This is an In Discussion working paper (July 2026, revised October 2026). It tracks a live political proposal rather than presenting a settled position. The institutional history in Section 1 is well established across several independent sources. The costed figures in Sections 2 and 3 are as reported in press coverage from January to October 2026, and the gap between the £35bn and £11.5bn estimates is unresolved; see the note at the end before treating either as final.
1. Why Bank of England Independence Matters
On 6 May 1997, four days after the general election, the new government gave the Bank of England operational independence over interest rates.1 The decision came without any manifesto commitment beforehand, reportedly for fear the opposition would claim it meant higher rates. It also wasn't full independence. The Chancellor still sets the inflation target; the Bank's job, through the newly created Monetary Policy Committee, is to hit it, and to explain itself publicly if inflation misses by more than a percentage point either way. The government keeps the goal and the Bank gets the tool.1
The case for that split rests on what came before. Before 1997 the Chancellor set interest rates, and they were repeatedly used for short-term political ends. The clearest example is the late-1980s Lawson boom, when low rates and looser credit fed a housing and spending boom around the 1987 election, followed by a sharp bust in 1990–92.1 That instability led into Black Wednesday on 16 September 1992. The UK had joined the Exchange Rate Mechanism in 1990, committing to hold sterling within a band against the Deutschmark, with inflation then running at roughly three times Germany's. On Black Wednesday it raised rates from 10% to 12% and announced a further rise to 15% the same afternoon, a rise that never took effect before the UK left the ERM and let sterling float.1 The immediate result was a reputational disaster for the government of the day. The medium-term result, once free of the ERM, is widely seen as positive: rates fell to 7%, the pound devalued, and together those are credited with helping the 1993–94 recovery. The line from that crisis to 1997 is direct. Within weeks of Black Wednesday the government adopted an explicit inflation target, with the Bank publishing its own inflation assessment, though the Chancellor still set rates. Five years later the Bank was given the rate-setting tool as well, bringing the UK into line with the Bundesbank and Federal Reserve model.1
An analysis by the San Francisco Federal Reserve, published within months of the 1997 announcement, found the change shifted market expectations of future UK inflation, a direct measure of the credibility independence was meant to produce.2 Two further points matter. First, supporters of independence have criticised particular Bank decisions since 1997, Gordon Brown among them over the Bank's handling of the 2007–08 crisis, without concluding that the design should change; criticising a decision is different from attacking the mechanism behind it. Second, independence has kept evolving rather than staying frozen: the Financial Services Act 2012 created the Financial Policy Committee in April 2013, giving the Bank a remit for the stability of the financial system as a whole that the original settlement didn't include.3
2. How QE Created This Bill
Between 2009 and 2021 the Bank of England ran quantitative easing, buying about £895bn of bonds from the financial system: £875bn of government gilts and the rest corporate bonds.4 It paid for them not with cash but by crediting the sellers' banks' reserve accounts at the Bank, an electronic balance rather than physical money. The Bank pays interest on those reserves at Bank Rate. That isn't a subsidy in origin: it is how Bank Rate is passed through to the wider economy. If reserves earned nothing, banks would lend them out below Bank Rate, and the Bank would lose control of the rate it is trying to set.5
The reason the cost has become a political issue is simple. During QE, Bank Rate was close to zero, so the interest on hundreds of billions of pounds of reserves cost very little. The rate rises from 2022, peaking at 5.25%, turned an almost costless mechanism into a bill of billions of pounds a year. Because HM Treasury covers any losses on the Bank's QE operations, this is a real cost to the public finances, not an abstraction on the Bank's balance sheet.5
3. The Proposal: What Would Change
Reform UK has proposed stopping interest payments on QE-created reserves altogether, claiming a saving of £35bn a year. It first set this out in 2024 and repeated it in January 2026.46 It presents the idea as different from a bank tax: the argument is that this is income banks never earned, rather than money taken from them. Its backers claim support from figures at the Financial Times, the New Economics Foundation, the IFS and two former Bank of England deputy governors. This paper hasn't been able to confirm which individuals that refers to, or in what form they gave support.
