Creating fiscal headspace by delivering outcomes and not just bills to taxpayers
I am watching the Conservative conference in Birmingham with some anticipation. The party chairman has promised more announcements about scrapping taxes before the week is out. Last week I wrote about Andy Burnham's first conference speech as Prime Minister. He offered a National Care Service, a public body to run the electricity grid, a restricted pensions triple lock and a break with forty years of policy. The bond market answered almost immediately. The ten-year gilt yield rose from 5.35% to almost 5.5% after the speech, and the thirty-year yield crossed 6% on Thursday for the first time since 1998, although a global sell-off tied to oil and inflation was pushing yields up at the same time.
I hope Badenoch will offer a more pragmatic prescription for an ailing economy, one that improves on Burnham's. My bar is very low. I just want it to outline who pays, what they get for it, and why a mobile taxpayer should stay. These three answers matter more than the headline tax rate and will decide where the next round of capital goes.
This issue is not an easy one to address. Start with the firms. Ferguson moved its primary listing to New York in 2022, CRH followed in 2023 and Flutter in 2024. This year, Wise made Nasdaq its primary market in May, AstraZeneca launched a direct listing on the NYSE in February whilst keeping London, and Flutter cancelled its London listing on 3 August. The number of companies listed or traded in London fell from about 2,365 in 2015 to just over 1,560 in 2025. Brexit is one clear cause, but valuation is another: Rathbones research published in 2024 found UK forward price-to-earnings ratios 32% below US peers on a like-for-like basis. I would not pin the exodus on one factor, but its direction can’t be disputed.
Capital is harder to contain. HMRC told the Lords in May that it holds no comprehensive data on high earners leaving, because not every taxpayer notifies it. What is visible is concentrated and anecdotal. Chris Rokos, reportedly the third-largest individual taxpayer after paying an estimated £330 million in 2025, confirmed in September that he is moving his tax residence to Greece, joining Lakshmi Mittal and Nassef Sawiris, who left after the non-dom regime ended. Meanwhile, academic evidence on earlier reforms points the other way: modest emigration followed the 2017 non-dom change, with a central migration elasticity of 0.02, and UK tax paid by the affected group rose by more than 150%. However, the 2025 abolition is a bigger shock than the one they studied, and HMRC will not publish returns for the first year under it before 2027.
Prime central London is where departures show up first atsay who pays, what they get a price. LonRes reports first-half 2026 transactions down 12.7% on the year, an average discount to asking price of 10.4% in June, and half of all sold properties having been reduced first. Savills puts prices 24.5% below the 2014 peak at the start of 2026, and some neighbourhoods are now more than 25% below it. Agents attribute part of the fall to departures after non-dom abolition.
The gilt market may be delivering the harshest verdict. The September 2022 mini-budget set the new baseline, and markets now price any fiscal signal against this baseline. Reports this week note that the thirty-year yield has passed the mini-budget peak. The Debt Management Office has cut the average maturity of the gilts it issues to just under seven years in the first three quarters of 2026, from about twenty years before the financial crisis, because long-dated demand is thin. Public sector borrowing was £18.3 billion in August, the second-highest August on record and £3.5 billion above the OBR's forecast. Borrowing for the year to August stands at £77.3 billion, £8.1 billion above forecast. John Healey is left with the hospital pass with Budget on 28 October.
Whilst the tax take has risen, despite reports and expectations to the contrary, from 34.5% of GDP in 2024–25 to 38.5% by 2030–31, a historic high, self-assessed income tax receipts in July were the highest for any July since monthly records began in 1999. The problem lies with the spending, debt interest, and a tax base that can move rapidly and fluidly.
Growth has to some extent stalled – GDP rose 0.6% in the first quarter of 2026 and 0.5% in the second. The weakness is cumulative. GDP is 6.2% above its late-2019 level after six and a half years, under 1% a year on my arithmetic, and the OBR forecasts 1.1% for 2026. Whilst the headline unemployment remains reasonable at just under 5%, the number of payrolled employees fell by 84,000 in the year to May–July, and vacancies stand at 702,000, roughly 2.5 unemployed people for every vacancy. Youth unemployment is 16.2%, the highest in over a decade, and 981,000 young people are not in education, employment or training (NEET). A state that cannot find work for its young has a harder time asking its older earners to fund them.
