The Average Person Does Not Exist

You have almost certainly been told, at some point this month, what the average person does.

The average person spends so many hours a day online. The average person checks their phone so many times. The average person sleeps less than they should, earns so much, owes so much, feels lonely this often. The figures come from careful surveys and are, for the most part, honestly reported. Some of them have appeared in pieces on this very shelf.

And then, very often, a small and quiet thing happens. You compare yourself to the average person, and you feel that you have failed. You are above the average where it is bad, below it where it is good. You are, by the measure of the figures, a slightly deficient copy of someone.

I would like to examine that someone. I would like to ask the average person the question I ask of every claim.

Who are you?

The question

Here it is in one line.

Does the average person exist, and if not, what is the average actually telling us about any one of us?

Let me proceed by steps.

Step one: what an average is

An average is a property of a group. You take many measurements, from many people, add them up and divide by how many there were. The result is a single number that describes the centre of the group.

It is a useful number. If you are designing a bus timetable, a hospital, a school, you need to know roughly how many people will use it and roughly how much of what they will need. Averages make that possible. They are one of the great tools of the modern world, and I have no quarrel with them.

But notice the first step. An average describes a group. It is not, in itself, a description of any member of the group. The average number of children per household can be a fraction, and no household has a fraction of a child. The fraction is a fact about the group. No individual has it.

So far, nothing controversial. The difficulty begins when we forget this.

Step two: the cockpit

There is a well-documented case that shows, better than any argument, what happens when the average is mistaken for a person.

In the early 1950s, the United States Air Force was measuring its flyers in great detail, for the design of cockpits and equipment that would have to fit them. A young researcher named Gilbert Daniels, the project scientist on that body-measurement work, asked a simple question of the numbers: how many of these men are actually average?

Daniels examined measurements from 4,063 Air Force flyers. He defined "approximately average" generously, as falling within the middle 30 percent of the range on a given measurement. He then asked how many men were approximately average on ten measurements relevant to design. The answer was none: "there was not a single individual remaining who fell within the 'average' range for all measurements." Even on just three measurements, stature, chest circumference and sleeve length, only 143 of the 4,063, about 3.5 percent, were approximately average on all three. (Gilbert S. Daniels, "The 'Average Man'?", Technical Note WCRD 53-7, Wright Air Development Center, December 1952.)

Read that slowly. Out of four thousand men, chosen from a single, relatively uniform profession, measured on ten dimensions, the number who matched the average on all ten was zero. A cockpit built for the average would have been built for a man who did not exist.

This is the second step, and it is not a matter of opinion. The more dimensions you measure, the less likely it is that any real person is average on all of them. On one dimension, plenty of people are near the middle. On ten, nobody is.

Step three: why this is a fact about mathematics, not pilots

It would be a mistake to treat the pilots as a curiosity. The same thing would happen to almost any group, on almost any set of measurements, and it is worth seeing why.

Suppose you are near the middle on one measure. Your chance of also being near the middle on a second, independent measure is only a fraction. Being near the middle on a third is a fraction of that fraction. Each additional dimension multiplies the odds downward. Measurements of a real person are not perfectly independent, which is why three dimensions still left a few men in the pilots' case, but they are independent enough that the product shrinks very fast.

Now apply this to the figures you are told about the average person: hours online, hours of sleep, hours of leisure, income, debt, number of friends, frequency of loneliness. Each one is a fair description of a group. But the person assembled from all of them, the one who sleeps the average amount, earns the average income, spends the average time online and feels lonely the average number of times, is exactly as fictional as the average pilot.

Settled at this step: a person who is average across many dimensions is, for practical purposes, nobody. The figures are real. Their combination into a single person is not.

Step four: what we do with the fiction

If the average person is fictional, why does it matter?

Because we use the fiction as a mirror.

Consider what happens when you read that the average person spends so many hours a day online. You do not usually think that is a fact about a population. You think I spend more than that, or less, and you feel something about it. Guilt, usually. Occasionally a small pride. In either case, you have taken a property of a group and used it as a standard for yourself, as though the average were a target someone set and you had hit or missed.

