Published on 2026-07-25
Why doesn't 'the average' tell you what's typical?
Add one very wealthy person to a room and the average savings jumps into the millions, though nobody got richer. That's the trick behind most 'average' claims.
Ten friends sit around a table comparing savings. Nine of them have a few thousand dollars put aside. Then an extremely wealthy guest joins the group, someone worth tens of millions, and suddenly the table's 'average savings' rockets into seven figures. Nobody at that table got richer in the last five minutes. The number describing them did, and that gap between the number and the people it claims to describe is exactly what makes 'the average' one of the most quietly misleading phrases in everyday reporting.
This isn't a trick that only happens around imaginary tables. It happens at national scale, every few years, in one of the most careful surveys of household finances that exists.
The gap, at national scale
The US Federal Reserve runs the Survey of Consumer Finances, a detailed count of what American households actually own and owe. Its 2022 edition, published in October 2023, put the mean net worth of US families at 1,063,700 dollars. The median net worth, for the same year, the same families, the same survey, was 192,900 dollars. One number is more than five times the other, describing exactly the same population at exactly the same moment.
Both figures are correct. Both come from the same careful research. They simply answer different questions, and most headlines only ever quote one of them.
One number bends, the other doesn't
A mean, what most people mean by 'average', is the sum of every value divided by how many values there are. That arithmetic makes it exquisitely sensitive to extremes: a handful of very large numbers in the pile can drag the whole result upward, even if almost everyone else sits far below it. Michigan State University Extension, in a plain explainer on the two measures, puts it simply: outliers distort the mean, while the median stays put, because the median is just the value sitting in the exact middle once everyone is lined up in order. Half the group is above it, half below, and one very unusual neighbour joining the line doesn't move that midpoint by much at all.
That's why the extension's advice is so blunt: whenever a distribution is skewed, lopsided by a few extreme values rather than evenly spread, the median is the number that describes the typical case, and the mean is the number that describes something closer to the total divided evenly, which is a different and often less useful question.
The mean tells you what everyone would have if the pile got split evenly. The median tells you what the person in the middle actually has.
Where this shows up beyond bank accounts
Household wealth is the cleanest example because a small number of very large fortunes reliably drags the mean far above where most families actually sit, but the same mechanic hides inside plenty of other 'average' claims. An 'average' home price in a city can be pulled upward by a cluster of luxury sales that have nothing to do with what a typical buyer faces. An 'average' commute, an 'average' hospital stay, an 'average' return on an investment: any of these can be quietly reshaped by a small number of extreme cases sitting at one end of the scale, while the number itself gives no hint that they're there.
The direction of the pull matters too. Wealth and income tend to stretch upward, a long thin tail of very large fortunes above a thick cluster of ordinary ones, which is why the mean for those figures usually sits above the median. Other measures stretch the other way. A handful of very short response times, or a run of unusually cheap sales, can pull a mean downward instead, below where most of the real cases actually sit. The lesson isn't that the mean always exaggerates upward. It's that a small number of extreme cases, wherever they sit, can move the average far more than they move the reality most people are living.
None of this means the mean is a bad measure. It has real, specific uses, especially when a total genuinely needs to be shared or estimated. The trouble starts when 'average' is offered as a stand in for 'typical', because those are two different words doing two different jobs, and only one of them describes the person standing in the middle of the crowd.
The two questions worth asking
Next time a figure is introduced as 'the average', it's worth asking a second, quieter question before accepting the first: what's the median, and how spread out is the data around it? If the two numbers sit close together, the average was a fair summary all along. If they're far apart, as they are for household wealth, something in the tail of the distribution is doing most of the talking, and it usually isn't the household that looks like yours.