None of this is peculiar to Britain. The OECD's Revenue Statistics 2025 put the average tax-to-GDP ratio of its 38 members at a record 34.1% in 2024, ranging from 18.3% in Mexico to 45.2% in Denmark, with France at 43.5% and the United States at 25.6%, down from 28.3% in 2000. The OECD's own headline credits higher labour taxes for the record. Governments are raising money where people cannot easily move. In the UK, the OBR expects income tax threshold freezes to pull an extra ten million people into higher rate bands by the end of the decade compared with 2023. The base that can leave pays less than it did, and the base that cannot pays more.
The same story holds for corporate taxes. The average combined statutory rate across the OECD/G20 Inclusive Framework fell from 28.0% in 2000 to 21.7% in 2019 and was 21.2% in 2025. The slide has levelled off because a 15% global minimum tax now applies to groups with revenue above €750 million. The OECD then agreed to a side-by-side arrangement on 5 January 2026 that exempts US-parented groups from most of those rules for fiscal years beginning on or after 1 January 2026. Competition has not ended. It has moved from the headline rate to certainty, substance and the cost of dealing with the state.
This same tax-arbitrage pressure shows up in the United States. IRS data compiled by the Economic Innovation Group put Florida's net gain of adjusted gross income at $39.2 billion and California's net loss at $29.1 billion between the 2020 and 2021 filings. By the 2022–23 filings, the gap had narrowed to a Florida gain of $20.6 billion and a California loss of $11.9 billion. California voters will decide on 3 November whether to levy a one-time 5% tax on the net worth of roughly 200 residents worth more than $1 billion, measured at 1 January 2026, so someone who left in December 2025 will owe nothing, but someone who left in February will. The state's legislative analyst expects more billionaires to leave and ongoing income tax revenue to fall by hundreds of millions of dollars or more a year. Advocacy bodies put the damage far higher, with the National Taxpayers Union Foundation counting half a dozen departures carrying an estimated $700 billion of wealth. Washington, Maryland, Massachusetts and Minnesota have taxes on millionaires in force or in debate.
My conversations with the firms and individuals leaving high-tax countries reveal that no bargain or value proposition is offered to them. It is not just tax reduction that drives companies and individuals - their reasons are manifold and rational. Governments should take these into account before resorting to punitive measures, as current tax regimes are increasingly poised to do so.
The biggest issue is lack of trust - the OECD's 2024 trust survey found that 44% of respondents across 30 countries had low or no trust in their national government, and even fewer said they felt they had a say in their government.
When the public sees the top paying little, governments reach for new taxes on the top, and the top has the means to leave. But labour income is reported by the employer and taxed at source, with no resort to any relocation - other than contracting or nomad work. Corporate profit and wealth, by contrast, can be moved, booked elsewhere, or valued at a time of the owner's choosing. Gabriel Zucman and his co-authors estimate that 36% of multinational profits were shifted to tax havens in 2015, and the EU Tax Observatory puts the 2022 figure at about $1 trillion, or 35% of profits booked outside headquarters countries. US multinationals shift twice as much, relative to their foreign earnings, as other multinationals. The same observatory reports that global billionaires pay effective rates equivalent to between 0% and 0.5% of their wealth. This creates a vicious circle, as the government reacts to public sentiment and reaches out to increase the taxes at the top brackets; increased migration of capital ensues – until the billionaires become the villains in the public narratives.
The other side of the tax continuum, low-tax jurisdictions, are not charities. A study by Tørsløv, Wier and Zucman found that tax havens collect disproportionate fiscal revenue relative to the size of their economies compared with higher-tax countries, despite low statutory rates, because they capture a base that others lose. Ireland shows the scale. Underlying corporation tax receipts were €32.9 billion in 2025, about triple the 2019 level, and the ten largest companies paid 56% of them. Around 84% came from foreign-owned multinationals. The UAE took the other route. It introduced a 9% federal corporate tax from June 2023 with a zero rate on profits up to AED 375,000, a 5% VAT since 2018 and no personal income tax. S&P estimated the new tax could add 1.5% to 1.8% of GDP to the emirates' revenue from 2025.
