Thursday, July 15, 2010

Income Distribution

I was curious about the individual income distribution in the US for people over 15 in 2008, and so I looked up some data from the US census bureau for individual income. It doesn't give additional breakdowns for people making over $100k so I did some extrapolations. This is mainly a thought exercise for what you think the income distribution is compared to what you expected it to be compared to what you might think is fair.

In these graphs, I extrapolated the distribution for people making over $100k. In this first graph, the distribution is if the total amounts of money made by the number of people making a certain income in $2500 increments were all equal. That is, 100k*ppl_making_100k = 102.5k*ppl_making_102.5k, and the shape of the distribution is 1/x. As you can see, the maximum amount of income people are making is around $460k.

That is not realistic because I know for a fact that there are many people making millions, but I don't know what is the maximum million of dollars people make in a year. This is the graph that is the same as above but with a few people added at $1 million and $11 million.

In contrast, this is usually the kind of graphic that is used when portraying income distribution. It is for the bottom 98% of households from


Anonymous said...


mirthbottle said...

yeah but what is the physical meaning of a log-log scale? I mean, if you make income log, it implies that the distance between 10 and 100 and 100 and 1k are the same, and I don't see how you can say that those are equivalent differences between people's income. it's not to say it's not worth plotting, but I think this is an important shape to look at for the first pass when looking at income.