When the average income of a person's reference group doubles, their spending on visible goods falls by roughly half at the same personal income, which makes conspicuous consumption a signal about relative position rather than a reliable tell for wealth

The premise of every list of this kind is that certain habits are reliable indicators, and that a trained observer can read wealth off behaviour the way a mechanic reads an engine by ear. Two well-established pieces of research make that harder to sustain, and they fail it in different directions.

One says that visible spending is not a measure of how much money a person has. The other says that the same behaviour can read as high status or low status depending on who is watching.

Visible spending tracks the neighbours, not the bank balance

Kerwin Kofi Charles, Erik Hurst and Nikolai Roussanov published Conspicuous Consumption and Race in the Quarterly Journal of Economics in 2009. The working paper version sets out the model and the data in full. They used Consumer Expenditure Survey records from 1986 to 2002, covering 48,758 households.

Their category of visible goods is the things other people see: clothing, jewellery, cars. The raw pattern in the data is that Black and Hispanic households spent roughly 32 and 31 per cent more on these goods than White households at comparable income, around 2,300 dollars a year in absolute terms.

That number gets quoted on its own fairly often. The paper’s actual finding is what happens to it next.

The authors add one variable: the mean income of the person’s reference group, defined as others of the same race in the same state. The coefficient for Black households falls from 0.32 to -0.09 and stops being statistically significant. The Hispanic coefficient falls to near zero. The gap does not survive the control.

What is doing the work is the relationship between visible spending and reference-group income, and it is strong. Holding a household’s own income constant, doubling the mean income of its reference group is associated with visible expenditure falling by something in the range of 48 to 56 per cent.

Read that slowly, because it inverts the premise of the topic. Two people with identical incomes will spend very differently on the things other people can see, and the variable that predicts the difference is how rich the people around them are. Signalling effort goes up as the surrounding average goes down.

Under this model the habits that broadcast wealth most loudly are, on average, coming from people lower in their local distribution, not higher.

The same clothes read two opposite ways

The second problem is at the receiving end. Silvia Bellezza, Francesca Gino and Anat Keinan published The Red Sneakers Effect in the Journal of Consumer Research in 2014, across five studies.

The cleanest is the first. A woman entered luxury boutiques in Milan either in elegant dress or in gym clothes. Shop assistants rated the version in gym clothes as higher status, 4.9 against 3.8. Pedestrians at a train station, shown the same thing, rated her the opposite way, 3.5 against 5.7.

Same woman, same clothes, reversed reading.

The mechanism the authors identify is that breaking a norm signals status only to an observer who knows the norm well enough to see that it is being broken deliberately. Shop assistants in a Milan boutique know exactly what someone is expected to wear. Passers-by do not, so they read the same outfit as simply being unable to afford better.

The moderators are as informative as the effect. When the deviation looked accidental rather than intentional, the status advantage disappeared entirely. In the professor study, T-shirt and beard read as more competent than suit and tie, but only at prestigious universities, where the norm was clear enough to be violated on purpose. Observers who scored low on need for uniqueness showed no effect at all.

So a habit does not carry status information by itself. It carries status information to a particular audience that knows the local rule.

Why the two findings fit together

Both are saying the signal is relational rather than intrinsic, from opposite ends of the transaction.

On the sending side, how much visible display someone does depends on the gap between them and their reference group. On the receiving side, what the display means depends on how fluent the observer is in the relevant norm. Neither leaves much room for a list of habits that reliably indicate money.

It also explains why the two most common claims in this genre contradict each other and both feel true. The claim that the wealthy signal loudly and the claim that real wealth is quiet are describing different reference groups and different audiences. Someone rich relative to their surroundings, being observed by people who share the local code, is in a different situation from someone rich in absolute terms being observed by strangers.

What neither study supports

The Charles, Hurst and Roussanov paper is an economic model fitted to survey data, and the authors are careful about what it does not establish. Their reference group is defined at state level by race, which assumes people compare themselves to a group they may not be neatly sorted into. They discuss alternative explanations including differences in preferences and housing-market discrimination, and they do not claim their model excludes those.

Expenditure survey data also depends on households reporting their own spending, which is a soft measurement for anything sensitive.

The Red Sneakers studies are the more standard kind of limitation: several were conducted with student samples, the status ratings are perceptions rather than outcomes, and the effect sizes are moderate. Nobody has shown that wearing trainers to a boutique improves how you are actually treated over time, only how you are rated in the moment.

Neither paper says that no habit correlates with money. Plenty do. What they show is that the correlation runs through position and audience rather than through the habit, which is why the reading changes when either of those changes.

The practical residue is small and slightly deflating. If you are trying to work out what someone has from what they do, the more informative question is not which habits you are observing but where they sit relative to the people they measure themselves against, and whether you happen to know the rule they are following.