This week, Jay Powell, Federal Reserve chair, is in the spotlight — and pulpit. A year ago, he delivered a speech at that gigantic economic tribal gathering known as the US Federal Reserve’s Jackson Hole conference, pledging that inflation pressures would only be “transitory.”
On Friday morning he speaks again — and it is now clear that this “transitory” message was totally wrong. It remains uncertain if he will apologise, or admit the mistake. But what is clear is that any numbers Powell cites will prompt investors to reappraise their macroeconomic models and forecasts.
Fair enough. However amid this frenzy of number-crunching, investors should also take note of some intriguing research floating around the edge of the Jackson Hole meeting about the importance of storytelling in monetary policy.
A couple of decades ago, this was a topic garnered that little attention at the Fed. For while anthropologists such as Douglas Holmes have long studied how central bank language and rituals influence the economy, economists used to assume this was less important than mechanistic models.
However, after the 2008 financial crisis, Robert Shiller, the Nobel Prize-winning economist, urged his colleagues to study how “narratives” shape sentiment and thus economic trends. And one encouraging post-crisis development in economics is that a swelling number of young behavioural finance economists have heeded Shiller’s call.
Last year a team spearheaded by Peter Andre surveyed 10,000 households and 100 economists and concluded that whereas economic “experts” attribute price jumps to cyclical demand swings, consumers blame supply issues such as government mismanagement. That perception, which is heavily influenced by media, causes consumers to project higher inflation for longer.
This year another group led by Yongmiao Hong used machine learning tools to analyse inflation narratives in 880,000 Wall Street Journal articles. They concluded that “narrative-based forecasts perform better in the long-run forecasting” than numerical models, seemingly because the latter miss subtle economic signals and shifts.
And this month Chad Kendall and Constantin Charles released research about extensive psychological experiments into how humans create explanatory frameworks to frame economic data. This shows that people almost always seek to fit new information into pre-existing narratives.
But what is most interesting is that the shape of our storytelling matters, since “narratives with a particular structure may affect people’s actions by influencing the subjective beliefs they form from the data they observe”. More specifically, the experiments suggest that “lever” narratives, which present linear causality frames (A leads to B leads to C, and so on) have the strongest grip on people’s minds. So-called threat narratives, which describe how offsetting forces both avoid and cause particular outcomes, are less potent. The research also notes that people love “to share their homegrown narratives with other subjects, who are then persuaded by them”.
All of this has practical implications for Powell, given that the Fed faces an increasingly bitter fight over inflation. Some Fed critics, such as former US Treasury secretary Lawrence Summers, are currently promoting one causality narrative — that this year’s spiralling inflation arose because of lax monetary policy. This implies that interest rates must rise to stop inflation.
Fed officials, however, prefer another causal story according to which prices have jumped because of high energy prices and supply side shocks. This suggests that prices will fall if (or when) the initial shock of the war in Ukraine subsides, and/or economic activity slows down.
There is a political tussle too: Republicans blame the Biden administration for inflation; Democrats blame it on external events. One consequence is that Pew research suggests that 84 per cent of Republicans think inflation is “a big problem”, but only 57 per cent of Democrats agree. This is a striking demonstration of why narrative causality matters.
Of course, an intellectually honest economist might note that all these simplistic causal stories are essentially fictions, given that economic phenomena arise from complex interactions in the economy. The causality narrative being presented by the White House around its Inflation Reduction Act is also somewhat fictitious. While some measures in the legislation are sensible, they are unlikely in themselves to influence short-term price trends much, if at all.
But Powell’s problem is that if he does not present a convincing causality narrative at Jackson Hole, others will offer one in its place. Simply describing what has happened in the past year, or proffering complex and competing explanations, is unlikely to pack much of a punch — or shape sentiment in the way that the Fed needs.
So if I were his speechwriter, I would take a leaf from Paul Volcker’s book and argue that inflation will fall when rates go up, and promise to keep raising those rates until price growth is at a sensible level. That will not be politically popular ahead of a crucial midterm election. But it is at least a crystal clear message — or it would be if the Fed actually does it.
gillian.tett@ft.com