Haven economies run on two engines: a captured base and a bundle of services that others cannot match at the same price. The gains are revenue, financial services employment, and rent.
So how can the OECD countries address this? Think tanks and academics have produced six families of prescriptions. None work alone, and each carries a risk.
Zucman's EU Tax Observatory proposed a global minimum tax of 2% on billionaires' global assets. The UK Wealth Tax Commission, too, studied a one-off 5% charge on wealth above £500,000, paid over five years. However, the evidence against this approach is clear and strong. Norway's 2022 increase led to a significant exit of its richest residents, and Switzerland's voters rejected a 50% inheritance tax above CHF 50 million by 78.3% in November 2025. The reality is that the results depend on design, on whether the target can move, on whether the state can credibly promise not to move the goalposts again, and on whether the state can deliver a relevant and proportionate value proposition.
Much like most right-of-centre economists. I like to argue for lower rates, simpler rules and lighter regulation, and an entrepreneurial economy. Evidence supports this: Moretti and Wilson's study in the American Economic Review found a long-run elasticity of 1.8 between the after-tax income gap and migration and no response at the median earner. Italy sells a fixed annual charge on foreign income: €100,000 in 2017, €200,000 from August 2024 and €300,000 for arrivals from 1 January 2026, with €50,000 per family member. The challenge, however, is that government borrowing at 6% on thirty-year money cannot cut rates faster than it cuts spending, and competition for capital continues to escalate.
Another school of thought argues to tax the assets that can’t be moved - land. The Mirrlees Review, which the Institute for Fiscal Studies ran in 2011, proposed replacing business rates and non-domestic stamp duty with a land value tax on business and agricultural land, if practicable, and abolishing stamp duty land tax. Land cannot emigrate, and a tax on its value does not change how much of it exists. The challenge is that this has a ceiling, bills shift by postcode, and cash-poor owners of valuable sites will need a deferral mechanism.
For the states that created endowment structures such as SWFs, the answer is to invest in assets and create annuities. Singapore's Net Investment Returns framework came into force in FY2009 after a constitutional amendment in 2008, and Temasek joined it in FY2016. It allows the government to spend up to 50% of the expected long-term real return on the net assets invested by GIC, the Monetary Authority and Temasek. The contribution was S$27.53 billion in FY2025 and is estimated at S$28.48 billion for FY2026. Corporate income tax, at S$35.24 billion, is larger, and the contribution is about 17% of total revenue of S$158.39 billion. The FY2026 estimate more than covers a projected primary deficit of S$2.6 billion, and the government ended FY2025 with a surplus of S$15.1 billion, 1.9% of GDP. The reserves came from budget surpluses and land-sale proceeds, and the constitution requires a balanced budget each term of government. Norway's fund stood at NOK 22.683 trillion in June 2026, and transfers from it will finance about 27% of the 2026 central budget under a 3% expected-return rule. The awkward fact is that Norway has the fund and still raised its wealth tax. An endowment reduces the need to tax the mobile. It does not remove the political urge to do it. However, few OECD nations have this advantage.
Perhaps the most effective strategy is to make state expenditure transparent and public-private partnerships belong here. But caution is needed, as building public assets with private money can look cheap upfront but cost considerably more later. British experience is a cautionary tale. HM Treasury data to March 2024 counted 665 Private Finance Initiative projects with about £50 billion of capital investment and £136 billion of annual charges to be paid from 2024–25. The fees mentioned here also cover maintenance and services, so it is not a like-for-like comparison with £50 billion, but it does show that the bill runs for decades after the ribbon is cut. The National Audit Office found in 2018 that private financing can cost 40% more than government borrowing.
