How to Lie with One Number

On the Trump Tax Cut

Yesterday I read a meme on Facebook (so it must be true), that stated "253,020 Michigan Families Paid More in Taxes Last Year Due to Trump's Tax Scam".  The organization that put out that meme made the mistake of stating referencing the source:  the Institute on Taxation and Economic Policy.  Thanks to the help of a friend I was able to track down the figures and to say the least the meme was misleading.

It is true that 253,020 taxpayers paid more in 2018 that in the year before.  It is also true that 4.026 million taxpayers in Michigan paid less; an additional 536,500 paid the same amount.  Also, despite arguments to the contrary it appears that the Trump tax cut did not adversely impact working class families.  People in the 80 to 95 percent income bracket - roughly $111,000 to $218,000 -  were most likely to see a tax increase.  Thirty percent of the top one percent - incomes above about $517,000 -  received a tax cut, but 36 percent saw a tax increase.  In the state of Michigan, the re-distributive aspects of the tax cut appear to be overstated.

How to Interpret Data Correctly

The first question to ask yourself when you are given one number is to ask "compared to what".  In the example above the question is how many is 253,020 compared to the total number of taxpayers in the state.  In this case, its about 5 percent, conversely more than 86 percent received a tax cut.  A single number never gives you the whole story.  Percentages and per capita data usually give a more accurate picture than a single raw number.  While disappointing, it is also helpful to determine whether or not the source of the number has an agenda.  In the above case it was a partisan organization.

To finish I would like to plug a book, written by Hans Rosling, a Swedish biostatistician. The title is Factfulness:  Ten Reasons We're Wrong About the World - and Why Things are Better Than You Think.  It covers how to interpret data and why most of our underlying assumptions, especially about the developing world, are wrong.  It's actually a pretty easy read and will change the way you gather and interpret information.  It also shows that we tend to be dumber than chimpanzees.