Let’s build a person from statistics.
Imagine a 43-year-old person living in France. They are employed and, if we make them a full-time private-sector employee, earn around €2,730 a month. They have 1.56 children. They can expect to live to around 83. The poverty rate is 15.4%.
Who is this person? They don’t exist.
Not because the numbers are wrong. Because they describe different populations, different distributions and sometimes even hypothetical people.
And yet, put them together and we get something that looks remarkably like a human being.
The average person is a statistical construction: a person assembled from the numbers of millions of other people, none of whom necessarily resembles them.
This is not a flaw.
We need averages. They allow us to see populations too large and complicated to understand one person at a time.
So we compress. And compression is useful because it makes the world legible.
But compression also means loss.
What disappears when we average?
At first, you see a single blue square. It looks like one thing.
But zoom in.
It becomes 16 smaller squares, each a slightly different colour. Together, they produce the blue you saw before.
Zoom in again. The 16 become 256. Then 4,096. Then 65,536.
The further you zoom in, the more variation you discover.
Eventually, there are individual squares. None of them is the blue you saw at the beginning. And yet, together, they produce it.
This is what an average can do. A population of millions of very different people can produce a number—or a colour—that belongs to nobody.
We can describe that population with a mean, a median, a standard deviation, percentiles. We can measure inequality and variation. We can build increasingly sophisticated models of the distribution.
And still, something remains outside the numbers.
Why is this person poor? Why did that child leave school? Why did this family move? Why did this business fail?
A statistic can tell us how many people are in a situation. It cannot, by itself, tell us what it means to be one of them.
The point is not that statistics are wrong. The point is that a population and the people who make it up are two different things.
The average lets us step back and see the whole. But when we step back far enough, the individuals disappear.
The average is not the person. It is what we see when millions of people become one colour.