Reading Levels of the State of the Union Addresses

I can remember when I was little and discovered a feature of my word processor (Microsoft Works, I think) that told me the reading level of a document. When you’re in middle school, this almost felt like a dare: could I make my paper as erudite as possible? The scoring methods were opaque, and I’m […]

I can remember when I was little and discovered a feature of my word processor (Microsoft Works, I think) that told me the reading level of a document. When you're in middle school, this almost felt like a dare: could I make my paper as erudite as possible?

The scoring methods were opaque, and I'm pretty sure I thought it was based on word length, where sesquipedalianism was richly rewarded. One of the methods I can recall is the Flesch-Kincaid grade level, and is a bit more complicated than I thought, based on both the length of sentences as well as the number of syllables per word. This ignores more subtle features, such as the structure of the sentence, but let's assume it's not terrible.

So what can we do with it? Well, one thing that I've always thought would be interesting would be to take the American State of the Union addresses(here's a database of all of them) and run them through this little readability test. I am not the first one to think of this, as The Guardian also stumbled upon this ideas and has helpfully already carried out the analysis (a similar analysis has also been conducted by Brad Borevitz):

As you can see, the grade level of the addresses as decayed quite a bit. Now, no doubt some of this is due to changes in the style of the address over the years—most notably, speaking the address aloud instead of submitting it in a written form—rather than a diminishing intellectual capacity of the presidents or the electorate, but it is quite interesting. The president whose addresses have the highest reading level is James Madison, not entirely surprising given his work in drafting the United States Constitution. And the lowest is George H. W. Bush, influenced by his 1992 address which reached an all-time low reading level of 7.6.

These readability scores must be taken with a certain grain of salt but there other similar measures. My favorite, simply for the name alone, is the Simple Measure of Gobbledygook (or SMOG), which is often used to "measure the understandability of written information in clinical and health settings."

An interactive version of the above graphic is available here.

Top image:Jessica F./Flickr/CC