A quarter of a century ago, the Nobel laureate economist Robert Shiller organised a survey-based study of consumers’ perceptions of inflation. It drew attention to three points that are highly relevant now.

First, ordinary people do not view prices like economists. Emotions, not just economic data, matter. Second, the emotions in question, such as anxiety and anger, are powerful but hard to track in models.

Third, these sentiments can shape perceptions. Shiller found that back then inflation was the most widely cited economic term in the media, even outstripping references to “sex”.

Conditions today seem different. Shiller conducted his research soon after the 1970s oil shock and stagflation. But Federal Reserve officials and investors should ponder his findings and consider repeating his research.

Consumer prices are rising at the fastest rate since 2008, prompting the Fed committee to reveal this week that it has raised core inflation projections for 2021 from 2.2 to 3 per cent. It is now signalling future policy tightening.

Unsurprisingly, this has sparked debate about hard data on inflation and unemployment trends, and about implied investor projections for future inflation rates, as measured by security prices. But there is another point that investors need to remember. Monetary policy is not run according to mechanical rules that are like Newtonian physics. Fed officials are in fact unleashing a gigantic experiment in psychology.

Inflation trends will depend not just on “real” economy activity, or the seemingly precise data on market expectations, but the murky mysteries of the minds and emotions of consumers. Will they accept retail price increases? View these as permanent? Translate these into higher demands for wages?

Economists sometimes seem ill-equipped to track sentiment, or how this is shaped by public narratives about it. So much so, that the economist Vincent Reinhart depicts consumer inflation expectations as the “unidentified flying object” of monetary policy: out there, but mysterious.

Thankfully, Fed officials increasingly seem to recognise the problem. Arguably, they are putting more emphasis on consumer psychology in decision-making today than at any point in recent history. Better still, they are trying to get more reliable data to track this.

Until recently, the main way to do this was to monitor consumer surveys, such as those conducted by the University of Michigan and New York Fed. But these are fragmented, use differing approaches and can be contradictory. Although some surveys depicted falling expectations in spring, a new New York Fed study suggests they are now surging.

Last year the Fed created a new data series called the “index of common inflation expectations” that collates these surveys, alongside other inflation signals. This move attracted scant attention, but it matters enormously. Having one easy-to-understand signal makes it easier for the Fed to focus on sentiment in policy discussions.

It is important for another reason. The CIE series shows that inflation expectations have stayed remarkably stable in recent years, near the Fed’s target of 2 per cent, irrespective of wider economic gyrations. That seems like a victory for the Fed and is a key reason why its officials have been relaxed about ultra-loose policy. (Which may have been a mistake for financial stability reasons, but that is another story.)

This matters for the future. Richard Clarida, deputy Fed governor, recently stated that if the CIE stays low, the “pace of policy normalisation . . . would be somewhat slower than otherwise”.

But there are two reasons for caution. First, the CIE has not been tested over time and is only published every three months. This means the latest result does not reflect recent inflation surges. Second, as Shiller observed, emotion cannot always be captured by just one number, least of all at times of flux.

So the Fed should go further and harness a wider range of intellectual tools. It should emulate data geeks in political and consumer industry analysis, and study cyber signals to track sentiment shifts. It should assess how the internet revolution is changing information networks, since digitisation not only affects the velocity of information as much as money, but confidence in this information, too.

More specifically, we live in what the Oxford professor Rachel Botsman describes as an era of “distributed trust”. People rely on their cyber peers for advice, and only partly or not at all on authority figures. Surveys by the Edelman public relations group have shown how this reshapes politics and corporate life. We need to know how it affects monetary policy, too.

Lastly, the Fed should commission qualitative ethnographic research, to study how consumers perceive inflation in their everyday lives. Asking questions about this in surveys is useful, but observation and ethnography could throw light on what consumers do not voluntarily talk about. Silences matter.

No, this is not quite “normal” economics. But “normal” monetary policy rules have already been ripped up. If there was ever a good moment for the Fed, and economics profession, to widen its intellectual lens, it is now. Without this, it will be hard to decode that monetary policy UFO — and right now ignorance is dangerous.