Kevin Warsh’s first assignment to redesign the workings of a central bank began with a British lawmaker wondering whether its officials were shouting at one another.
In March 2014, Andrew Tyrie, chairman of the House of Commons Treasury Committee, asked Bank of England officials what happened to the recordings of their monetary-policy meetings. Paul Fisher, the Bank’s executive director for markets, explained that the meetings were recorded to help produce the official minutes. Once the minutes were finished, the recordings were destroyed.
Tyrie was incredulous. The discussions had “huge historical significance,” he said. The Federal Reserve managed to publish transcripts of its meetings after a delay. Why couldn’t the Bank?
When Fisher suggested that the free-flowing discussions were difficult to transcribe, Tyrie asked whether committee members were shouting and throwing things at one another. Most organizations, he observed, somehow managed to produce a record of their meetings.
Seven weeks later, Bank of England Governor Mark Carney appointed Warsh, a former Federal Reserve governor, to conduct an independent review.
The resulting document, published in December 2014, now reads like a 64-page guide to the institutional changes Warsh is considering as chairman of the Federal Reserve. The report says little about where interest rates should be set. Its subject is how central bankers should think, argue, decide, and explain themselves.
That makes it especially relevant today. Warsh has begun reconsidering the Fed’s meeting schedule, communications, and internal policy process. Minutes released Wednesday show that he asked colleagues whether the Federal Open Market Committee should hold six rate-setting meetings a year instead of eight, allowing two full months of economic data to accumulate between decisions. One version under consideration would preserve two additional meetings for deeper discussion of important economic questions.
Warsh confronted a similar problem in Britain 12 years ago. He ultimately concluded that a good central bank needed two distinct kinds of meetings: one for deliberation and another for decision.
The distinction came from listening to recordings of the Bank’s Monetary Policy Committee. Its monthly meeting was then conducted over two days. On the first day, members worked through financial markets, the international economy, domestic demand, employment and inflation. The conversation was informal. Members interrupted one another, tested competing explanations and pursued puzzles in the data.
Warsh described this as “inquiry.” Its purpose was to discover what members should think.
The second day was more formal. Members had absorbed the staff analysis and heard their colleagues’ arguments. They stated their preferred policy, assembled the evidence supporting it and attempted to persuade the rest of the committee. Warsh called this “advocacy.” Its purpose was to decide what the central bank should do.
The difference mattered because transparency changes behavior. Officials who know that every speculative remark will eventually become public have an incentive to avoid uncertainty, polish their interventions and defend positions they previously expressed. The meeting begins to resemble a series of speeches delivered for the historical record.
Warsh had seen this happen at the Fed. After Congress discovered in 1993 that the FOMC maintained transcripts, the central bank agreed to release them after five years. Research subsequently found that officials increasingly relied on prepared remarks. Some of the real discussion migrated into private conversations before the formal meeting.
The transcript policy made officials prepare more carefully, which Warsh regarded as valuable. It also made them more cautious about disagreeing, appearing mistaken or changing their minds. The challenge was to preserve the discipline produced by eventual disclosure without allowing it to suppress the argument needed to produce sound decisions.
His answer was to protect the first stage and expose the second.
Warsh recommended that the Bank stop recording the first day of deliberations. The published minutes should instead provide a richer account of the competing hypotheses, unresolved questions and weight members placed on different interpretations of the evidence.
The second day should be recorded and transcribed. Those transcripts should be published after five to 10 years, accompanied by the staff materials used to reach the decision. Individual votes and an explanation of the committee’s decision should be released immediately.
This was a carefully constructed division of responsibilities. Deliberation required privacy because officials needed room to be uncertain. Decisions required accountability because officials exercising public power had an obligation to explain what they had done.
Warsh also recommended reducing the Bank’s policy meetings from 12 a year to eight. Monthly meetings consumed staff resources, encouraged excessive attention to noisy data and left too little time for reflection. If a crisis required action between scheduled meetings, the governor could call an additional one.
Carney accepted the recommendations immediately. The Bank chose an eight-year delay for transcripts, began publishing decisions and votes with the supporting minutes, and agreed to release the relevant staff papers. It also supported moving to eight scheduled policy meetings.
The political reception was unusually strong for an outside review of a central bank. Chancellor George Osborne said the recommendations would place British monetary policymaking at the forefront of international practice. The government incorporated the reforms into the legislative program accompanying the 2015 Queen’s Speech. Parliament subsequently changed the law through the Bank of England and Financial Services Act of 2016.
The reforms endured. In 2024, the Bank published the first batch of eight-year-old transcripts and staff materials from meetings held in 2015.
There were objections. Some economists warned that releasing the decision, minutes and quarterly Inflation Report simultaneously would bury the public in information. Others worried that stretching the meeting process across several days would increase the risk of leaks or leave members debating an outlook overtaken by new data.
The central distinction between deliberation and decision nevertheless survived. Warsh has continued to embrace it.
In a 2023 interview reported by Reuters this year, Warsh said the Fed’s tape recorder still “looms large” over its meetings. Officials worried about looking foolish with the benefit of hindsight tend to hedge their arguments, he said. Warsh wanted the initial debate to resemble a “family fight,” with members presenting the strongest case for competing positions. The final decision round should be recorded and eventually released so each member could be held responsible for his judgment.
His early moves at the Fed follow the same pattern. One of the five outside task forces he created is examining how the Fed communicates “policy deliberations and decisions amid uncertainty.” Its members include Mervyn King, the former Bank of England governor. Warsh is considering fewer rate-setting meetings while creating more space for extended discussion of broader economic questions.
The British experience does not provide a precise roadmap for what Warsh plans for the Fed. The FOMC is larger, its institutional traditions are different, and Warsh’s views have continued to develop in the 12 years since his Bank of England reform proposals. But the report reveals a durable conviction about how central banks go wrong.
Officials need a place where they can test an argument without committing themselves to it. They also need a separate moment when uncertainty ends, a judgment is rendered and responsibility attaches.
Warsh first put that principle into practice at the Bank of England. Twelve years later, he appears ready to try an updated version of it at home.