My Vote Should Count, but It Doesn’t, and I Can Prove It.
A Data Scientist’s Objection to the Electoral College
My vote for President is worth less than that of most Americans.
It’s un-democratic. It’s un-American. It’s absurd. But until I sat down with the numbers, I didn’t realize how plain the situation really is. In fact, the narrative I had encountered seemed to cast the Electoral College as archaic and quaint, but not patently unjust.
I hope the following 4 charts capture the heart of the matter. For me it was the missing part of the story.
A System Built for Smaller States
As a pillar of democracy—if not common sense—each of us as citizens should be afforded one equal vote in the election of our President. Plain as day. The Electoral College obscures this in most states, aggregating all of a state’s voting power and casting it en mass to the state’s winning candidate.
Fair enough, if electoral votes were allocated equitably based on population, but this is not the case. Let’s start by simply comparing state populations to the number of electoral votes assigned to each state.
- Winner: Wyoming gets 1 vote for every 192k citizens
- Loser: California gets 1 vote for every 692k citizens
- Alarming: Smaller states have up to 3.6x more voting power
Interpreting the chart above, a lower number on the green line would be preferable, indicating that electoral votes are spread over fewer people.
Indeed, the Electoral College was designed to ensure that states with smaller populations still had a say in federal elections, and clearly this is still “working,” as you can see smaller population states have more of a say than larger states. A citizen of Wyoming, in fact, has 3.6 times the raw voting power of a citizen of California.
This was designed as a grand compromise, but it seems to be grossly out of proportion.
A System Delivering for Swing States
Despite the intentions of the system’s architects, we could hardly argue that our national elections come down to either California or Wyoming.
This is where Margin of Victory plays such an important role. States with tremendous margins of victory are difficult to sway (read “safe”), while those with smaller margins of victory “swing” large tranches of electoral votes from one side to another over disproportionately small changes in voting behavior.
To wit, New Hampshire and Hawaii both have 4 electoral votes and comparable populations around 1.3M. In the 2012 Presidential election, New Hampshire was decided by a margin of less than 40k (in play as a swing state), while Hawaii was decided by over 185k (ignored as a safe state). As a candidate, you would have to change the minds of 4.6 times as many people in Hawaii to secure the same volume of electoral votes. Plus, it’s a long flight.
Let’s quantify this. If we define “voting power” as margin of victory per electoral vote, we are really calculating the number of people whose position a candidate must sway in order to win 1 electoral vote. The story is illustrated starkly by applying this measure to the 2012 results for each state.
- Winner: Florida can move an electoral vote with only 2,562 voters
- Loser: DC needs 81,896 voters to move an electoral vote
- Outrageous: Swing states have up to 30.96x more voting power
Great news for swing states! Interpreting the chart above, a lower number on the green line would be preferable, indicating that changes in electoral votes occur with changes in fewer citizens’ votes. Not unrelated: The left side looks like a candidate tour schedule for October, whereas states on the right see almost no campaigning at all. Absurdly, by this measure a Florida citizen has 30.96x more voting power than a D.C. citizen.
Same As It Ever Was
This is not a new problem. Let’s put all this in a historical context.
Looking at the last 150 years of Presidential elections, the overall margin of victory for the popular vote varies anywhere from 0.10% on the low end (Garfield ekes one out in 1880) to 26.24% on the high end (landslide for Harding in 1920), though over time it averages around 9.53%. This is the green area in the chart above.
If we characterize “very safe state” voters as those in states whose margin of victory in the election exceed 20%, we have a very reasonable classification of disenfranchised voters we can observe over time.
This poor flock, shown in the blue area in the chart above, accounted for an average of 25.82% of the popular vote over the last century and a half.
That is correct: Year after year, over a quarter of our voters stand negligible chance of moving the electoral needle. Fully a quarter of our democracy has an empty voice in selecting its President. Worse, these voters have outnumbered the election’s overall margin of victory nearly 3-to-1. As the data above shows, this muted quartile could very well have changed the face of each election in the modern era.
All this to say, the system broken, it is broken with material consequences, and it has been for a long time.
Fixing the Problem
The solution is simple: The National Popular Vote bill would guarantee the Presidency to the presidential candidate who receives the most popular votes in all 50 states (and the District of Columbia). Done and done.
The bill requires ratification by states comprising 270 Electoral Votes. With New York joining this month, the bill has 165 already (i.e. we are now 61% there).
Let’s look at the half of our states that we’ve identified as under-represented:
2 light blue columns voted Democrat in 2012 but have not signed the bill.
13 light red columns voted Republican in 2012 but have not signed the bill.
Dots
indicate electoral votes.
This chart requires a bit of unpacking. We are looking at the 26 states with the least voting power (as measured by Margin of Victory / Electoral Votes). This is the value we are plotting in the columns themselves. The dots are the number of electoral votes each of the states carry. Finally, blue indicates states that voted Democrat in 2012, whereas red went Republican.
So, what do we observe?
States with Voting Power Don’t Mind the Electoral College
As we expect, you don’t see Florida on the list. Or Iowa, North Carolina, Virginia, Pennsylvania, or Ohio. The swing states not only get attention, but also funding for the projects they want. It’s good to be king.
So far, all 11 states that have signed the bill are all from states the Electoral College leaves behind, and this makes sense.
Thus Far, Red States Don’t Mind the Electoral College
The surprise: So far, the 11 states that have signed the bill all voted Democrat in 2012. Looking at the chart above, the dark blue columns represent states that have signed the bill. Of the blue states in this group, only Connecticut and Oregon remain.
The light red columns represent states that voted Republican in the last election. Utah, Oklahoma, Idaho, Alabama, and Kentucky—perennially disregarded by candidates—would all benefit from a national popular vote, and yet not one has signed on.
This is all the more surprising given the bill’s bipartisan advisory board, which includes former Utah Senator Jake Garn and former Alabama Congressman John Buchanan, and whose informal supporters include high profile conservatives including Newt Gingrich and Fred Thompson.
I am not in the business of speculating as to why underrepresented red states haven’t pushed for a national popular vote. In fact, I have no interest in alienating people or pointing fingers. I just want my vote!
Nor, as it happens, am I inclined contemplate the potential benefits of giving a minority of the citizens in our country a disproportionate say in the election. I am just a data guy. And to me, the data speaks clearly:
- The Electoral College is creating material iniquities in voter power and has been doing so for at least 150 years.
- Blue states bearing the brunt of the iniquities have signed the National Popular Vote bill to resolve this.
- Red states bearing the brunt of the iniquities have not yet done so.
The bill needs another 105 electoral votes. Connecticut and Oregon make 14, and Michigan is right behind with another 16. Perhaps Minnesota adds their 10. If needed, you could pick up 40 there. But clearly the path forward requires bringing underrepresented red states to the table.
If you’re reading this as a citizen of one such state, I hope you reach out to you legislators to right the ship and help strengthen our democracy. This is your chance to have a voice.
In either case, if you found the article useful, I would humbly appreciate your recommendation, your sharing with others, and even your comments, if you think I’ve gotten it wrong or would like me to clarify.
- Data sources: David Liep’s Atlas of Presidential Elections, US Census
- Analysis source code (please feel free to fork and adapt!)
- Font used in charts: Concourse (300 weight, small caps)
