Advice on how to “look upper class” usually focuses on things that can be bought or copied: the right clothes, a good watch, a polished table manner. Research on how people actually judge each other’s social class tells a different story. Strangers do size up class quickly, often within seconds, but many of the cues they rely on are habits built up over a lifetime. Those are much harder to change than a wardrobe, and the purchasable signals turn out to be less reliable than many people assume.

A minute of body language

In a 2009 study published in Psychological Science, psychologists Michael Kraus and Dacher Keltner of the University of California, Berkeley filmed volunteers from a range of backgrounds having a short getting-to-know-you conversation with a stranger. They then showed 60-second clips of these conversations to observers who knew nothing about the people on screen and asked them to estimate each person’s socioeconomic status.

The observers’ guesses were not random. They tended to agree with one another, and their estimates matched the volunteers’ actual family income, their mothers’ education and where the volunteers placed themselves on the social ladder.

The clues were in small nonverbal habits. Volunteers from wealthier backgrounds showed more signs of disengagement, such as doodling or fiddling with personal objects. Those from less wealthy backgrounds showed more signs of engagement, such as nodding, laughing and making eye contact. The researchers linked this to dependence on others. People with more resources may feel less need to win over a stranger, and it shows in small ways they are rarely aware of.

That finding is awkward for anyone trying to project status. The behavior that read as “upper class” in this study was not polish or charm. It was a kind of mild inattention that most people would consider less socially skilled, not more.

Seven words are enough

Speech carries even stronger clues. In a 2019 paper in the Proceedings of the National Academy of Sciences, Kraus, by then at Yale University, and colleagues ran five studies on how class comes through in the way people talk.

In one, 229 listeners heard 27 different speakers each say just seven words, taken from a dialect archive. The listeners judged the speakers’ social class better than chance. Their guesses about class were less accurate than their guesses about a speaker’s age, gender or race, but they were still meaningfully above a coin toss.

Another study compared listening with reading. When participants heard speakers describe themselves, their judgments of class lined up with the speakers’ real backgrounds more closely than when they read word-for-word transcripts of the same descriptions. In other words, the giveaway was mostly in how people spoke, especially pronunciation, rather than what they said. Speech that sounded closer to “standard” English, the kind used by digital assistants such as Alexa, was associated with both higher perceived class and higher actual class.

The final study showed why this matters. People with hiring experience listened to brief recordings of job candidates. Without seeing any résumés, they rated the candidates who sounded higher class as more competent and a better fit, and suggested higher starting salaries and signing bonuses for them.

Accents can shift over time, but the study points to how deeply the cues are embedded. Pronunciation is learned in childhood, long before anyone thinks about how they come across.

Even a resting face

The face may carry clues too. In a 2017 study in the Journal of Personality and Social Psychology, Thora Bjornsdottir and Nicholas Rule of the University of Toronto photographed students with neutral, expressionless faces. The students came from families earning either under $60,000 or over $100,000 a year. A separate group of participants then looked at the photos and guessed who was richer and who was poorer.

They were right about 53 percent of the time. That is only a small edge over guessing, but it was consistent. The effect disappeared when the students were smiling or showing other emotions. The researchers found that the neutral faces of richer students showed slightly more positive affect, and they suggested that habitual expressions gradually leave a faint mark on the face, even by the late teens and early twenties.

As with speech, these impressions had consequences. Faces judged to be richer were also rated as more employable.

What about designer labels?

The obvious shortcut is to buy the signals. Here the evidence is mixed. In 2011, Rob Nelissen and Marijn Meijers, social psychologists at Tilburg University in the Netherlands, reported seven experiments in Evolution and Human Behavior testing the effect of luxury brand logos on clothing.

The results were striking. When a research assistant wore a sweater with a Tommy Hilfiger logo, 52 percent of shoppers agreed to answer her survey, compared with 13 percent when she wore an identical sweater without one. In a video of a job interview, the same man was rated more suitable and was suggested a higher salary when his shirt showed a designer logo.

But one result points to the limits of faking. In an economic game, participants were more generous to a partner wearing a logo, until they were told the shirt had been given to that partner by the experimenters. Once the logo was known to have cost the wearer nothing, it lost its effect. The signal worked only when people believed it was expensive to send.

The broader effect has also proved fragile. In 2017, researcher Joël Berger published a replication in PLOS ONE using five field experiments. He found no positive effect of luxury labels in any of them, and in poorer neighborhoods there was even a tendency for the luxury label to work against the wearer. How a status symbol is received appears to depend heavily on who is looking.

Why the deeper cues are hard to copy

Taken together, the research suggests a hierarchy of class signals. The easiest to change, such as clothing labels, have the weakest and least consistent effects. The most telling, such as pronunciation, conversational style and the resting expression of a face, are built up over years and are mostly invisible to the person sending them. In the Toronto study, participants could not even say what they were picking up on.

The findings come with limits. The effects are averages across groups, and in the face study the accuracy was only modestly better than chance. Several of the studies used university students or relatively small samples, and much of it was done in the United States, where class and speech interact in particular ways. Kraus’s speech studies relied partly on education as a stand-in for class.

The researchers behind these studies tend to draw a different lesson from the one implied by “how to look upper class.” Their concern is less about how individuals can game these judgments and more about what the judgments do. If employers and strangers read class from a few words or a neutral face, and then treat people accordingly, those snap impressions can help keep people in the positions they started from. Knowing that these judgments happen, and how little they reveal about a person’s actual ability, may be more useful than any attempt to fake the signals.

Representative image. Photo: Ketut Subiyanto / Pexels.