Everyone misunderstands Jay Powell

The headline reflects some rather interesting research about the Federal Reserve that shows how difficult central banks find landing their expert and technical messages with the public.

For example, however hard Fed chair Jay Powell tries to say absolutely nothing controversial in a polarised US, people often hear news mediated through a partisan outlet and take offence, assuming he is speaking for the other side.

The research by Pei Kuang, Michael Weber and Shihan Xie sampled more than 5,000 people just after Donald Trump was inaugurated in January. It finds that the public is not only extremely polarised, but that this colours their view of the Fed. Democrats thought US inflation was moderate in 2024 but expect it to be high in 2025; Republicans think the opposite.

As such the results are rather like the University of Michigan consumer sentiment index, below, which shows consumer confidence is governed more by who is running the country than economic events. This is troubling because economic surveys then become difficult to interpret.

What is more difficult for the Fed is the public’s view about the institution’s impartiality and its ability to take expert decisions. Some 66 per cent of people identifying as Democrats thought the Fed was run by a bunch of Trump-loving Republicans and 60 per cent of Republicans thought the Fed was a hotbed of woke Democrat lefties.

This means that most Americans are not inclined to trust the Fed. Interestingly, those identifying as independents are not some non-partisan oasis of trust, but the most sceptical group of all. In total, almost two-thirds of respondents thought the Fed was essentially out to get them.

That result is not necessarily disastrous, since independent central banks are supposed to take difficult decisions to foster low inflation and financial stability. The problem is that trust in the Fed is governed by how much people think it is aligned with their thinking.

Trust matters. Powell mentioned the importance of anchored inflation expectations eight times in his press conference last week. Success relies on people trusting the Fed will keep inflation low.

In 2023, European Central Bank President Christine Lagarde said trust was extremely important. “Conveying our policy messages to the wider public is critical for the legitimacy of independent central banks in democratic societies and for the effectiveness of monetary policy.”

The problem for the Fed is that those who thought the Fed was against them were much less willing to trust it.

The good news in the research is that when respondents were presented with basic information about the appointment process at the Fed, its objectives or recent record, trust scores went up significantly, as did the survey respondents’ perception of the Fed as an independent institution acting in the interests of the US.

The bad news is that the survey participants were forced to consume this information. And the research found that those who did not trust the Fed preferred partisan news sources rather than factual information. So, however factual and impartial the Fed is, it has to reach people who prefer to have their priors reinforced by partisan outlets.

I got in touch with one of the authors, Michael Weber at the University of Chicago. His conclusion: “Communication can be successful in winning people over . . . [but] many consumers do not want to learn about Fed policy.”

A sermon on hawkishness and dovishness

With apologies to anyone who has heard me bang on about this in the past, last week’s Fed meeting stuffed my inbox with analysts disagreeing on whether the Fed had been hawkish, dovish or neutral.

None of the people emailing me were idiots, but in describing economic events or data, anyone in the communications game needs to answer one basic question. Is it big or small? In this case: is it dovish or hawkish? And to do that you need a relevant comparison. You need to answer the “relative to what?” question.

Given this framework, there are three relevant comparisons to last Wednesday.

Compared with the Fed’s December meeting, the decision and the summary of economic projections were clearly hawkish. Growth was revised down and inflation revised up. Alongside these contradictory moves, the centre of gravity in the Federal Open Market Committee shifted towards expecting fewer rate cuts than previously, even though the view of the median FOMC member was unchanged. Mostly, this came from some officials removing their view of rapid interest rate cuts, as the chart below shows.

Compared with financial market expectations, the meeting and forecasts were also hawkish. Before the meeting, bond and interest rate futures markets expected roughly three rate cuts in 2025, but Fed officials suggested a maximum of two. Given this, financial markets should have seen the decision as hawkish, but weirdly took the meeting as dovish. There is an inconsistency here. Never mind.

Compared with a view that tariffs made rate cuts impossible, the meeting was clearly dovish, since the Fed still signalled two rate cuts as most likely this year.

The meeting was therefore either hawkish and dovish depending on the relevant comparison. Before pronouncing, you need to articulate your frame of reference.

Here endeth my sermon.

What I’ve been reading and watching

• Read Barry Eichengreen’s Saturday essay questioning whether the dollar can remain king of the currencies. No one can ever say a straight “no” to this question, but Eichengreen comes close and melds wonderful history and facts doing so.

• So desperate is everyone to think there is some grand plan behind Trump’s economic policy that there has been an explosion of Mar-a-Lago Accord explainers (even though a plan does not even exist). To save yourself the trouble of many takes, read Martin Wolf’s two-part series. The first lays out the theory. The second, publishing as this newsletter goes out, Martin promises will trash it.

• President Recep Tayyip Erdoğan’s detention of his political rival shows where authoritarian tendencies lead. It is also bad for Turkey’s economic prospects.

• If you want to know roughly what will be in the UK’s Spring Statement on Wednesday, I wrote about it in a column last week.

A chart that matters

Younger employees are twice as likely to be using artificial intelligence at work than their older colleagues, the European Central Bank reported last week. Regardless of age, attitudes towards AI were overwhelmingly positive, especially among those who used it.

The Bundesbank also highlighted last week that artificial intelligence was pretty good at tracking dovish and hawkish statements from the European Central Bank (see chart below). It does not reflect a dovish bias at the bank, but an appropriate dovish outlook when inflation was below target in the 2010s and more hawkish noises before the ECB started raising rates.

The even better news is that my colleagues and I over at the FT’s Monetary Policy Radar independently undertook a very similar analysis in December last year and got consistent results. AI is proficient at reading and synthesising a large amount of information. Chapeau to the robots.