Australia, perhaps, offers a better model for infrastructure. Its Asset Recycling Initiative put A$5 billion behind a 15% bonus to states that sold mature assets and reinvested the proceeds. New South Wales was due A$1.965 billion on an estimated A$13.1 billion of lease proceeds reinvested, most of it in Sydney Rapid Transit. The Australia Institute disputed the net value, but the logic holds: private capital buys assets with a revenue history, and the state spends the proceeds on new construction.
The same logic could carry social programmes through outcomes contracts, in which providers are paid when an agreed result occurs. The Oxford Government Outcomes Lab counted 250 impact bonds worldwide by July 2022, and its July 2021 data showed more than $496 million raised across 214 of them. Admittedly, this is quite meagre compared to the state budget, and I have not seen evidence that they beat conventional commissioning at scale. What they do well, however, and should be used for, is turn a taxpayer’s bill into an outcome, which is what the mobile taxpayer says he is missing. England's planning obligations already do a version of it by making the developer who benefits from a permission fund the roads and schools it requires.
Global collaboration in setting a floor will remain important for OECD countries. International agreements can stop countries from undercutting each other on tax rates. They cannot stop a country from winning business by being more reliable and easier to deal with than its neighbours. The January 2026 exemption for American groups weakened the 15% floor and the OECD agreement. The recourse for countries like Singapore, the UAE and Ireland is to make their own system more trustworthy and their processes easy to navigate – sell assurance, not arbitrage.
Singapore's record is not a story about low tax. Its statutory corporate rate is 17%, and corporate income tax is its largest revenue line. What it did right was write a robust set of rules before the fiscal pressure materialised: spend no more than half the expected return, balance the budget over each term, and run a net asset position. Switzerland ties tax competition between cantons to a voting public that defends the bargain, as the November referendum showed. Ireland captured a base and now has to manage its concentration. The UAE added a corporate tax and a VAT rather than defend an exemption it could no longer afford.
The key was that, in each case, these countries' success came from a credible promise that the rules would remain firm and predictable, along with a clear link between the tax bill and delivery of a value proposition to the taxpayer, and building an asset base or institution that did not depend on the political cycles.
None of this can be copied quickly. Singapore's reserves took decades of surpluses, and Norway's fund took decades of oil. A country starting in 2026 with a 30-year yield above 6% can copy the sequence, a rule first, a disclosed surplus second and an asset base third, but not the date.
I will listen for four things from Birmingham.
First, whether Badenoch binds the spending. Last year she set a golden rule that at least half of any saving goes to cutting the deficit. The test now is whether she can hold that line so that a bond investor can check it, because Burnham's triple-lock restriction, which official estimates put at £11 billion a year by 2040, did not stop yields from rising.
The second is who will pay for the tax cuts. The tax take stands near 38.5% of GDP on the OBR's forecast, so there is room in principle, but borrowing is running £8.1 billion above forecast, and the thirty-year yield is above 6%. Therefore, in practice, this will be pulling a rabbit out of a hat.
The third, and perhaps most important, is whether she addresses value delivery: unit costs, contract overruns, and outcomes, so that a large taxpayer can see value in paying tax.
Finally, the fourth is whether she can deliver a budget response on 28 October that the bond market will listen to.
Whatever she says today will be marked first by the gilt market and then by the Chancellor's arithmetic on 28 October. The evidence cannot yet say whether a state can lower the bill and raise delivery in the same parliament. Nobody has done it starting from a 30-year yield above 6%. I like Kemi, and I wish her all the best.
#MacroObserver #UKEconomy #PublicFinance #CapitalMobility
More in this area
New Zealand Needs a Government That Can Decide
The current election results have brought back all the reasons I have never been comfortable with MMP as the platform for New Zealand’s democracy. I think it brings back the arguments for a different architecture, not a minor…
Youth, Tariffs, and India’s Strategic Autonomy Story
Two recent issues – the youth movement, new tariffs imposed ondemanded and Russian oil purchases and future tariffs of 100% on generics – made me reflect over the foreign policy posture of India. I have never been comfortable…
Coming home to the same boat, just a different ocean
Landing back in Sydney this month, the first thing that struck me wasn't jet lag. It was how many conversations with colleagues here picked up exactly where the ones in Europe left off — energy costs, fiscal headroom, a…