Earned Electoral Allocation
Whether you like it or not, Americans are getting into the swing of another presidential election cycle! That means it’s time to prepare for phone calls, mailers, attack ads, corn dogs, and debates about the Electoral College!
Lots of people have opinions about the Electoral College and I’m sure you’ve heard them all. It’s out of date. It’s not representative. It gives small states too much power. It ruined the 2000 election. Whatever your gripe may be, the fact is it’s our system and politicians must navigate it. But that doesn’t mean we can’t do some analysis against it.
The number of votes that each state has in the Electoral College is based off of the census, which is taken once a decade. The larger a state’s population, the more votes it gets (and the more members the state gets in the House of Representatives). For all the problems one might have with the system, this arrangement does a pretty okay job of accomplishing its goal: distributing votes proportionally. Populous states like California do, in fact, have the most votes, while smaller states like Wyoming have the fewest.
Another Way
But here’s the rub: a state’s population doesn’t necessarily imply its voter turnout. Sure, California has the most residents, but what if only a few thousand people in the state bother to show up to vote? Does it still deserve all of those 55 Electoral votes? You might argue yes, that as long as the sample size is large enough, it could still be a statistically significant representation of the state. But it would depend on the spread.
What if states earned their standing in the Electoral College on election day? What if we made their Electoral College allocation based on the day’s turnout? There are plenty of hypotheses to be made around such an idea. Maybe it would encourage higher turnout as states compete to maintain their standing. Maybe certain states would lose some of their sway in deciding the election, while others emerge as new battlegrounds.
Some Real Data
These ideas are impossible to know without actually changing the rules. But we can look back at historic data to see how things might be different. Let’s start with 2012 (all of the data is available on GitHub).
It’s important to note that we’re going to keep the total number of Electoral votes at 538 in this thought-experiment. That will keep our data somewhat normalized when drawing comparisons.
If we were to adjust the results as described, we can come to a few immediate conclusions. First, we find that Wyoming is way over-allocated in the Electoral College. Wyoming voters made up 0.193% of the total popular vote in 2012. That would earn them a total of one electoral vote. In reality Wyoming gets three. You can thank Article I, Section 3, Clause 1 and Article II, Section 1, Clause 2 of the Constitution for that. Since Wyoming made out like a bandit, I colored them dark green.
So which state seemingly has the most to gain from our Earned Electoral Allocation system? Ohio. In reality, Ohio has 18 Electoral College votes — 3.3% of the total — but 2012 Ohio voters made up 4.3% of the popular vote. By that metric, Ohio would be allocated 23.3 (we’ll call it 23) Electoral College votes, or an increase of over 29%. As a consolation prize for Ohio’s bamboozlement, I colored them red.
Would this reallocation have impacted the 2012 election results? No. Barack Obama would have won by an even larger margin: 365–173, compared to the actual results of 332–206, over Mitt Romney.
Let’s Make Things More Interesting
But the 2012 election wasn’t really a close one. What if we apply this transformation to the 2000 election? Lest you forget, Al Gore won the popular vote in that election, but did not clinch the Electoral College, and thus lost the election.
In that election, Wyoming was still way over-allocated by our metric, with its voters making up 0.203% of the popular vote. Ohio was still under-allocated, but to a lesser extent (holding 21 of its earned 24 votes). In that election, Minnesota was the worst-represented, with a missed upside of 25% (it was granted 10 votes that year).
But would that be enough to tip the scales for an Al Gore victory? In fact, yes! With our Earned Electoral Allocation system, Al Gore would have seized the Electoral College in an equally close result of 270–268 over George W. Bush.
The Data
None of this is to say that our Earned Electoral Allocation is right; it’s just different. And thanks to Michael P. McDonald of The United States Elections Project, there are lots of ways we can bend the data. I’ve created a GitHub repository of the presidential elections from 2000–2012 with the Earned Electoral Allocations applied. Analyze, transform, and graph to your heart’s content.
