Mean

The average value of a variable, commonly used to summarize data and connect samples to expectations.

The mean is the average value of a set of numbers. In economics, it is used to summarize data (average income, average inflation) and to connect empirical averages to theoretical expectations.

Arithmetic mean (the usual definition)

For observations (x_1,\dots,x_n), the arithmetic mean is:

[ \bar{x} = \frac{1}{n}\sum_{i=1}^{n} x_i. ]

  • (\bar{x}) is a sample mean (computed from data).
  • (\mu) is often used for the population mean (the mean of the underlying distribution).

Weighted mean

When observations carry different importance (population shares, expenditure weights, portfolio weights), economists use a weighted mean:

[ \bar{x}w = \sum{i=1}^{n} w_i x_i \quad \text{with} \quad w_i \ge 0,; \sum_{i=1}^{n} w_i = 1. ]

Examples include inflation measurement (price indices) and average wages computed across groups with different employment shares.

Mean as an expected value

In probability and econometrics, the mean is the expected value of a random variable (X):

[ E[X] = \sum_x x,P(X=x) \quad \text{(discrete)}, \qquad E[X] = \int x f(x),dx \quad \text{(continuous)}. ]

Under standard conditions, the sample mean (\bar{x}) converges to (E[X]) as the sample gets large (law of large numbers). This is why averages are so central to empirical work.

Why the mean can mislead

The mean uses all values, so it is sensitive to outliers and skewed distributions.

Example: incomes (30, 30, 40, 200) (in thousands) have:

  • mean (= (30+30+40+200)/4 = 75)
  • median (= 35)

If the distribution has a long right tail (as income often does), the mean can be much higher than what a “typical” person experiences.

Knowledge Check

### What is the arithmetic mean of \(x_1,\dots,x_n\)? - [x] \(\bar x = \frac{1}{n}\sum_{i=1}^{n} x_i\) - [ ] \(\bar x = \sum_{i=1}^{n} x_i\) - [ ] \(\bar x = \max_i x_i\) - [ ] \(\bar x = \min_i x_i\) > **Explanation:** The arithmetic mean is the sum divided by the number of observations. ### In a weighted mean \(\bar x_w=\sum w_i x_i\), what condition do the weights satisfy? - [x] \(\sum_{i=1}^{n} w_i = 1\) (with \(w_i\ge 0\)) - [ ] \(\sum_{i=1}^{n} w_i = 0\) - [ ] \(w_i\) must all be equal - [ ] \(w_i\) must be negative for outliers > **Explanation:** Normalizing weights to sum to 1 makes the weighted mean a convex average of the observations. ### Why can the mean differ a lot from the median in income data? - [x] Income distributions are often right-skewed, so high outliers pull the mean upward - [ ] The mean ignores extreme values by construction - [ ] The median is always larger than the mean - [ ] The mean is only defined for discrete data > **Explanation:** Because the mean uses all values, a long right tail can move it far above the “typical” (median) observation.