Monday, October 23, 2006

Turnout vs. Age: Who's gonna vote on November 7th?

I recently needed to know US voter turnout rates verses age for an analysis I did of Governor Doyle's popularity. A quick web search turned up answers that were all over the map, with few people citing their sources or giving historical information. So I decided to do the research myself and make it available to my many loyal readers. I'm also going to post it on Daily Kos, just to get a little more exposure...

Follow me down for some useful and interesting information on trends in US voter turnout verses age over the last thirty years.


(Click on chart for larger version)

The chart above uses information from the US Census Bureau as published in its Statistical Abstracts for 1988, 1992, and 2006. It covers only presidential years, although I may also do midterms in the future. It would be interesting to see if midterms follow the same age patterns as presidential years.

A couple of things are obvious from the data. The older you are the more likely you are to vote, and this is a very strong effect. The 65+ crowd is 1.6 times more likely to vote than those 24 or under. And while overall voter turnout started trending upward in 2000, young voters have increased their turnout more than any other group, by nearly 50% in the 18-20 year old group.

The table below combines the voting rates above with population data also from the 2006 Statistical Abstract to calculate what percentage of 2004 voters were represented by each age group. As you can see, older groups are the dominant voters in the US, with a solid majority of votes coming from those 45 and over, and about 3/4 of the total votes from people over 35.


(Click on table for larger version)

Caveats: While the Statical Abstracts are great information sources, you have to keep in mind that they are the result of surveys, and have the same flaws as all surveys. For example, the 2006 Statistical Abstract reports a 4% larger voter turnout in 2004 than there actually was. Also, I ran into obvious typographical errors in the Statistical Abstracts when I was doing my research, so it's wise to cross check data whenever possible.

No comments: