How to compare average and median salary information

How to compare average and median salary information

I recently described how to calculate a percentile. Today I will clarify the difference between the average and the median.

Technically, the average is the sum of all data points, divided by the number of data points. The median is the 50th percentile, or the value that splits the data points in two equal groups (each group with the same number of data points).

When you look at your internal equity or study compensation survey results, both are very useful and should be used in cunjunction in order to understand your population.

Let me give you an easy example of an internal equity and distribution analysis.

  • A population is made of one CEO earning 100 million per year, 2 managers earning 100, 000 each, and 70 blue collars earning 10,000 each.
  • The average is (100,000,000+2*100,000+70*10,000)/73 = 1,382,192.
  • For the median, first calculate 73 * 50% = 36.5. Once you have sorted all your population by increasing salary, you look for the average value between the 36th and 37th lowest or highest salaries. Given that 70 out of 73 employees earn 10,000, employees number 36 and 37 both earn 10,000 and the median is therefore 10,000.

None of these results is wrong. The average salary in the company is 1.4 million while the median salary is 10,000. Checking both numbers helps you understand what is going on in your organisation. In that case, you may consider that the median information is more relevant to describe your organisation than the average salary.

Similarly when you receive compensation survey results, always check out both the median and the average results. If they are relativeley close to one another, you can safely assume that most of the employee data submitted to the survey is pretty clustered around the median, and if there are outliers, they are as “strong” in the high salaries as in the low salaries and therefore they “cancel” each other out.

If median and average are quite apart however, you can assume that something is skewing the results, and you need to investigate (maybe even speak to the survey provider to gain a better understanding) : is there a small group or very highly or very poorly paid employees that affect the average data ? Could it be that these few employees may have been matched to the incorrect job, one too high (hence their relatively lower salary as they should have been matched one grade lower) or one too low (hence their relatively high salary as they should have been matched one grade higher) ?

Or is there really a big distorsion and disparity of pay in the market ? In that case you will also need anecdotal data such as salaries of recent hires in the same position, or feedback from headhunters. This will give you a better idea of what to say to managers regarding the level of pay in that particular position, especially if you are planning to recruit or are facing retention issues.

In short, whether checking internal equity or analysing market data, always compare average and median, and if they are quite different, try to understand the reason behind so that you can give relevant feedback and provide meaningful data to your HR partners and the line managers.


Liked this post ? Why don’t you subscribe by email to receive my posts 3 times per week ? You will never miss the Sunday Skills post every two weeks and will continue to learn practical technical tips for an effective use of your time in Compensation & Benefits !

Related post :

Percentiles – an easy way to “figure it out”

Print Friendly, PDF & Email

Speak Your Mind