The UK Labour Party is fighting in public, generating uncertainty over Sir Keir Starmer’s future as prime minister and the direction of British politics. This is undermining sterling and raising government borrowing costs.
This week I want to bring you up to speed with the policy fights at the Bank of England and the uncertainty these public disagreements have cast over monetary policy and interest rates. The entirely calm reaction reveals the UK economic community as more grown-up than its political counterpart, and demonstrates that policymakers can find a way to air their differences in public without the sky falling in. This is a good news story.
Last Thursday, the BoE decided to leave interest rates on hold at 3.75 per cent. That decision was expected, but the five-to-four vote against a cut was closer than I, or pretty much anyone else, expected. For my part, I expected the Monetary Policy Committee to take more notice of its own forecast evaluation report, which found that it had underestimated inflation persistence, as well as the economic data so far this year, which has been broadly positive.
Why was the vote so close? One short answer is that, although the BoE staff have turned rather dovish and would have voted to cut interest rates if they had a say, they could not persuade five of the MPC’s members.
We know this because since November, the central bank has moved to a full transparency mode of communicating, with each MPC member given 200 words to explain their vote in public. These paragraphs tend to explain individual votes with reference to the topical issues the BoE staff have outlined in the quarterly Monetary Policy Report.
There were five issues outlined. I asked ChatGPT for a short summary. It was reassuringly rubbish, so this is mine:
-
Box A: The level of UK wage growth consistent with the 2 per cent inflation target is about 3.25 per cent
-
Box B: There have been few changes in the process of UK wage-setting, so MPC hawks should pipe down a bit
-
Box C: The productivity outlook has improved a touch, which is, on balance, bad news for MPC hawks
-
Box D: The effect of fiscal policy on the likely level of slack in the economy is incredibly uncertain but quantitatively important
-
Box E: Interest rate-sensitive parts of the economy are still being hit, so it is likely interest rates at 3.75 per cent are still restrictive
MPC members viewed the staff analysis as being decidedly dovish. Among the hawks who voted to hold rates, I counted five dissents from the staff analysis, five neutral references and one extra dissent by implication. Among the doves, there were five approving references, six neutral and one sign of implicit support.
Box A got many neutral references, as it constituted helpful and uncontentious analysis of the current level of wage growth consistent with 2 per cent inflation. This depends on productivity and import-price assumptions, but no one felt the need to get into a fight about the calculation.
Box B delighted the MPC’s doves and was criticised by the hawks. It examined whether the Covid-19 pandemic and post-pandemic inflation surge had changed the wage-setting process to add persistence to inflation.
In what appears to be detailed and difficult microdata work on a one in 100 annual sample of the same UK employees, it grouped employers into four types by the way they set pay: “reputational firms”, “incentive payers”, “cost minimisers” and “bargaining firms”. It found that the size of each group had not changed after Covid-19, concluding that “this suggests that the pandemic and the subsequent period of high inflation have not induced many UK firms to fundamentally change their approach to wage setting”.
This helped assuage MPC doves’ fears that high wage growth would lead to persistently high inflation.
I will be blunt here. The UK wage-setting process might well have stayed the same. But concluding from unchanged firm cluster sizes that wage-setting has not changed is an astonishing non sequitur. I am not in the least bit surprised that BoE chief economist Huw Pill very politely wrote that he thought there had been “structural changes in wage bargaining and productivity, notwithstanding the conclusions in Box B”, adding that he drew “different implications from the careful and important analysis”.
The underlying staff work is good and shows that the cost-minimising group of employers, who hire many minimum wage employees, still had high median wage growth in April 2025, related to the 6.7 per cent rise in the legal minimum. The “bargaining firms”, which are often unionised and in the public sector, also had high wage growth in April 2025. We do not know yet if these will sufficiently moderate, because the survey is undertaken once a year. The jury here is still out.
Box C related to productivity growth and generated some MPC dispute. First, the staff looked back at recent productivity trends, which were weak and volatile. Then they looked forward and were a bit more optimistic.
The forward-looking analysis was mostly educated guesswork, so it is entirely reasonable that the hawks should disagree, as Clare Lombardelli and Megan Greene did. Sarah Breeden, who voted for a cut, also expressed some scepticism, saying that “potential productivity growth could prove weaker than expected (Box C)”.
Most of the MPC mentions about Box D on fiscal policy were neutral and related to the regulatory changes that will mechanically bring down UK inflation, probably to around the 2 per cent target in the April data.
More interestingly, the staff recognised that, despite the UK’s plans for a big fiscal tightening, they could not be certain of the scale or the timing of its effect on the output gap. Although the BoE did not quite highlight this in its modelling, there is a big difference between the effects of a 1.6 per cent of GDP output gap in 2026-27 and a 0.2 per cent one on inflation and the economy. The former assumes tax rises depress activity more than standard assumptions, while the latter assumes tax rises take time before they affect consumer spending behaviour. The implicit scenarios in the chart below would have a much larger effect on forecast inflation than the formal scenarios examined by the BoE.
In Box E, the BoE examined the effect of higher interest rates on consumption and investment. As the chart below shows, staff found that higher rates had much more impact in curbing interest rate-sensitive elements of consumption and investment, and that the effects were ongoing. This suggests interest rates at 3.75 per cent are still restrictive, although they are clearly not holding activity back as much as when they were 5.25 per cent in 2023 and 2024.
The questions that some of the hawks raised, both explicitly and implicitly, was whether the data shown below was out of date, and whether that meant rates were no longer restrictive, even if interest rate-sensitive parts of the economy were still weak.
Overall, the important thing about the latest BoE Monetary Policy Report is that the staff had a view, MPC members commented, often critically, and it was all done in public. Nothing bad happened.
You can be cynical and say “if this is the disagreement aired in public, just imagine the dreadful fights going on behind closed doors”. I am not going to do that.
Last week, the BoE demonstrated how to disagree politely and properly in public. So long as people understand this is all in the pursuit of its monetary stability remit, that is what we want from powerful unelected economic institutions.
What I’ve been reading and watching
-
Banque de France governor François Villeroy de Galhau announced he would quit his role early in June to chair a children and social inclusion charity. This will allow President Emmanuel Macron to appoint a successor before the 2027 election.
-
Shrinking the Federal Reserve’s balance sheet, as desired by nominee Kevin Warsh, will be difficult, Andrew Whiffin concluded for Monetary Policy Radar.
-
In further bad news for Warsh, academic economists are sceptical that an AI boom will enable interest rate cuts.
-
The European Central Bank still thinks it is in a “good place”.
One last chart
Everyone has engaged large language models and the words of Warsh to produce hawk-dove scales. Some say he is just hawkish. Others, such as Nobel laureate Paul Krugman, call him a weathervane, in that he is in favour of tight monetary policy when the Democrats have the White House and loose policy when the Republicans are in charge.
In our chart, produced by my colleague on MPR, Joel Suss shows that Warsh has been generally more hawkish at most times. But his words were much more dovish than the Fed average at the start of Covid-19, as well as when he campaigned for the nomination last year.
You can check for yourselves. Click on the chart and hover over the dots and you can see the exact words that are shifting the AI algorithm. Most, but not all, make a lot of sense.
Central Banks is edited by Harvey Nriapia