By now, you have probably heard people claim that Hillary Clinton mainly lost due to low turnout. It's been kind of hard to miss that line of thinking, since graphs such as this are getting shared several thousand times on Twitter:
This graph is just a public-facing example of a wider trend. Personally, as political professional, I have seen the claim that Clinton lost because of low turnout circulating pretty widely behind the scenes within progressive circles.
As an example of how widely it is believed among progressives that low turnout cost Clinton the election, Keith Ellison, a progressive champion who is the favorite to become DNC chair, made that claim in the email announcing his candidacy. In fact, Ellison stated that reversing low 2016 turnout was the very reason he was running for DNC chair. Here is a screenshot of that email:
Ellison claims that turnout in 2016 was the lowest it has been in 20 years--lower than 2012, 2008, 2004, and 2000.
But here is the thing--that is not true. Turnout in 2016 will likely be slightly higher than it was in 2012, much higher than 2000, and well above the historic average. Here are the numbers:
In 2012, the "voting eligible" population (that is, the number of citizens registered to vote or legally allowed to register) of the United States was 222,474,111. That year, there were 129,070,906 votes in the presidential election, for a voting eligible population turnout of 58.016%. (Source: United States Election Project)
In 2016, the voting eligible population of the United States is 231,556,622. As such, in order to equal 2012 turnout levels, there would need to be 134,340,229 votes in the 2016 presidential election. (Source: United States Election Project)
While the counting is not yet done, the current estimate for the total number of votes in the 2016 presidential election is 134,537,600. Thus, turnout in 2016 is likely a bit higher than it was in 2012. (Source: United States Election Project)
It is worth adding that turnout went up slightly in 2016 despite numerous obstacles being placed on voting in several states following the Supreme Court striking down key sections of the voting Rights Act in 2013.
Moving back to Ellison for a moment, here are other recent presidential elections for a comparison of turnout by voting eligible population (Source: United States Election Project):
Ellison's claim that 2016 represented a 20-year low in turnout is way off. In fact, 2016 it was closer to the post-1972 turnout peak of 2008 than it was to turnout in 2000. Further, while counting is still underway, it looks like 2016 will finish basically tied with 1992 (58.1% VEP turnout) and 2012 for the third highest turnout election since the passage of the 26th amendment.
In short, 2016 was not a low turnout election, and low turnout was not the cause of Clinton's defeat. Those claims are based on incomplete data and should be retracted.
It is possible that turnout was down among key demographics that lean toward Democrats, while up among demographics that are more favorable toward Republicans. That is a hypothesis that deserves more investigation. However, we are months away from having accurate information on that via the Census Bureau's Current Population Survey, so we should reserve judgement at this time.
I have been in politics for 14 years now, and I can tell you that there are a lot of progressives who want to believe that Democratic defeats are always caused by low turnout among an uninspired electorate. From where I sit, that is why the low turnout narrative took off--because it was what progressives wanted to believe. However, it is just not true in this case, and we are not going to start winning elections again by inventing our own facts in order to support our desired reality. That might be what the other side does, but it won't work for us.
Please share this post the next time you hear that Clinton lost because of low turnout, because that is just not true and should not be allowed to stand.
(Note: The United States Election Project lists 2012 VEP turnout for the presidential election as 58.6% here, and 58.0% here. However, as discussed above, doing the math shows that the number is 58.016%, or 58.0% rounded off).