Ten years after the referendum, the economic reckoning on Brexit is finally arriving in a form that's hard to argue with. The Bank of England's Decision Maker Panel - a major business survey set up specifically in 2016 to track the corporate fallout from the vote - has now produced a landmark working paper. Co-authored by Stanford professor Nicholas Bloom alongside four Bank of England and university economists, "The Economic Impact of Brexit" (NBER Working Paper 34459, November 2025) concludes that by early 2025, Brexit had reduced UK GDP by 6% to 8%, business investment by 12% to 18%, employment by 3% to 4%, and productivity by 3% to 4%.
These aren't projections. They're measured outcomes, drawn from nearly a decade of firm-level data covering more than 7,000 UK companies and compared against a basket of 33 peer economies. And buried in the paper is something the insurance industry should be paying close attention to.
Finance and insurance was one of the four most EU-exposed industries in the entire UK economy - in the same bracket as manufacturing, wholesale and retail, and other production. On the paper's composite EU exposure index - which covers exports to the EU, imports from the EU, migrant labour share, regulatory reliance, EU directors and EU ownership - finance and insurance sits well above the UK business average.
Nobody has done the specific calculation of what Brexit cost insurance. This article attempts a first version, using verified ONS Blue Book data and the Bloom paper's methodology. The short answer: it's a lot more than the industry has ever publicly acknowledged.
The paper's methodology tracks what happened to firms with high EU exposure before the referendum versus after 2016, and compares them to similar firms with lower EU ties. The higher the pre-referendum EU exposure, the worse the performance on investment, employment and productivity in the years that followed.
Finance and insurance firms were among the highest-exposure businesses in the survey. EU regulatory exposure was particularly pronounced - which makes sense, because Solvency II, the Insurance Distribution Directive and passporting rights directly determined how UK insurers could operate across the single market. Remove those rights, and you don't just inconvenience the sector. You restructure its operating model.
The investment numbers bear this out. The paper shows that a typical firm with average EU exposure experienced investment growth around 1.7 percentage points lower per year than it would have otherwise. Finance and insurance firms, sitting above the average, experienced headwinds at least as large. The paper's own analysis shows investment effects were among the most severe in the three highest-exposure industry groups: manufacturing, retail and finance.
The directional signal is unambiguous: insurance absorbed a disproportionate share of Brexit's economic costs.
The Bloom paper captures the macroeconomic and firm-level impact. What it doesn't fully quantify are the direct structural costs specific to insurance - the real cash spent adapting to a world where passporting no longer existed.
Those costs were significant, and most have never been added up in one place.
When the UK left the EU single market on 31 December 2020, UK-authorised insurers lost the right to write business across the EEA on the basis of a single UK licence. According to Lockton's analysis, around 27 million insurance contracts were transferred from UK to EU entities to maintain coverage continuity. An estimated 35 new insurance subsidiaries were established in EEA countries from 2016 onwards - each requiring its own capital base, regulatory authorisation, board structure, compliance infrastructure and operational staff. These costs generate no new business. They're pure friction - the price of maintaining market access that previously came for free.
Lloyd's of London - which writes roughly 10% of global commercial insurance and reinsurance - established Lloyd's Insurance Company S.A. in Brussels, now operating 18 branches across the EEA. Before the vote, Lloyd's EU business generated around 11% of its premiums. That business didn't disappear, but restructuring 25 years' worth of European policies and standing up an entirely new subsidiary isn't cheap. Lloyd's Brussels launched in January 2019 with initial capital of approximately €100 million, with further capital injections to follow. The ongoing cost of running a dual regulatory structure - London and Brussels in parallel - represents a permanent addition to Lloyd's operating base that simply didn't exist pre-2016.
Then there are the jobs. The then-Lord Mayor of the City of London, Professor Michael Mainelli, told Reuters in October 2024 that Brexit had cost the Square Mile close to 40,000 jobs: "We had 525,000 workers in 2016. My estimate is that we lost just short of 40,000." The City has since grown to around 615,000, Mainelli noted - "buoyed by the expansion of the insurance and data industries." That recovery is real, but it's been driven largely by domestic demand and the global hardening cycle, not a return of the EU-facing business that moved to Dublin, Amsterdam and Frankfurt.
ONS Blue Book 2025 · Table 2.2 (dataset KKK9) · £ billion, current prices
The sector was on a strong growth trajectory before the referendum, then went sideways for five years while the total UK economy expanded around it
To build a cost estimate, you need a GVA base. Previous versions of this analysis used a £75bn figure for insurance sector GVA, which was wrong. Here's what the ONS Blue Book 2025 - the definitive official source, published 31 October 2025 - actually shows.
The entire finance and insurance sector (SIC K - which includes banking, insurance and financial auxiliaries combined) contributed £153.3bn to UK GVA at current prices in 2016, rising to £204.0bn by 2023. That consistently represents around 8–9% of the total UK economy.
