Imagine two neighbors filling out the same survey about where consumer prices are headed over the next year. They read the same headlines, shop at the same stores, and pay similar mortgages. Yet one predicts mild inflation while the other braces for a painful spike. What explains the gap? A team of economists set out to test an old idea: that our beliefs about the broader economy are shaped not only by the facts in front of us, but by which personal memories happen to be top of mind.
That investigation appears in an NBER working paper titled “The Psychology of Macroeconomic Expectations.” The authors report that nudging people to recall a personal hardship, one that carries no real information about prices, pushed their inflation and home-price forecasts higher. The finding offers evidence that emotions and context can move economic beliefs even in the absence of news.
An old puzzle about confidence
Nearly a century ago, the economist John Maynard Keynes argued that economic decisions ride on “animal spirits,” a kind of confidence or mood that standard number-crunching cannot fully capture. Economists have long documented that sentiment predicts booms and busts, but they have struggled to pin down how a feeling, absent any new data, could change what people actually expect to happen.
The research team behind this paper includes Pedro Bordalo of the University of Oxford, Nicola Gennaioli of Bocconi University, Andrei Shleifer of Harvard University, and colleagues, along with Maarten van Rooij of the Dutch Central Bank. Their proposed answer draws on the psychology of memory. When a person is asked to forecast inflation, the authors argue, the mind does not run a tidy statistical calculation. Instead, it retrieves relevant past experiences and uses them to imagine, or “simulate,” what the future might look like.
Two kinds of memories matter here. The first the authors call domain-specific: experiences directly tied to the forecast, like the prices you remember from years ago. The second they call non-domain-specific: experiences from unrelated parts of life, like a job loss or a health scare. The team’s claim is that a personal financial struggle, though it tells you nothing about national price trends, can still color an inflation forecast. Because such a memory feels bad, and high inflation also feels bad to many people, the mind treats them as similar and uses one to imagine the other.
An experiment built on memory, not information
To test this, the researchers fielded a survey experiment on more than 4,000 households participating in the Dutch National Bank Household Survey, a long-running panel run through Tilburg University. The survey was conducted in June 2024, with a follow-up wave that October.
The design was straightforward. Before asking people for their expectations about inflation and home prices, the team randomly sorted respondents into three groups. One group was prompted to recall a time when they or a loved one struggled with financial hardship. A second group was asked to recall a serious illness or accident. A third group, the control, received no prompt at all. Those who were prompted described the event, the difficulties involved, and the emotions they remembered.
The key feature of this setup is that the prompt carried no information. People were asked to recall events they had already lived through, and a personal illness or money trouble says nothing about where national prices are heading. So in any model where beliefs depend only on facts or on stable personal preferences, the prompt should have no effect. If it did move beliefs, the authors reasoned, the cause had to be a shift in mental context: a change in which memories were active when people answered.
The survey also measured each person’s other life experiences, asked them to rate how similar each experience felt to high inflation and high home-price growth on a one-to-seven scale, and collected written explanations of the reasoning behind their forecasts.
What the prompts did to people’s forecasts
Recalling a hardship made people more pessimistic. Among respondents who had actually lived through the primed experience, being prompted to recall a financial struggle raised inflation expectations by about 0.97 percentage points against an average forecast of 5.46 percent, and lifted home-price forecasts by roughly 1.08 percentage points against a 6.51 percent average. The health prompt produced similar, slightly smaller shifts.
The financial prompt moved beliefs more strongly than the health prompt. This lines up with the survey’s similarity ratings: respondents rated financial difficulty as much more similar to high inflation and rising home prices (around 4.5 on the seven-point scale) than illness or accident (closer to 2.7). The team interprets this gradient as support for their central mechanism. The closer an adverse memory feels to a bad economic outcome, the more it pulls forecasts in a pessimistic direction.
The effect also showed up in people’s existing experiences, not just the prompted ones. In the control group, those who reported having faced financial hardship gave higher inflation and home-price forecasts than those who had not, even after accounting for age, income, education, and other characteristics. Because individual hardships do not actually predict national price trends, the authors read this pattern as evidence of a psychological driver rather than an informational one.
Changing the story people tell themselves
The prompts did more than nudge a number. They changed how people reasoned. Using a language model to classify the written explanations, the researchers found that recalling a hardship made respondents more likely to ground their forecasts in personal experience and negative emotion, with words about affordability and making ends meet, rather than in news reports or central bank policy.
The prompts also raised what the team calls “backcasts,” people’s estimates of how high inflation and home prices had been in the recent past. Prompting a financial memory increased recalled past inflation by about 1.8 percentage points. The authors describe this as mood-congruent recall: a bad memory cues the retrieval of other bad memories, including memories of high past prices, which then feed a gloomier forecast.
From individual minds to the broader economy
The authors suggest this memory mechanism could help explain several long-standing observations: why households tend to overestimate inflation, why people who see identical news still disagree, and why beliefs can swing sharply in response to vivid stories rather than dry statistics. They connect it to the work of the economist Robert Shiller on “narratives,” arguing that a viral story works by changing which experiences come to mind.
They also sketch a broader idea they call a “confidence multiplier.” If one person’s spending acts as a cue that signals good times to others, it could prompt those others to feel optimistic and spend more, which cues optimism in still more people. This feedback, the authors argue, could in principle generate self-sustaining booms or slumps driven by shifting confidence rather than by hard economic data.
A few caveats are worth keeping in mind. The experiment was conducted with Dutch households, and the patterns may differ elsewhere. The strongest belief shifts appeared among people who had genuinely lived through the primed hardship, and the priming effect faded by the second survey wave, consistent with the idea that whatever is top of mind at the moment of forecasting is what counts. The larger claims about business cycles and the confidence multiplier are theoretical extensions the authors propose rather than results they directly tested. What the experiment does show is narrower and concrete: an uninformative reminder of a personal hardship can reshape how people imagine the economic future.