And the average, by step one, is not a target. Nobody set it. It is simply where the middle of the crowd happens to fall. Being above it or below it says that you are not in the middle. It says nothing about whether you are doing well or badly, because on any real dimension, roughly half of everyone is on each side of the middle, by construction.

The cockpit was a physical version of this error. It was designed for the middle, and so it fitted no one. The modern version is psychological. We hold ourselves against a composite of middles, and so we all feel, in some respect, that we do not fit.

Step five: the worth that has no average

This is the step at which the House's premise becomes relevant.

The books on this shelf rest on one claim: that a person's worth begins with their first breath, and that nobody can add to it or take it away. I have argued for that claim at length elsewhere, and I will not repeat the argument here. I want only to show how it bears on the average.

If worth were a measurement, it would have an average, like everything else that can be measured. There would be people above it and people below it, and the comparison would carry the same sting as all the others.

But if the premise is right, worth is not a measurement. It does not vary from person to person. Each person's is complete, and complete in the same way. There is no distribution, no middle, no spread. There is no average worth, because there is nothing to average.

Which means that on the one dimension that actually matters to how a person should regard themselves, the question am I above or below average? does not arise. It is not that everyone is average. It is that the concept does not apply. Asking for the average worth is like asking for the average of a single number repeated: there is only the number.

Step six: the strongest objection

The strongest objection, as I understand it, runs like this.

This is all very well, but averages are how we notice that something is wrong. If the average person is lonely, or tired, or in debt, that tells us about a problem that affects real people. Dismissing the average as a fiction risks dismissing the problem. And for an individual, knowing that you're far from the average, sleeping three hours a night, say, can be a genuine warning that something needs attention.

That objection is correct on both counts, and I want to grant it fully.

First, averages are indispensable for seeing problems in a population. If about half of adults report loneliness, as the US Surgeon General's 2023 advisory said, that is a fact about a society, and it should move us to act on the society. Nothing in my argument weakens that. A fictional average person can still be the sign of a very real shared condition.

Second, for an individual, distance from the middle can be useful information, especially when it concerns health. If you sleep far less than nearly everyone, that is worth noticing and perhaps worth discussing with a doctor.

But notice the form of both uses. The first uses the average to understand a group. The second uses the distribution, how far you are from where most people fall, as one piece of information about a particular dimension of your life, to be weighed with others. Neither use requires you to imagine a single average person and measure yourself against them as a whole. Neither use says anything about your worth.

So the answer to the objection is: keep the averages, and use them for what they are good for. Do not use them as a mirror.

What is now settled

Let me set out the conclusions plainly.

Settled. An average is a property of a group, not of any member of it.

Settled, by a well-documented case and by elementary reasoning. The more dimensions are measured, the fewer real people are average on all of them. Of 4,063 Air Force flyers measured in Daniels's study, none was average on ten dimensions.

Settled. The average person assembled from many separate figures is a fiction. Comparing yourself to that fiction as a whole is a comparison with nobody.

Settled. Averages remain indispensable for understanding populations, and distributions can be useful information about single dimensions of a single life.

Settled, if you accept the House's premise. Worth is not a measurement and has no average. The question of being above or below it does not arise.

Not settled. How to live in a world that constantly offers us the average as a mirror. I have no proof to offer there, only a practice.

A practice, not an exhortation

I do not end with instructions, but I will leave you with something that can be tried and observed.

The next time you meet a figure about the average person, and you will, probably today, perform one small translation before you let it touch you. Turn the sentence about a person into a sentence about a crowd.

The average person spends so many hours online becomes across the crowd, the middle falls at so many hours.

The average person feels lonely becomes a great many people in the crowd feel lonely.

Then ask a second question: Is this telling me about the crowd I belong to, or is it telling me who I am?

Almost always, it is the first. Which may be important. Which may even be a reason to act, for yourself or for others. But it is not a verdict on you, because no figure about a crowd can be.

The average pilot never sat in the cockpit. The average person will never sit in your chair. There is only you, measured on more dimensions than any survey will ever count, matching the average on none of them.

That is not a deficiency. Like the four thousand and sixty-three pilots, it is simply what it means to be a particular person.

I am content to call that proven.