows a steady accumulation of economic costs: weaker investment, slower productivity growth.”

The measurement challenge: why no single number wins

Anyone claiming to know Brexit’s exact cost to the penny should be treated with caution, and the OBR itself is careful to stress the uncertainty around its assumptions. The core problem is the counterfactual. To measure Brexit’s impact you must estimate what would have happened in a Britain that never left, and no such Britain exists to observe.

Economists use several techniques to build that shadow economy. “Doppelgänger” or synthetic-control methods construct a weighted basket of comparable countries whose combined pre-2016 behaviour tracked the UK, then measure the gap that opens after the referendum. Gravity models, favoured in the CER’s latest trade work, predict bilateral trade flows from the size of economies and the distance and barriers between them, then compare predicted with actual flows. Each approach has strengths and weaknesses, and each can be attacked on its choice of comparators or its handling of one-off shocks.

Those shocks are the second problem. The UK formally left the single market and customs union on 1 January 2021, in the depths of the pandemic. The energy price spike of 2022 hit gas-dependent European economies hard, complicating any comparison with EU neighbours. Sceptics argue, reasonably, that some of the “Brexit effect” measured in the data is really a Covid or energy effect wearing a Brexit mask. Defenders of the mainstream estimates counter that the studies deliberately exclude the most distorted pandemic years, and that the UK’s persistent underperformance on trade openness and investment predates the energy crisis and outlasts the pandemic recovery.

The honest position is that a range, not a point, is the responsible way to describe Brexit’s cost, and that reasonable analysts differ within that range.

Migration and the labour market

If trade is the most quantified effect, migration is the most visible. The end of free movement in January 2021 and the introduction of a points-based immigration system transformed the composition of Britain’s workforce.

According to the Migration Observatory at the University of Oxford, immigration from EU countries fell by more than 80% between 2016 and 2025. EU net migration turned negative in 2022, and the number of EU citizens in payrolled employment dropped to around 2.3 million by the end of 2024, a fall of roughly 13% from the 2019 peak. The decline was sharpest in lower-paid sectors that had relied on EU labour: hospitality saw employment of EU workers fall by more than a third.

Yet the headline story of the past decade is not falling migration but rising migration of a different kind. As EU flows contracted, non-EU migration surged, and total net migration reached a record of around 944,000 in the year to March 2023 before government restrictions brought it down towards 200,000 by mid-2025. Britain did not become a lower-migration country after Brexit; it became a differently sourced one, drawing more heavily on workers, students and dependants from outside Europe. For some sectors that shift eased shortages; for others, particularly seasonal agriculture and hospitality, it left persistent gaps and upward pressure on wages and costs.

The 2025 UK-EU “reset”

The political mood a decade on is markedly less confrontational than during the years of negotiation. The Trade and Cooperation Agreement (TCA), which has governed the relationship since 2021, was always a thin, tariff-free but friction-heavy deal, and its own review provided the occasion for a rethink.

At a summit in May 2025, the UK and EU agreed a “reset” built around a new strategic partnership. The centrepiece for business was a commitment to a sanitary and phytosanitary (SPS) agreement, intended to strip away many of the veterinary checks and certificates that had throttled agri-food trade, alongside a reciprocal fisheries deal extended to 2038 and continued energy cooperation. As UK in a Changing Europe has documented, progress in the months since has been mixed: the fisheries and energy elements moved forward, while talks on defence-fund participation and youth mobility proved harder, and full implementation of the food-standards deal is targeted only for later in the decade.

Economists broadly welcome the reset as sensible housekeeping, but caution against overstating it. An SPS agreement could meaningfully help food and drink exporters and ease pressure at the border, yet it does not restore single-market membership, and most analysts judge its macroeconomic effect to be modest relative to the scale of the original loss. The reset narrows the gap; it does not close it.

Who has been affected

The costs of Brexit have not fallen evenly. Smaller exporters, for whom fixed compliance costs loom largest, have borne a disproportionate share of the burden; many simply stopped selling to the EU rather than navigate new certification and customs regimes. Food, drink and agricultural producers faced some of the heaviest non-tariff barriers. Manufacturers reliant on just-in-time European supply chains absorbed new frictions and inventory costs.