ONS Blue Book 2025 · Tables 2.2 (KKK9) and 4.4.2 (B.1g) · Index: 2016 = 100
Both series rebased to 100 at the referendum year. By 2023 the total economy had grown 40 points; insurance and pensions grew 48 — but from a depressed post-2015 base, with significantly more volatility
Insurance corporations and pension funds specifically - the closest ONS breakdown to "pure insurance," though it includes pension funds - contributed £30.1bn in 2016 and £44.5bn in 2023. Pure insurance underwriting without pension funds would be a subset of that, but the Blue Book doesn't break it out separately. For this exercise, the £30.1bn–£44.5bn range is the right base.
Applying the Bloom paper's implied Brexit cost for a high-EU-exposure sector - somewhere between 7% and 12% above the economy-wide average impact - gives an annual output drag in the region of £3bn to £5bn by 2023, relative to what the sector would have contributed without the referendum result.
Illustrative estimate · Bloom et al. (NBER WP 34459, 2025) × ONS Blue Book 2025, table 4.4.2
Annual drag on insurance and pensions output estimated at £3–5bn by 2023, accumulating to approximately £17–29bn over the decade to 2025. Based on Bloom's documented gradual build pattern applied to actual ONS GVA each year
Methodology note: Annual drag estimated as 7% (low) to 12% (high) of insurance corporations and pension funds GVA (ONS table 4.4.2), scaled by a gradual build factor derived from Bloom et al. Figure 8 — approximately 15% of full effect in 2017, reaching 100% by 2023. This is illustrative arithmetic, not a peer-reviewed estimate. The range reflects data uncertainty: the ONS groups insurance with pension funds and pure insurance GVA cannot be isolated without further work. Source: Bloom, Bunn, Mizen, Smietanka and Thwaites, "The Economic Impact of Brexit," NBER Working Paper 34459, November 2025; ONS Blue Book 2025.
That's the lost annual output from Brexit's dampening effect on investment, employment and productivity in insurance and pensions specifically. Accumulated across the decade since 2016, accounting for the gradual build-up the Bloom paper documents, the cumulative figure is likely somewhere between £20bn and £40bn.
This is openly illustrative arithmetic. It's not a peer-reviewed estimate. But it's grounded in official ONS data and a methodology that has withstood serious academic scrutiny. The wide range reflects a genuine data gap: the Blue Book combines insurance with pension funds, so pure insurance GVA can't be cleanly isolated. Closing that gap is one reason the industry needs to commission a proper study.
For context: at the low end of that range, the cumulative Brexit cost to insurance and pensions is roughly equivalent to wiping out the sector's entire annual output once over. It didn't arrive as a single cliff-edge shock - it built quietly, year by year, exactly as the Bloom paper predicts these costs tend to.
Source: Bloom, Bunn, Mizen, Smietanka and Thwaites, "The Economic Impact of Brexit," NBER Working Paper 34459, November 2025
How Brexit reduced insurance sector output — per the Bloom et al. methodology, applied to insurance specifically
The paper identifies four main ways Brexit hurt high-exposure firms. All four hit insurance harder than most sectors.
Uncertainty was the dominant early driver of lower investment. Insurers hold large investment portfolios and price risk for a living - both activities are acutely sensitive to prolonged policy uncertainty. The paper's Brexit Uncertainty Index, tracking firms that rated Brexit as a top-three business concern, peaked at around 55% in early 2019 and stayed above 40% until mid-2021. Finance and insurance firms, with their above-average EU exposure, consistently reported higher-than-average Brexit uncertainty. The investment that didn't happen during those years - in new products, systems and market development - shows up in the sector's productivity gap today.
Management time is the one that never makes headlines. The paper found that time senior managers spent on Brexit preparations directly correlated with lower subsequent productivity growth. For insurance, that burden was exceptional. CFOs and CEOs at Lloyd's syndicates, composites and brokers spent years navigating passporting loss, Part VII transfers for legacy EU policies, EU subsidiary set-ups, policy wording rewrites for territorial scope, country-by-country compliance across 27 member states, and reinsurance restructuring for business ceded to Brussels entities. None of that time generated revenue.
Trade barriers built gradually after 2021 and are now structural and permanent. UK insurers cannot passport into the EEA without an EU-authorised entity. That changes the capital efficiency of serving European clients in ways that compound over time: more capital tied up across multiple jurisdictions, higher operational overhead, and a fundamental friction that London's competitors in Paris, Dublin and Amsterdam don't face.
Innovation suppression is the hardest to measure. The paper found tentative evidence that high-EU-exposure firms became less likely to register patents after 2016 - a proxy for reduced innovation. For insurance, the equivalent is product development for EU markets: every new product now requires routing through an EU subsidiary, adding cost and complexity that dull the incentive to develop EU-facing lines.
