Salaries: Not Normally Distributed

An example of something that does not represent a Gaussian distribution is salary. If it were salary, we would expect to see people making the median salary, plus or minus a thousand. Some people may make $20,000 annually, while others may make $1,000,000 annually. However, we know that very few people make upwards of $500,000 much less $1,000,000 annually. We also know that few people make $10,000 to $20,000 annually and almost no one makes less than $10,000 annually (I’m referring to the average household). Thus, the true distribution of salaries is not symmetric around the mean. There are few people living in severe poverty, but there are even fewer living with millions. Thus, I believe that similar to the Bachelor’s degree example, a better representation of this data would be a gamma distribution. I believe this because in the case of salaries, it does have a mean between two possible extremes (poverty and wealth). However, this mean would be closer to the “minimum” (minimum not being clearly defined) salary than to the “maximum” (maximum not being clearly defined, as well) salary.