Services, by contrast, have proved more resilient. Professional, financial and digital services, delivered across borders without a physical shipment, have grown despite the loss of passporting rights that many in the City feared would be devastating. Some financial activity relocated to EU hubs, but the wholesale exodus predicted in 2016 did not materialise. Meanwhile, consumers have paid indirectly through a modestly higher cost of imported goods and food, and through the opportunity cost of investment and productivity growth that never happened.

Historical context

It is worth remembering how far expectations have travelled. In the referendum campaign, the Treasury and most international institutions forecast an immediate recession that did not arrive in the form predicted. Brexit’s critics were, in the short run, wrong about the timing and the mechanism. What the past decade suggests is that the cost has been slower, subtler and cumulative rather than sudden, a gradual erosion of trade intensity, investment and productivity rather than a single crash.

This matters for how the debate is conducted. The absence of an overnight catastrophe allowed proponents to claim vindication, while the steady accumulation of underperformance allowed critics to claim theirs. Both can point to real data. The lesson is that structural economic change plays out over years, and that the true test of Brexit was always going to be the long-run growth path, not the quarterly numbers immediately after the vote.

Opportunities, challenges and material risks

Brexit’s defenders point to genuine freedoms secured. The UK has struck trade deals of its own, joined the CPTPP trans-Pacific bloc, and gained control over its own regulation, tariffs and immigration policy. In principle, regulatory divergence could allow Britain to move faster than the EU in areas such as technology, life sciences and financial innovation. Whether those freedoms are used well is a policy choice, not a foregone conclusion, and to date the measurable trade gains from new agreements have been small relative to the loss of frictionless access to a market on the doorstep.

The material risks now are less about a single shock and more about persistence. If the productivity and investment gaps prove durable, they compound: lower investment today means a smaller capital stock and weaker productivity tomorrow. A thin trading relationship also leaves the UK more exposed to any future tightening of EU rules on which it has no vote. Against that, the reset shows the relationship can be improved incrementally, and a more pragmatic political consensus reduces the risk of renewed rupture.

The counter-case

A serious article must give the counter-case its due, and it is not merely political. The Institute of Economic Affairs and other sceptics have directly challenged the OBR’s methodology, arguing that its 15% trade assumption was set early, calibrated to external studies, and risks becoming a self-fulfilling forecast rather than a measured outcome. They note that UK goods exports to the rest of the world have grown, that services exports have been strong, and that global trade patterns since 2020 owe more to pandemic disruption and shifting supply chains than to Brexit.

Critics also make a fair methodological point: synthetic-control and gravity studies are sensitive to which comparator countries and time windows are chosen, and small changes in assumptions can move the headline number by several percentage points. Some argue the long-run benefits of regulatory autonomy and independent trade policy will only accrue over decades and are, by their nature, hard to capture in models built on the past.

The mainstream response is that the weight of independent evidence, from the OBR to the Bank of England’s own surveys to multiple academic teams, points in the same direction, and that the burden of proof lies with those claiming the frictions are costless. But the existence of a credible, evidence-engaging counter-argument is exactly why this remains a debate rather than a settled fact.

What happens next

The immediate agenda is implementation. Delivering the SPS agreement and the other strands of the 2025 reset would offer the clearest near-term route to easing costs for exporters, particularly in food and drink, and the 2026 review of the TCA provides a formal moment to widen cooperation. Business groups will press for progress on mutual recognition of professional qualifications and smoother arrangements for touring workers and short-term mobility.

The deeper question is whether Britain can lift its underlying investment and productivity rates, with or without closer EU ties. Brexit sharpened a pre-existing weakness rather than creating it from nothing, and the policy levers, planning reform, capital allowances, skills and infrastructure, sit largely in domestic hands. A decade on, the debate is shifting from whether to leave to how to make the settled relationship work as well as it can.

Conclusion

Ten years after the referendum, the responsible conclusion is neither triumph nor catastrophe. The best available evidence, from the OBR’s 4% productivity and 15% trade assumptions to the CER’s detailed trade estimates and academic findings of 12-18% missing investment, points to a real and continuing cost, concentrated in goods trade, business investment and productivity. That cost is genuinely hard to separate from Covid and the energy shock, and honest analysts express it as a range rather than a single figure. The counter-arguments about methodology and long-run opportunity deserve to be heard, and the 2025 reset shows the relationship can still be improved. What Britain can no longer do, a decade on, is pretend there is no cost to weigh at all.