To a large and under-appreciated extent, the job of the US Federal Reserve entails chasing an elusive number: r*, or the neutral short-term interest rate. When the Fed’s target rate is above r*, it should restrict growth. When it’s below, it should stimulate economic activity.

I think r* is a lot higher than the Fed recognizes — which means the central bank isn’t doing enough to fight inflation.

R* isn’t directly observable. It must be inferred from how the economy responds to short-term interest rates. This isn’t easy, for three main reasons. First, the impact of short-term rates depends on other financial phenomena such as long-term rates, stock prices and credit spreads – all of which vary considerably on their own. Since October, financial conditions have eased significantly even as the Fed has held short-term rates steady. Second, rate changes operate with long and variable lags: It’s possible, for instance, that the full effects last year’s Fed tightening have yet to be felt. Third, there’s always a lot going in the economy other than monetary policy — such as, right now, a wave of investment in artificial intelligence.

Still, r* is too important to ignore. A judgment must be made, to assess where the Fed stands and what it needs to do. And lately, officials’ approach hasn’t been compelling.

Chair Jerome Powell has mostly pushed back against r* questions at his press conferences. The Fed’s median projection of r* (adjusted for 2% inflation) has hardly budged in years, finally moving up 10 basis points to 0.6% in March. Governor Christopher Waller inexplicably argued that r* mainly depends on the demand for safe assets such as US Treasuries. Formal econometric models that estimate r* can’t quickly incorporate the change in the post-pandemic economic environment: Data from recent years are too noisy and provide too little information, as efforts to update the Holston-Laubach-Williams model have shown.

There’s a strong case that r* has risen substantially. For one, the persistent strength of the US economy suggests that monetary policy isn’t very restrictive. The relatively slow growth in the first three months of 2024, an annualized rate of 1.6%, understates the true momentum. One measure of underlying demand — real final sales to private domestic purchasers — rose 3.1%.  The Atlanta Fed’s GDP Now model forecasts 3.5% growth in the second quarter.

Also, various factors are driving down desired saving and boosting desired investment, which in turn pushes up r*. On the saving side, high stock prices are making people feel more inclined to spend, baby boomers are tapping their retirement funds and the US government is borrowing vastly (the opposite of saving) to fund its budget deficits. On the investment side, the Biden administration has jump-started capital expenditures on everything from chip factories to green technology, and renewable energy (wind and solar) is much more capital-intensive than coal or natural gas.

Put it all together, and r* could be as high as 2%, which was the conventional wisdom prior to the 2008 financial crisis (embodied in the Taylor Rule, a mechanical estimate of the optimal federal funds rate). If so, the neutral short-term rate currently would be about 5% (a 2% r* plus 3% inflation), meaning that the current fed funds rate of 5.25% to 5.50% is exerting negligible restraint on growth and inflation.

Perhaps the Fed’s mantra, instead of “higher for longer,” should be “higher indefinitely” until inflation moves more convincingly in the desired direction.

Bill Dudley, a Bloomberg Opinion columnist, served as president of the Federal Reserve Bank of New York from 2009 to 2018. He is the chair of the Bretton Woods Committee, and has been a nonexecutive director at Swiss bank UBS since 2019.