So About That Popular Vote

Just under two weeks ago, in response to some protesters at an event of hers, Hillary Clinton shot back with a pretty serious and bold statement in which she claimed that “[she has] two and a half million more votes than Bernie Sanders.” This number seemed pretty high, so I felt the need to look into it (as everyone should do when they are unsure of the validity of a statement), and as of the time of her statement, according to, Clinton did in fact have a commanding 2,512,598 vote lead over Sanders. Her statement was valid. But is it really true that despite an increase in National Polls that show Sanders is neck and neck, if not ahead of Clinton, that she is really poised to dominate the popular vote by this much by the end of the race?

Not Exactly.

Despite wanting to write this post since shortly looking into the larger scope of this issue, the impetus really came from a recent article on fivethirtyeight. First, during their live stream of Wisconsin, one of their journalists made a comment about the popular vote total. I posted a bit of my information in the comment section, but as one would suspect, nothing came of it. Then, yesterday morning, the same journalist, David Wasserman, wrote a very detailed article about the “Will of the people” and the popular vote. And while I respect what fivethirtyeight does with their data, when comparing two things that are of different values (i.e. primaries v caucuses), different methods must be enacted. In an attempt at doing this, I have some calculations of my own that shows a different story.

After taking a deeper look at the popular vote argument, I have discovered that if Bernie Sanders can pull off a pledged delegate win (not an easy task), that he is actually poised to hold the popular vote lead, or at the very least, the Will of the People, by the end too. Here are three reasons why.

1. The Total from Green Papers Is Incomplete

And this might seem minor, but the Green Papers total notes that it excludes IA, NV, ME, and WA. While IA and NV were essentially ties vote-wise from what we know about them, ME and WA were a bit lopsided. So much so, that based on estimates of turnout (over 230,000 and 50,000 respectively) and the win margins in each state, this would add an additional 120,000 votes into Sanders’ column. Additionally, since Wisconsin voted earlier this week, Sanders shaved an additional 135,000 votes off of that deficit.

But this is only about 10% of the ginormous lead that Clinton has amassed; what else ya got?

2. Clinton’s Best States Have Already Voted

This is not news. The only reason that anyone has considered Sanders even remotely in the race at this point, is because it was pretty clear that the states that Clinton was supposed to do best in were heavy loaded into the front of the race (Before anyone cries foul, this is not some sign of manipulation by the party, but comes from a variety of factors that has nothing to do with who is running). This means, that if Clinton and Sanders were meant to be tied nationally, at this point, Clinton should have a popular vote lead (albeit not by 2.25 million). While I would have loved to try and figure out some way of predicting how much Sanders should gain in the remaining states, I do not have as much resources, information, nor the statistical acumen for this task like someone at fivethirtyeight does. Fortunately for me, Wasserman did just this in his article. Wasserman looked at what the popular vote change would be, if moving forward, Sanders acquired the right amount of delegates to secure a pledged delegate win at 2026-2025 delegates (presumably using Nate Silver’s adjusted goal chart). This chart shows that if Sanders were to come back and win, he would acquire about 1.7 million additional votes over Clinton. Bringing her actual lead down to about 525,000 votes (which is less than Bush lost by in the 2000 General Election against Gore).

Okay, so the lead isn’t as demanding as Clinton and many outlets make it seem, but she still has a popular vote lead, and therefore, the will of the people… right?

3. Primary v Caucus

Before we get into the statistical mess that this entails, I want to ask you to imagine this situation for me.

Imagine there are two classrooms, each with 50 students, that need to vote on a theme for prom. The person in charge of counting votes tells each class that they can decide on how to vote in any way they want, but that they must provide a count.

Classroom A decides to allow students to come in and drop off their vote anytime during the day into a box.

Classroom B thinks that there should be a discussion, and plans to meet after school to talk about the themes, and then vote on it.

Classroom A gets 20 of their students to drop off votes, whereas Classroom B gets 7 or so students to meet after school and discuss what theme best fits the desires of their class. These are the results:

A: 12-8     Theme 1
B: 5-2       Theme 2

Popular vote places Theme 1 as the winner (14-13). But is this really the will of the students? If these vote totals were compared to the population they represent, Theme 2 would win 55-45. Because Classroom B decided to have a discussion about how to vote before they voted, do they no longer represent the same amount of people?

These questions are not rhetorical. What constitutes the “Will of the people”? If it is simply popular vote, then why do we have an electoral college?

In states with a Caucus, which Sanders happens to do very well in, there is a much lower turnout than those with a Primary. My example from above? Those numbers were proportionally based off turnout and voting preference of Virginia and Washington respectively.

The delegate totals for each state is determined by their democratic voting record in the last three general presidential elections, and that state’s overall electoral total. Which means that they are proportional to how democratic each state is and their population. The Democratic Party created these rules to say that the will of the people nationally should be determined by representation, not by popular vote. Because of this, many states decide how to do this differently, and those states’ citizens’ opinions should not be considered less than others because of how they choose to vote. Caucuses have less turnout, but each vote represents thousands more than some primaries.

Take my home state of NH and adopted state of MN for example.

NH Democratic Primary 2016 = 256k
NH Democratic General Election Votes 2012 = 369k
or 66.67% of democratic general election voters voted in their primary

MN Democratic Caucus 2016 = 191k
MN Democratic General Election Votes 2012 = 1,546k
or 12.35% of democratic general election voters caucused in their caucus

It takes almost 5 NH votes to match 1 MN vote in delegate worth. If MN had the same turnout rate as NH, with its win margin, Sanders would have accumulated another 200,000 votes over Clinton. Had turnout in Washington been like NH? An additional 425,000 OVER the 104,000 that I awarded him earlier. Are you seeing a pattern? And, while in the general election WA may not have the same turnout at NH (65.8% to 70.9%), MN has one of the highest general election turnouts in the nation with 76.4%.

In an attempt at calculating this, I tried a few different measures. First, with data totaled after the WI primary, I calculated the AVERAGE ratio of turnout for primary states and compared it to their awarded delegates (the best representation for how democratic a state is). NH has a high primary turnout, and it wasn’t fair to compare each caucus to them so I felt the average was a better measure. In doing so, I discovered that had every caucus (not just the ones Sanders won) had the same turnout as the average primary, and maintained similar win margins, Clinton’s lead would be down to 1.6 million (this figure includes all races thus far), which when added to Wasserman’s goal sheet, would give Sanders a 100,000 representative popular vote lead over Clinton by the time that he wins the election IF he gets enough pledged delegates.

I am in no way saying that this is a guarantee, or even that it is the most likely of scenarios. But Wasserman’s article, as well as Clinton’s campaign, are misconstruing the information at hand by comparing two things that really have no business being compared. The Popular Vote is misinforming, and has zero impact on the general election, and anyone who tries to claim that the popular vote is the same as the will of the people without regarding the will of caucus states, is telling millions of voters in caucus states that their votes are worth less than the votes in primary states.

Are you really so sure that a popular vote count is an accurate portrayal of the will of the people? I’m not.


Hillary Clinton’s Claim:
Popular Vote Tally:
National Polls:
Washington State Turnout:
Maine Turnout:
538 Goal Voter Turnout:
Nate Silver’s Sanders’ Goals:
Determination of Democratic Delegates:
2012 Voter Turnout by State:


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s