Mean vs. Median

The average US household income is around $105,000. More than half of households earn less than $75,000. Both numbers are technically correct. One of them is being used to mislead you.

This is not a rounding error or a political argument. It's a mathematical fact about how averages behave — and once you see it, you'll notice it everywhere.

Two ways to find the middle

You probably learned both of these in school, but the difference matters more than it seemed at the time.

The mean is the arithmetic average: add everything up, divide by the count. The median is the middle value: sort all the values in order, find the one in the center.

For a symmetric dataset — one that looks the same on both sides — the mean and median are nearly identical. But most interesting real-world data isn't symmetric.

What happens when there's a billionaire in the room

Imagine ten people sitting in a bar. Their incomes range from $30,000 to $80,000. The mean income is around $50,000. The median is also around $50,000. Both feel like fair summaries.

Now Jeff Bezos walks in.

His income — let's say $10 billion — gets added to the sum. Divided across eleven people, the mean income in that bar is now over $900 million. The median? Still around $50,000.

Which number better represents what a typical person in the bar earns? The median, obviously.

The mean got pulled toward the outlier. That's always what happens: the mean is sensitive to extreme values, and the median is not. Neither is wrong — they're answering different questions.

Mean vs. median puller

63.9

Mean

61.0

Median

MeanMedian

Drag the right-side dot far out. Watch which line chases it.

Reset

Skew is the key concept here

When the mean is significantly higher than the median, the distribution is right-skewed: there's a long tail on the right side (the high end). When the mean is lower than the median, it's left-skewed.

Right-skewed distributions are common wherever there's a hard floor but no ceiling:

• Income (can't earn less than zero, but no upper limit)
• House prices
• Social media followers
• Company lifetimes

In all of these, the mean is inflated by the extreme high end. The median tells you what a randomly selected person or item actually looks like.

When the mean is the right choice

The median isn't always better. The mean is the right tool when:

• You need a total. If you want to know how much everyone in a country earns combined, you need the mean times the count — the median doesn't help.
• The distribution is roughly symmetric and there are no extreme outliers.
• You're averaging things that add linearly, like distances or time.

The key question is: are you asking about the typical individual, or the aggregate? Median for the individual. Mean for the aggregate.

The formal definitions

For a dataset of $n$ values $x_1, x_2, \ldots, x_n$:

$\bar{x} = \frac{1}{n} \sum_{i=1}^{n} x_i$

The median is the value at position $\frac{n+1}{2}$ after sorting (for odd $n$), or the average of the two middle values for even $n$. There is no tidy closed-form expression — it's defined by rank, not arithmetic.

Key takeaways

• The mean is pulled toward outliers; the median is resistant to them
• For right-skewed distributions (income, prices), the median is almost always the more honest summary
• When you see "average" in a headline, ask: would the median tell a different story?
• Mean for totals and aggregates. Median for typical individuals.