The £35bn figure is contested. An independent fact-check found a more defensible estimate of about £11.5bn, with the gap depending heavily on which reserves would be exempted and how any tiered system would work.7 The proposal also sits within wider political pressure on the Bank, which separately includes calls to halt quantitative tightening (the Bank's sale of bonds) on cost grounds, and to stop work on a digital pound.8
4. International Comparison
Paying interest on only part of banks' reserves is not a fringe idea. Since September 2023 the European Central Bank has paid 0% on the minimum reserves banks are required to hold, equal to 1% of certain liabilities, mainly customer deposits, while continuing to pay its deposit rate on reserves above that minimum. The ECB described the change as reducing the interest it pays without weakening monetary policy.9 That is a far more limited version of the idea than the "stop it entirely" proposal above: interest is still paid on the bulk of reserves, which is what keeps market rates anchored to the policy rate.
This paper hasn't yet looked in detail at how the Federal Reserve or the Bank of Japan treat reserves, or at the ECB's earlier tiering system, so no figures for them are given here.
5. Steel Man: The Case for Reform
The strongest case for change doesn't rely on "free money" rhetoric. Banks didn't choose to hold these reserves for the return; they are the mechanical result of the Bank buying bonds from the financial system under QE, which some economists reasonably describe as a windfall rather than an earned commercial return. A partial reform, such as an unpaid minimum reserve requirement like the ECB's, is a real, less disruptive policy already in use, so the choice doesn't have to be between paying Bank Rate on every pound of reserves and paying nothing. And the sums involved, tens of billions of pounds in total since rates rose, are large enough that even a partial reform would matter for the public finances compared with other revenue and spending proposals on this site.
6. The Case Against
The Bank's own objection, made by Governor Andrew Bailey, isn't institutional defensiveness. Paying nothing on reserves risks breaking the mechanism that holds market interest rates at Bank Rate, which is the whole point of the tool.5 A second risk is that the saving doesn't disappear so much as move. Banks losing income on reserves have an obvious way to recover it, through lower savings rates, more expensive lending or higher fees. The cost would then fall on customers rather than being removed.
A third risk concerns how markets read the Bank's independence. Opponents of the proposal, including other political parties and market commentators, see it as eroding the Bank's independence in practice, whatever it is called. A Rabobank note warned that if markets treat reserve-interest reform as a threat to that independence, the result could be steeper gilt yield curves, a weaker pound and weaker performance in shares sensitive to interest rates. A saving on one line could then be partly offset by higher government borrowing costs on another.10 Supporters of the proposal themselves point to the 2022 mini-budget as a warning, arguing that any reform that unsettles markets needs to come with credible fiscal discipline. That leaves an unresolved question: would reserve-interest reform on its own trigger a similar market reaction, whatever spending package came with it?
7. Two Possible Reforms, Not Yet Decided
A. Full stop: the proposal as currently framed. Stop paying interest on QE reserves entirely. The largest claimed saving (a contested £11.5bn to £35bn), and the largest risk to how interest rates are passed through and to the Bank's credibility.
B. Partial reform: an ECB-style approach. Pay no interest on a required minimum level of reserves, and Bank Rate on the rest. A smaller, more defensible saving; keeps rate transmission working on most reserves; harder to portray as ending central bank independence; and with a direct international precedent.
How large a UK minimum reserve requirement would be, and what it would save, hasn't been worked out yet. The £35bn and £11.5bn figures both apply to the "full stop" version and shouldn't be used as a stand-in for what a partial reform would save.
8. Links to the Rest of the Site
Whatever figure proves defensible, the saving claimed for this proposal needs to sit alongside this site's other revenue proposals, including the Two-Jar Fiscal Framework and the Revenue Architecture pillar, so the overall fiscal picture doesn't double-count a contested saving that may not fully materialise. Section 1's history is also the clearest case study on this site of shielding a technical function from short-term political pressure, which is directly relevant to the Public Office Covenant's wider argument for independent institutions, in the same spirit as judicial independence or the independence of the Office for National Statistics.
Sourcing note: the 1992–1997 independence history in Section 1 draws on the Bank of England's own retrospectives and the San Francisco Fed's 1997 analysis, and is treated as reliable. The QE and reserve-interest figures in Sections 2 and 3 reflect press reporting from January to October 2026. The gap between £35bn and £11.5bn should be resolved against a primary source, such as the Bank's annual report or an OBR costing, before this note moves out of "in discussion" status.
For public discussion. Not affiliated with any political party. | generationalreset.org
The Generational Reset | In Discussion: BoE Reserve Interest | For public discussion. Not affiliated with any political party. | generationalreset.org