There are legitimate criticisms of Brexit cost studies - separating the referendum's impact from Covid, the energy shock and global uncertainty is genuinely hard. The Bloom paper addresses this carefully, running separate firm-level controls for pandemic effects and using multiple weighting methods for the comparator-country analysis.
But the Bank of England's own position has shifted markedly. In a May 2025 speech in Dublin to the Irish Association of Investment Managers, Governor Andrew Bailey was as direct as a serving central bank governor gets on the subject: "The evidence on Brexit suggests that in the UK the changing trade relationship has weighed on the level of potential supply," he said, adding that this meant the UK should "do all we can to minimise negative effects on trade." In a separate Mansion House speech, Bailey went further: "I do have to point out consequences. As a public official, I take no position on Brexit per se - but the changing trading relationship with the EU has weighed on the economy."
These aren't throwaway comments. They're prepared speeches from the governor of the Bank of England saying, in plain terms, that Brexit has made the UK economy structurally smaller and that the government should work to reduce the damage.
The BBC's reporting on the Bloom paper notes that Starmer is due to meet EU counterparts at a July summit covering food and farm exports, electricity and emissions trading. The paper itself draws a direct parallel between Brexit and US tariff policy - both represent trade fragmentation between major developed economies, with costs that are "substantial, protracted and underestimated at the point of decision."
Arabella Ramage, Legal and Regulatory Director at the Lloyd's Market Association (LMA), said the diplomatic mood had shifted ahead of the talks. "We are pleased to hear a more optimistic tone from the UK and EU delegations before the July summit. This more positive approach has also been reflected in our own interactions with other European officials and associations over the past year."
A Centre for European Reform (CER) report published on 18 June 2026 found that Brexit substantially reduced the UK's trade with the EU, with finance and insurance exports down by 24%. Ramage added that the LMA was among the first to call for equivalence or mutual recognition agreements for insurance between the UK and EU, and that the industry's existing international regulatory standards put it in a strong position to absorb closer alignment.
For insurance, the ask is specific: financial services passporting equivalence. The Trade and Cooperation Agreement negotiated in December 2020 explicitly excluded financial services. The Memoranda of Understanding on regulatory cooperation are limited in scope. A genuine equivalence agreement would eliminate much of the structural cost the dual-entity model imposes - the 35 subsidiaries, the duplicated capital, the split regulatory structures, the ongoing friction.
It's not currently on the agenda. It should be. The industry's case is commercially coherent: equivalence reduces cost, improves capital efficiency and would serve both sides of the Channel. Whether it's Starmer pushing for it at the July summit, or Burnham - if he wins the Labour leadership challenge covered elsewhere in this publication - making it a priority in a future government, the ask from industry needs to be explicit and quantified.
To move from this illustrative estimate to something publishable and politically usable, three things need to happen.
First, the Bank of England's DMP team should publish sector-level Brexit cost estimates. The data exists - the paper explicitly confirms that finance and insurance shows results that differ significantly from the economy-wide average - but the sector breakdown hasn't been released. It should be.
Second, the ABI and the London Market Group should commission an updated economic analysis applying the Bloom methodology to insurance-specific data: premium growth counterfactuals, EU market share loss by class of business, investment suppression, and the annual recurring cost of running 35 EU subsidiaries. The last comparable exercise was the ABI's 2017 Volterra analysis - produced before the full impact of Brexit had landed. A 2025 version, using the Blue Book data and the Bloom framework, would be the most important economic document the UK insurance industry has published in a decade.
Third, the industry needs to aggregate the direct structural costs into a single documented number. Thirty-five EU subsidiaries carrying their own capital requirements, regulatory fees, board costs and operational overhead represent real annual expenditure that would largely disappear under a passporting equivalence deal. The industry knows these numbers. Publishing them would change the political conversation.
The Bloom paper gives the insurance industry a methodology it can use and a number it can no longer look away from. Finance and insurance was one of the most EU-exposed sectors in the UK economy. High EU exposure reliably predicted worse post-2016 performance. The structural disruption - passporting loss, 27 million contracts transferred, 35 new EU subsidiaries, years of management time diverted, Lloyd's Brussels launched from scratch - was larger in insurance than in almost any other sector.
Graeme Trudgill, CEO of the British Insurance Brokers' Association (BIBA), said the impact on brokers was foreseeable. "BIBA advised the Government at the time there would be a significant effect on our sector if UK insurance brokers, particularly those in Northern Ireland, were no longer permitted to directly access our EU clients." BIBA's current position on EU market access is set out in its 2026 manifesto.
Based on ONS Blue Book 2025 data and the Bloom framework, Brexit has likely cost the insurance and pensions sector somewhere between £20bn and £40bn in cumulative lost output over the decade since the vote - building gradually rather than arriving as a single shock, exactly as the research predicts.
The industry knew the costs were real. After 10 years, there's enough data to quantify them properly. It's time someone did.
