Statistics Definitions > Markov’s Inequality

**Markov’s Inequality** gives an upper bound on the probability that a non negative random variable is greater than or equal to some positive constant.

Given a non-negative random variable and a positive constant , Markov’s Inequality states that the probability that is greater than or equal to must be less than the Expected Value of divided by .

Markov’s Inequality is often expressed in a slightly different form, with the constant being some multiple of the Expected Value of .

Substituting we get:

With this more familiar form we can look at an interesting and useful application of Markov’s Inequality.

## Application of Markov’s Inequality

Suppose we wanted to know, *what percentage of people make more than 4 times the average income?*

Here our non-negative random variable is **income **(all income is positive) and the Expected Value of is the same as the *average* income , assuming the income distribution is symmetric about the average income.

Substituting and we get:

**N****o more than 1/4 of the population (25%) can make more than 4 times the average income.**

Intuitively it makes sense that there must be an upper bound for the percentage of people making more than 4 times the average income. If this number were too high (say 80%) the average income itself would be much higher than what it is.

Markov’s Inequality allows us to calculate this upper bound exactly, with the surprisingly simple result that the upper bound is just the * reciprocal *of the

*of the*

**multiple****:**

*average*- no more than
**1/3**of the population can make more than**3 times**the**average** - no more than
**1/5**of the population can make more than**5 times**the**average** - no more than
**1/10**of the population can make more than**10 times**the**average** *etc…*