WASHINGTON, May 6 (Reuters) - The release of transcripts of Federal Reserve rate-setting meetings, a cornerstone of its transparency for more than 30 years, undermines the debate needed to set good monetary policy, incoming U.S. central bank chief Kevin Warsh says in an upcoming book, remarks that echo his wider desire to overhaul the Fed.

Warsh, in a 2023 interview with New York University Stern School of Business Professor Simon Bowmaker, advocated limiting the taped sessions and published transcripts to a final "decision round of discussions" where officials can spell out the rationale for their policy votes.

Policymakers "do not want to appear wrong with the benefit of hindsight, and so they instinctively tend to hedge their bets" when their comments are taped for release, Warsh told Bowmaker for his book, "Fed Reckoning: Conversations on America's Central Bank," to be published early next year.

Warsh, who served as a Fed governor from 2006 to 2011, went into detail about a study he conducted for the Bank of England, where "as a result of the work I did in 2014, the recording device has been turned off" for the initial round of policy meetings to allow for more free-flowing discussions, while offering more public disclosure of the actual policy votes by releasing transcripts of the day-two proceedings.

"The tape recorder, however, still looms large at the Federal Reserve ... If we want that deliberation to be robust, we need a family fight. If people think the decision is 60-40 one way, I would prefer them to argue as if it were 95-5. I want to hear the best arguments," Warsh said in the interview, which was shared with Reuters.

The Fed since the early 1990s has released full transcripts of its two-day sessions after a five-year delay, with the lag considered a compromise between public transparency and the concern that quicker release would stifle debate.

Though U.S. central bank officials have unvarnished discussions with each other - outgoing Fed Chair Jerome Powell is known for his extensive canvassing of colleagues before meetings - "it would be preferable if the fierce deliberation happened among a larger group. So, if you want sound policy decisions, you have to create an environment in which sound policy fights can happen," said Warsh, who is expected to be confirmed as the next Fed chief by the U.S. Senate this month.

The 56-year-old lawyer and financier said the second day of discussions, by contrast, "should be recorded," and "the transcript should be made available because it is a judgment of what each member believes and his rationale. The historical record should ensure accountability for the decisions."

REVERSING THE TRANSPARENCY TREND?

Warsh did not respond to questions from Reuters about changes he may seek in the Fed's handling of transcripts or other communications policies that he recently told the Senate Banking Committee needed retooling to encourage "messier meetings" and "a good family fight."

But his comments for Bowmaker's book echo longstanding critiques Warsh has made about Fed communications he feels were an obstacle to responding faster to rising inflation following the COVID-19 pandemic.

Changing the policy on transcripts would rehash a controversial moment for the Fed. In the early 1990s it was revealed that, unbeknownst even to many at the central bank, meetings were not just recorded - a longstanding practice to help render minutes, with the tapes then erased - but that since 1976 the recordings had been transcribed and the transcripts retained.

The revelation drew comparisons to former U.S. President Richard Nixon's secret White House recordings, and added to pressure building among Democrats then in control of Congress for the Fed to change the aura of secrecy around its crafting of monetary policy.

Former Fed Vice Chair Donald Kohn, a long-serving U.S. central banker involved as a staff member in the initial discussions about how to treat the transcripts, said that along with the value to historians, the recordings are important in the creation of meeting minutes. While noting that Warsh's point about Fed debates isn't inaccurate, Kohn said policymakers' greater reliance on prepared statements after the decision to release the transcripts meant that their level of preparation also increased.

"Did it impede discussion? Yes, to some extent," Kohn said, but the practice among recent chairs of hearing out policymakers ahead of time "takes away spontaneity too ... His thing about messier meetings suggests there may be less of that beforehand."

Another change could involve the press conferences that Powell has held eight times a year after each policy meeting - his immediate predecessors, Janet Yellen and Ben Bernanke, held them on a quarterly basis, while former Fed Chair Alan Greenspan held none.

Warsh also may seek to curtail or eliminate the release of the Fed's quarterly economic projections, which he feels are a form of constraining "forward guidance." During his April 20 confirmation hearing he did not rule out reducing the number of meetings, which the law permits to be as few as four annually.

Fed communications "are not a light switch, on and off, it's a dial. Powell was incredibly transparent ... Should Warsh be confirmed, it will be turned down a few notches," said Michael Arone, chief investment strategist at State Street Investment Management. "As a consumer of information, more is better than less. It would increase the risk of misinterpretation. It may increase the volatility around expectations."

The release of the first transcripts marked the start of steadily expanding efforts by the Fed to both de-mystify its arcane world and leverage the power of public communications to improve policy effectiveness. The better central bankers explain themselves, and the more they are trusted, the faster and more effectively their rate decisions will be felt in the economy, the argument goes.

Dialing back what's recorded and released would not just reverse the usual bias towards greater disclosure, which for the Fed has included more expansive policy statements, regular press conferences, and frequent public speeches by Fed officials. It might also resurface some of the old suspicions about the Fed at a time when U.S. President Donald Trump seeks more influence over it, said Sarah Binder, a political science professor at George Washington University and author of a history about the Fed and politics.

Changes on disclosure "are hard to take back ... The big, broad movement at the Fed is from very little transparency to a pretty broadly transparent institution where the members like to explain themselves because they think it is helpful in setting expectations," Binder said. "The moment it becomes known that they are turning off the tape recorder, suspicions grow. How did they reach that decision? People's minds can go pretty conspiratorial."