So a series of news articles are being spread around that purportedly attributes some of Bernie Sanders' losses in states due to low voter turnout. A quick search on Google reports tens of articles all claiming the same or similar findings:

However, I recently saw an article on Reddit that says this claim is being falsified by purposely misattributing one statistic as another, but I don't know how verifiable the information is considering it's not a news source I'm familiar with.

If anything, I am more wary of the quantity of news reports. Earlier this month, Snopes.com published a fact-checking article disproving a wide swathe of news reports that Corona beer sales dropped due to fear about the coronavirus, and this is something I am not sure is also going on with the Sanders/youth voter turnout issue.

In any case, did low youth turnout affect Bernie Sanders' results in the Super Tuesday primaries? Or, as the 2nd linked article claims, is it not true?

  • Do states actually keep statistics on the age of people who voted? Seems like an invasion of privacy to me.
    – jamesqf
    Commented Mar 8, 2020 at 6:37
  • 1
    @jamesqf There's always exit polls. There's never any shortage of detailed demographic breakdowns on voting. Commented Mar 8, 2020 at 10:13

2 Answers 2


There's a lot of spin going on in all camps, including the Sanders camp. This is par for the course in a political election year.

The article you cited is guilty of the same criticisms it makes of other articles. It fails to provide any concise information about how many voters in the 18-29 age range in 2020 turned out compared to 2016.

In fairness, this is partly because election officials do not track voters by age when they vote. So the only way to know anything about this kind of metric is to look at the same exit polls the article rejects as inaccurate.

The Sanders campaign, like other campaigns, do their own polling and sampling. So they have pretty much the best idea of anyone as to how well they've really done. So listen to what Bernie himself said:

"Have we been as successful as I would hope in bringing in young people in? And the answer is 'no'," he told reporters at a press briefing on Wednesday.

"We're making some progress but historically everybody knows that young people do not vote in the kind of numbers that older people vote," Sanders went on. "I think that will change in the general election. But to be honest with you, we have not done as well in bringing young people into the process. It is not easy."

Source: https://www.businessinsider.com/bernie-sanders-admits-hes-not-inspiring-enough-young-voters-2020-3

However, that's not the only thing we can infer from available data. If you put together a spreadsheet state-by-state to tally vote totals compared to votes received by Sanders, you will see a strong pattern. The data shows that Sanders' overall vote count is running 20-30% LOWER than in 2016 in most states that have voted so far, despite the fact that across-the-board vote count is running 10-30% HIGHER than in 2016. Sanders defenders argue that this is because the field was so much bigger and stronger than in 2016, but I see this as a throwaway argument. If Sanders was really a strong vote-getter, he wouldn't be bleeding support compared to 2016 because the ones who voted for him then already knew him, whereas his opponents this year were mostly newcomers to the primaries. He already had name recognition that the others didn't have among Democratic voters.

So what does this broader insight tell us? Well, all of the polls we've seen say that Sanders' support levels are strongest among the 18-29 age range. So if he's hemorrhaging that much voter support in the overall numbers, it either means that the number of voters in the 18-29 range matched 2016 while the rest of his supporters disappeared almost completely into the woodwork, or (much more likely) his support among 18-29 year olds sagged compared to four years ago along with the other age groups.

Given Sanders' remarks Business Insider quoted, I'd go with the latter likelihood.

  • Please clarify whether "LOWER... HIGHER" refers to proportions, (i.e. three to five in 2016, two to five in 2020, or whatever -- those last two proportions are made up), or to actual numbers, (i.e. 2000000 votes in 2016, 1500000 in 2020, or whatever -- those last two numbers are also made up), or both.
    – agc
    Commented Mar 10, 2020 at 16:10
  • @agc By definition, a percentage is a proportion because it is an expression of a ratio. So saying 10-30% higher states proportion. To calculate proportion, take the difference between his 2020 and his 2016 vote counts for a given state, then divide the result by his 2016 total vote count and multiply that result by 100 to get the percentage. So for example, in 2020 Sanders received 80,121 votes in Vermont, and in 2016 he received 115,900 votes in Vermont. 115,900 - 80,121 = 35,779. Now divide by 80,121 and you get a ratio of 0.44655. Multiply by 100 and you get 44.655% lower in 2020. Commented Mar 10, 2020 at 17:44
  • It would help to include any relevant aggregate vote count numbers in the answer. Regarding your calculation of Sanders' VT decrease, it's suggestive yet inconclusive because it neglects to factor in a comparison of how many primary votes were cast for all candidates in those two cycles. That is, if Sanders 2020 seems to be down by 35K votes, did those go to Biden, Trump, (whether by defectors or cross-voting), or some third party, or did those people stay home, or did they move to some other state, ... etc.?
    – agc
    Commented Mar 10, 2020 at 18:18
  • @agc I disagree. If those voters voted for Sanders in 2016, but failed to do so in 2020, then it's incumbent upon his defenders to explain why he lost them. Blaming the loss on other competitors is basically another way of saying that they really never supported Sanders in the first place because they were just casting protest votes against Clinton. Besides, it's not up to me to do all of the data collection work. That's yuritsuki's job as the question asker if he wishes to look more closely at it. The data is widely available with a simple Google search. Commented Mar 10, 2020 at 19:35
  • @agc Interesting follow-up to our discussion. Tonight is the second major set of Tuesday primaries after Super Tuesday, this time in Michigan, Missouri, Mississippi, North Dakota, Idaho, and Washington. So far, despite the fact that he is now down to just one opponent (Biden), Sanders is losing a LOT of support compared to 4 years ago. Commented Mar 11, 2020 at 1:13

I think you're misunderstanding the issue. I've not seen anything indicating there was a significantly lower turnout among young voters. The issue is that Bernie Sanders predicted he was going to turn out larger numbers of young voters

It is the most politically provocative part of Senator Bernie Sanders’s campaign pitch: that his progressive movement will bring millions of nonvoters into the November election, driving record turnout especially among disaffected working-class Americans and young people.

As Super Tuesday proved, that did not happen

Consider Texas: According to NBC News’s exit polls, the Democratic electorate actually skewed older in Tuesday’s primary compared to past primaries. In 2008 and 2016, 13 and 18 percent of the electorate, respectively, was 65 and older. In 2020, it was 24 percent.

And this report

But Sanders' actual younger-than-30 turnout isn't quite living up to that promise. While Sanders is winning big margins among young voters, they aren't making up significantly higher shares of the electorate than in past elections.

Before Tuesday, voters younger than 30 were not keeping pace with the overall increase in voter turnout. In fact, young voters’ share of the electorate went down in three of the first four states compared with 2016.

On Tuesday night, not a single state saw an increase in young voters’ share of the electorate, according to exit polls conducted by Edison Research and sponsored by several of the television networks.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .