The day after Democrats got crushed in their second mid-term election in a row, President Obama said in a press conference he got the message.
“To everyone who voted, I want you to know that I hear you. To the two-thirds of voters who chose not to participate in the process yesterday, I hear you, too.”
Wow, two-thirds of voters didn’t vote?! If true, that would be a serious indictment of the system.
But it’s not true.
Calculating the percent of voters who voted should be no more difficult than a simple division problem: Take number of votes cast divided by number of registered voters, and you get turnout. The hardest part—projecting all the votes cast—has been helpfully done by the University of Florida’s Professor Michael P. McDonald and his United States Election Project. (You can check out his work here.)
The denominator is where it gets a little complicated. There are at least two different bases that can be used to calculate turnout, registered voters or the Voting Eligible Population (VEP). The VEP is roughly the number of adults nationwide minus those who are ineligible to vote (e.g. non-citizens and incarcerated felons). The Voting Eligible Population (VEP) base is the appropriate way to best measure election results over time, as Dr. McDonald discusses on his website, but in the everyday shorthand that has been adopted by the press and common use it is not a measurement of the turnout of “voters.”
Estimated Number of Votes Cast: 82.3 million
Estimated Number of Registered Voters: 153.2 million*
Estimated VEP: 227.2 million
Turnout Among RVs: 82.3 million ÷ 153.2 million = 53.8%
Turnout Among the VEP: 82.3 million ÷ 227.2 million = 36.2%
The President is wrong to say two-thirds of voters did not participate. While about two-thirds of the VEP did not participate, roughly 54% of registered voters did.
This year’s turnout was down slightly from 2010 (Professor McDonald counted 89.1 million votes cast for the highest office in each state four years ago), but three-fourths of the drop in turnout come from five states that account for just more than a quarter of the U.S. population. Compared to 2010, turnout was down by approximately 2.6 million in California, 800,000 in New York, 700,000 in Ohio, 500,000 in Missouri, and 500,000 in Pennsylvania. In addition to being relatively large, what these states had in common was no Senate election and either no Governor’s race or a non-competitive one.
In fact, due to a quirky luck of the draw, only 52% of the country lives in a state that did have a Senate election this year, by far the smallest percentage of any of the three Senate class elections. The lack of top-of-ticket races in this Senate class dampens turnout in a predictable manner: The last three times time this class of Senators was up for election in a non-presidential year—2002, 1990, and 1978—turnout among the VEP was closer to the 36.2% we saw this year—39.5%, 38.4%, and 39.0%, respectively. In fact, the lowest VEP turnout since the Great Depression occurred in 1942, another year this Senate class was up for election.
So, contrary to President Obama’s math, we did not see an unusually high number of voters sit this election out, and the drop in turnout we did measure can largely be attributed to the lack of competitive races in many of the most populous states.
Dr. McDonald’s work is always done with care and precision. His website makes abundantly clear the difference between the Voting Eligible Population and registered voters. It explains the reasons VEP is used, but, for goodness’ sakes, in popular lexicon it is time to retire the canard that “two-thirds of voters did not vote” or that this election represented a unique rejection of the political system.
*This estimate of registered voters comes from the U.S. Census, but is based on a 2012 estimate. It may not be a perfect number and, yes, some states allow Election Day Registration (EDR), but the point remains the same—a majority of registered voters actually voted in this election.