Do inflation expectations matter in a social-media echo chamber?

The writer is chief economist at UBS Global Wealth Management

The Iran war has once again thrust inflation to the forefront of popular consciousness. Earlier this year, searches on Google Trends for “inflation” were almost twice the level of 2022. Most economies’ current inflation pressures are due to energy prices passing down supply chains plus the lingering hangover of tariffs in the US.

By tradition, a central banker is supposed to adopt the detachment of the character Vicomte de Valmont in the film Dangerous Liaisons and coldly dismiss such things as “beyond my control”.

But the current risk of rising inflation expectations is creating some concern among policymakers. For some of them, expectations are perversely idolised as a policy objective in themselves. This reverence overlooks two points: expectation creation is very different from the past; and expectations alone are just meaningless noise.

Consumer and corporate inflation expectations are measured by survey evidence. Surveys today are not answered in the way that they used to be (if they are answered at all). Social media sensationalises everything it touches. In the Michigan consumer survey of May 2025, 29 per cent of respondents believed that US inflation would be above 15 per cent within a year. Inflation at the time was 2.4 per cent. The only other time 15 per cent plus inflation expectations were close to that reading was 1980, when inflation was at 14.8 per cent.

When a consumer’s phone is an echo chamber, and rising mocha prices produce a stream of irate influencers complaining about affordability, it is easy to abandon economic reality. Add ever more extreme political partisanship, and surveyed expectations become pure theatre. The corporate underling filling in a sentiment survey form may also pay more attention to their phone than their corporate strategy. Policymakers who refuse to recognise that social media has changed things risk overstating the importance of today’s higher reported inflation expectations.

If a genuine signal lurks in the warped data, do expectations then matter? They only do if they change economic behaviour. If people say one thing and do another, expectations are just unnecessary noise. For two groups in the economy, that is exactly the situation today.

Consumers’ inflation expectations matter if they affect wage bargaining or alter consumption patterns. No major economy comes close to suggesting the existence of a wage-price spiral. Wage growth is benign. Discounted by inflation expectations, real US wage growth is actually falling. Higher inflation expectations might shift demand earlier, say by encouraging people to buy now to avoid paying more later. There is some evidence of this taking place in the US in early 2025, in anticipation of the tariffs. However, current consumer trends are not suggesting frontloading demand. Consumers are complaining, and then just carrying on as before.

Corporate inflation expectations tend to be about producer prices. Inflation expectations are relevant if companies have pricing power. Because companies tend to raise prices at set intervals (for instance, on a six-month cycle), they may try to add inflation expectations to current pricing. That would lead to “inflation-beating” price increases — a second-round effect central banks should worry about. There is no evidence of this in the major economies today. Costs are being passed on (with an almost indecent alacrity). Beyond that, pricing power seems to be limited.

Recommended

Consumers and companies can expect whatever they want about inflation — for now they lack the ability or the will to act.

The group that does have the power to turn expectations into action is bond investors. If their expectations rise, the nominal cost of borrowing will rise. What is more economically troubling is that the real cost of borrowing may also rise.

In a darkly dystopian world where inflation is unpredictable and economists are not believed, investors may demand not only compensation for inflation but an insurance premium for inflation uncertainty risk. That adds directly to the real cost of borrowing for both the government and private sector, affecting fiscal policy and investment (and through that, economic growth).

The lesson is that social media is spoiling survey-based inflation expectations — and for now policymakers can set aside consumers’ and companies’ expectations as neither group is acting on them. But bond investors still have the ability and willingness to act. For now, bond vigilantes are the key group whose inflation expectations should carry weight in policy setting.